{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 2133,
  "global_step": 88820,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002420625985138482,
      "grad_norm": 9.723807334899902,
      "learning_rate": 4.029990627928772e-05,
      "loss": 4.6027,
      "step": 43
    },
    {
      "epoch": 0.004841251970276964,
      "grad_norm": 1.4147453308105469,
      "learning_rate": 8.059981255857544e-05,
      "loss": 3.9392,
      "step": 86
    },
    {
      "epoch": 0.007261877955415447,
      "grad_norm": 1.4368598461151123,
      "learning_rate": 0.00012089971883786317,
      "loss": 3.6142,
      "step": 129
    },
    {
      "epoch": 0.009682503940553928,
      "grad_norm": 3.807770252227783,
      "learning_rate": 0.0001611996251171509,
      "loss": 3.4853,
      "step": 172
    },
    {
      "epoch": 0.012103129925692412,
      "grad_norm": 1.006263017654419,
      "learning_rate": 0.00020149953139643863,
      "loss": 3.4273,
      "step": 215
    },
    {
      "epoch": 0.014523755910830894,
      "grad_norm": 1.2452696561813354,
      "learning_rate": 0.00024179943767572634,
      "loss": 3.3955,
      "step": 258
    },
    {
      "epoch": 0.016944381895969376,
      "grad_norm": 8.02540111541748,
      "learning_rate": 0.00028209934395501406,
      "loss": 3.3753,
      "step": 301
    },
    {
      "epoch": 0.019365007881107857,
      "grad_norm": 6.356571674346924,
      "learning_rate": 0.0003223992502343018,
      "loss": 3.3527,
      "step": 344
    },
    {
      "epoch": 0.02178563386624634,
      "grad_norm": 6.280053615570068,
      "learning_rate": 0.00036269915651358954,
      "loss": 3.3288,
      "step": 387
    },
    {
      "epoch": 0.024206259851384825,
      "grad_norm": 6.516995429992676,
      "learning_rate": 0.00040299906279287726,
      "loss": 3.3172,
      "step": 430
    },
    {
      "epoch": 0.026626885836523305,
      "grad_norm": 3.9377992153167725,
      "learning_rate": 0.0004432989690721649,
      "loss": 3.2898,
      "step": 473
    },
    {
      "epoch": 0.02904751182166179,
      "grad_norm": 3.169243335723877,
      "learning_rate": 0.0004835988753514527,
      "loss": 3.2728,
      "step": 516
    },
    {
      "epoch": 0.03146813780680027,
      "grad_norm": 4.908263206481934,
      "learning_rate": 0.0005238987816307405,
      "loss": 3.292,
      "step": 559
    },
    {
      "epoch": 0.03388876379193875,
      "grad_norm": 2.3755059242248535,
      "learning_rate": 0.0005641986879100281,
      "loss": 3.2715,
      "step": 602
    },
    {
      "epoch": 0.03630938977707723,
      "grad_norm": 2.9217305183410645,
      "learning_rate": 0.0006044985941893159,
      "loss": 3.2886,
      "step": 645
    },
    {
      "epoch": 0.038730015762215714,
      "grad_norm": 2.473240613937378,
      "learning_rate": 0.0006447985004686035,
      "loss": 3.2767,
      "step": 688
    },
    {
      "epoch": 0.0411506417473542,
      "grad_norm": 1.9684624671936035,
      "learning_rate": 0.0006850984067478912,
      "loss": 3.2678,
      "step": 731
    },
    {
      "epoch": 0.04357126773249268,
      "grad_norm": 0.4937862455844879,
      "learning_rate": 0.0007253983130271791,
      "loss": 3.2665,
      "step": 774
    },
    {
      "epoch": 0.04599189371763116,
      "grad_norm": 1.2119194269180298,
      "learning_rate": 0.0007656982193064667,
      "loss": 3.2553,
      "step": 817
    },
    {
      "epoch": 0.04841251970276965,
      "grad_norm": 1.4475995302200317,
      "learning_rate": 0.0008059981255857545,
      "loss": 3.2512,
      "step": 860
    },
    {
      "epoch": 0.05083314568790813,
      "grad_norm": 1.0512193441390991,
      "learning_rate": 0.0008462980318650422,
      "loss": 3.2611,
      "step": 903
    },
    {
      "epoch": 0.05325377167304661,
      "grad_norm": 0.4626305103302002,
      "learning_rate": 0.0008865979381443298,
      "loss": 3.2393,
      "step": 946
    },
    {
      "epoch": 0.05567439765818509,
      "grad_norm": 0.32701531052589417,
      "learning_rate": 0.0009268978444236177,
      "loss": 3.2347,
      "step": 989
    },
    {
      "epoch": 0.05809502364332358,
      "grad_norm": 0.31743124127388,
      "learning_rate": 0.0009671977507029054,
      "loss": 3.1999,
      "step": 1032
    },
    {
      "epoch": 0.06051564962846206,
      "grad_norm": 0.878101110458374,
      "learning_rate": 0.001,
      "loss": 3.1991,
      "step": 1075
    },
    {
      "epoch": 0.06293627561360055,
      "grad_norm": 0.851373016834259,
      "learning_rate": 0.001,
      "loss": 3.1926,
      "step": 1118
    },
    {
      "epoch": 0.06535690159873903,
      "grad_norm": 0.32422763109207153,
      "learning_rate": 0.001,
      "loss": 3.1776,
      "step": 1161
    },
    {
      "epoch": 0.0677775275838775,
      "grad_norm": 0.7243884205818176,
      "learning_rate": 0.001,
      "loss": 3.1778,
      "step": 1204
    },
    {
      "epoch": 0.07019815356901599,
      "grad_norm": 0.37140366435050964,
      "learning_rate": 0.001,
      "loss": 3.1871,
      "step": 1247
    },
    {
      "epoch": 0.07261877955415447,
      "grad_norm": 0.31872037053108215,
      "learning_rate": 0.001,
      "loss": 3.1752,
      "step": 1290
    },
    {
      "epoch": 0.07503940553929295,
      "grad_norm": 1.065709114074707,
      "learning_rate": 0.001,
      "loss": 3.1657,
      "step": 1333
    },
    {
      "epoch": 0.07746003152443143,
      "grad_norm": 0.2842984199523926,
      "learning_rate": 0.001,
      "loss": 3.1695,
      "step": 1376
    },
    {
      "epoch": 0.07988065750956992,
      "grad_norm": 0.3832158148288727,
      "learning_rate": 0.001,
      "loss": 3.1505,
      "step": 1419
    },
    {
      "epoch": 0.0823012834947084,
      "grad_norm": 0.33350056409835815,
      "learning_rate": 0.001,
      "loss": 3.1504,
      "step": 1462
    },
    {
      "epoch": 0.08472190947984688,
      "grad_norm": 0.2758004665374756,
      "learning_rate": 0.001,
      "loss": 3.1451,
      "step": 1505
    },
    {
      "epoch": 0.08714253546498536,
      "grad_norm": 0.2981252074241638,
      "learning_rate": 0.001,
      "loss": 3.1562,
      "step": 1548
    },
    {
      "epoch": 0.08956316145012384,
      "grad_norm": 0.29148155450820923,
      "learning_rate": 0.001,
      "loss": 3.1444,
      "step": 1591
    },
    {
      "epoch": 0.09198378743526232,
      "grad_norm": 0.925895094871521,
      "learning_rate": 0.001,
      "loss": 3.1328,
      "step": 1634
    },
    {
      "epoch": 0.0944044134204008,
      "grad_norm": 5.448670387268066,
      "learning_rate": 0.001,
      "loss": 3.1479,
      "step": 1677
    },
    {
      "epoch": 0.0968250394055393,
      "grad_norm": 0.33397766947746277,
      "learning_rate": 0.001,
      "loss": 3.1432,
      "step": 1720
    },
    {
      "epoch": 0.09924566539067778,
      "grad_norm": 0.28908681869506836,
      "learning_rate": 0.001,
      "loss": 3.1483,
      "step": 1763
    },
    {
      "epoch": 0.10166629137581626,
      "grad_norm": 0.296283096075058,
      "learning_rate": 0.001,
      "loss": 3.1447,
      "step": 1806
    },
    {
      "epoch": 0.10408691736095474,
      "grad_norm": 0.28633320331573486,
      "learning_rate": 0.001,
      "loss": 3.1399,
      "step": 1849
    },
    {
      "epoch": 0.10650754334609322,
      "grad_norm": 0.42049241065979004,
      "learning_rate": 0.001,
      "loss": 3.1328,
      "step": 1892
    },
    {
      "epoch": 0.1089281693312317,
      "grad_norm": 0.41925516724586487,
      "learning_rate": 0.001,
      "loss": 3.124,
      "step": 1935
    },
    {
      "epoch": 0.11134879531637018,
      "grad_norm": 0.2977161705493927,
      "learning_rate": 0.001,
      "loss": 3.1214,
      "step": 1978
    },
    {
      "epoch": 0.11376942130150867,
      "grad_norm": 0.36108505725860596,
      "learning_rate": 0.001,
      "loss": 3.1223,
      "step": 2021
    },
    {
      "epoch": 0.11619004728664716,
      "grad_norm": 0.2616022527217865,
      "learning_rate": 0.001,
      "loss": 3.1217,
      "step": 2064
    },
    {
      "epoch": 0.11861067327178564,
      "grad_norm": 0.3144634962081909,
      "learning_rate": 0.001,
      "loss": 3.1077,
      "step": 2107
    },
    {
      "epoch": 0.120074307588381,
      "eval_ag_news_bleu_score": 3.5988036588204895,
      "eval_ag_news_bleu_score_sem": 0.13683642710467986,
      "eval_ag_news_emb_cos_sim": 0.9901920557022095,
      "eval_ag_news_emb_cos_sim_sem": 0.0002824538123635439,
      "eval_ag_news_emb_top1_equal": 0.6299999952316284,
      "eval_ag_news_emb_top1_equal_sem": 0.027921293391044915,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 4.186891555786133,
      "eval_ag_news_n_ngrams_match_1": 4.452,
      "eval_ag_news_n_ngrams_match_2": 0.752,
      "eval_ag_news_n_ngrams_match_3": 0.198,
      "eval_ag_news_num_pred_words": 23.99,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.22800393279490827,
      "eval_ag_news_runtime": 16.2173,
      "eval_ag_news_samples_per_second": 30.831,
      "eval_ag_news_steps_per_second": 0.123,
      "eval_ag_news_token_set_f1": 0.2090330996304002,
      "eval_ag_news_token_set_f1_sem": 0.004938500594480564,
      "eval_ag_news_token_set_precision": 0.19835261305625249,
      "eval_ag_news_token_set_recall": 0.2283663871931453,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 2133
    },
    {
      "epoch": 0.120074307588381,
      "eval_anthropic_toxic_prompts_bleu_score": 4.284504115991871,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.18242524496076415,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9898161888122559,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00028848800877239426,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6399999856948853,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02775911810844162,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.3968095779418945,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.092,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.114,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.404,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.606,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.23592763524982738,
      "eval_anthropic_toxic_prompts_runtime": 6.6187,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.543,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.302,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2792294503681125,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006204854063466297,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.32116096135564215,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2755568003237843,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 2133
    },
    {
      "epoch": 0.120074307588381,
      "eval_arxiv_bleu_score": 3.2745132822420198,
      "eval_arxiv_bleu_score_sem": 0.10154372163943692,
      "eval_arxiv_emb_cos_sim": 0.9895544648170471,
      "eval_arxiv_emb_cos_sim_sem": 0.00031133135864745465,
      "eval_arxiv_emb_top1_equal": 0.49666666984558105,
      "eval_arxiv_emb_top1_equal_sem": 0.028915102862112945,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 4.14932107925415,
      "eval_arxiv_n_ngrams_match_1": 4.292,
      "eval_arxiv_n_ngrams_match_2": 0.582,
      "eval_arxiv_n_ngrams_match_3": 0.1,
      "eval_arxiv_num_pred_words": 18.31,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.22075669986639473,
      "eval_arxiv_runtime": 7.2124,
      "eval_arxiv_samples_per_second": 69.325,
      "eval_arxiv_steps_per_second": 0.277,
      "eval_arxiv_token_set_f1": 0.21496146405667527,
      "eval_arxiv_token_set_f1_sem": 0.0045674330729410535,
      "eval_arxiv_token_set_precision": 0.17846752873752172,
      "eval_arxiv_token_set_recall": 0.29248740235363296,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 2133
    },
    {
      "epoch": 0.120074307588381,
      "eval_python_code_alpaca_bleu_score": 4.447988340670456,
      "eval_python_code_alpaca_bleu_score_sem": 0.1422183678899435,
      "eval_python_code_alpaca_emb_cos_sim": 0.9878453612327576,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00046002157198417093,
      "eval_python_code_alpaca_emb_top1_equal": 0.6299999952316284,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027921293391044915,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.5153965950012207,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.006,
      "eval_python_code_alpaca_n_ngrams_match_2": 0.95,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.264,
      "eval_python_code_alpaca_num_pred_words": 20.326,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.27046534238720343,
      "eval_python_code_alpaca_runtime": 6.5583,
      "eval_python_code_alpaca_samples_per_second": 76.239,
      "eval_python_code_alpaca_steps_per_second": 0.305,
      "eval_python_code_alpaca_token_set_f1": 0.29997801753987247,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005085291601135751,
      "eval_python_code_alpaca_token_set_precision": 0.2922904806182288,
      "eval_python_code_alpaca_token_set_recall": 0.3404390094702117,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 2133
    },
    {
      "epoch": 0.120074307588381,
      "eval_wikibio_bleu_score": 3.5833690367276794,
      "eval_wikibio_bleu_score_sem": 0.1250208779395924,
      "eval_wikibio_emb_cos_sim": 0.9906015992164612,
      "eval_wikibio_emb_cos_sim_sem": 0.00029099565029545204,
      "eval_wikibio_emb_top1_equal": 0.7200000286102295,
      "eval_wikibio_emb_top1_equal_sem": 0.025966275374921653,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.198541164398193,
      "eval_wikibio_n_ngrams_match_1": 3.124,
      "eval_wikibio_n_ngrams_match_2": 0.782,
      "eval_wikibio_n_ngrams_match_3": 0.164,
      "eval_wikibio_num_pred_words": 21.308,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2434664824094146,
      "eval_wikibio_runtime": 6.4077,
      "eval_wikibio_samples_per_second": 78.031,
      "eval_wikibio_steps_per_second": 0.312,
      "eval_wikibio_token_set_f1": 0.19842533052350697,
      "eval_wikibio_token_set_f1_sem": 0.004997321389778937,
      "eval_wikibio_token_set_precision": 0.23712279827170935,
      "eval_wikibio_token_set_recall": 0.17767859722358414,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 2133
    },
    {
      "epoch": 0.120074307588381,
      "eval_nq_5round_bleu_score": 8.351402571239474,
      "eval_nq_5round_bleu_score_sem": 0.38654799678007434,
      "eval_nq_5round_emb_cos_sim": 0.9915448427200317,
      "eval_nq_5round_emb_cos_sim_sem": 0.0002740306177908206,
      "eval_nq_5round_emb_top1_equal": 0.596666693687439,
      "eval_nq_5round_emb_top1_equal_sem": 0.028370197491492683,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.368,
      "eval_nq_5round_n_ngrams_match_2": 2.686,
      "eval_nq_5round_n_ngrams_match_3": 1.134,
      "eval_nq_5round_num_pred_words": 24.22,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3252025914407234,
      "eval_nq_5round_token_set_f1": 0.3706483928451811,
      "eval_nq_5round_token_set_f1_sem": 0.006096572296879725,
      "eval_nq_5round_token_set_precision": 0.33587436050160324,
      "eval_nq_5round_token_set_recall": 0.43386109190214545,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.752232123492721,
      "eval_nq_bleu_score_sem": 0.3924947841821986,
      "eval_nq_emb_cos_sim": 0.9920123815536499,
      "eval_nq_emb_cos_sim_sem": 0.0002490647266594065,
      "eval_nq_emb_top1_equal": 0.6133333444595337,
      "eval_nq_emb_top1_equal_sem": 0.028163139369651306,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.861788272857666,
      "eval_nq_n_ngrams_match_1": 8.74,
      "eval_nq_n_ngrams_match_2": 2.828,
      "eval_nq_n_ngrams_match_3": 1.2,
      "eval_nq_num_pred_words": 24.156,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3428534706234191,
      "eval_nq_runtime": 18.6643,
      "eval_nq_samples_per_second": 26.789,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.3829739342674168,
      "eval_nq_token_set_f1_sem": 0.005932648524783356,
      "eval_nq_token_set_precision": 0.3520807475470799,
      "eval_nq_token_set_recall": 0.4347611845923932,
      "eval_nq_true_num_tokens": 32.0,
      "step": 2133
    },
    {
      "epoch": 0.12103129925692412,
      "grad_norm": 0.30394721031188965,
      "learning_rate": 0.001,
      "loss": 3.108,
      "step": 2150
    },
    {
      "epoch": 0.1234519252420626,
      "grad_norm": 0.28194472193717957,
      "learning_rate": 0.001,
      "loss": 3.109,
      "step": 2193
    },
    {
      "epoch": 0.1258725512272011,
      "grad_norm": 0.3140788972377777,
      "learning_rate": 0.001,
      "loss": 3.1022,
      "step": 2236
    },
    {
      "epoch": 0.12829317721233957,
      "grad_norm": 0.28860148787498474,
      "learning_rate": 0.001,
      "loss": 3.1053,
      "step": 2279
    },
    {
      "epoch": 0.13071380319747805,
      "grad_norm": 0.25884610414505005,
      "learning_rate": 0.001,
      "loss": 3.1095,
      "step": 2322
    },
    {
      "epoch": 0.13313442918261653,
      "grad_norm": 0.2748049199581146,
      "learning_rate": 0.001,
      "loss": 3.104,
      "step": 2365
    },
    {
      "epoch": 0.135555055167755,
      "grad_norm": 0.2905972898006439,
      "learning_rate": 0.001,
      "loss": 3.1061,
      "step": 2408
    },
    {
      "epoch": 0.1379756811528935,
      "grad_norm": 0.26175904273986816,
      "learning_rate": 0.001,
      "loss": 3.1001,
      "step": 2451
    },
    {
      "epoch": 0.14039630713803197,
      "grad_norm": 0.3153071999549866,
      "learning_rate": 0.001,
      "loss": 3.101,
      "step": 2494
    },
    {
      "epoch": 0.14281693312317045,
      "grad_norm": 0.2775098979473114,
      "learning_rate": 0.001,
      "loss": 3.1088,
      "step": 2537
    },
    {
      "epoch": 0.14523755910830893,
      "grad_norm": 0.2857923209667206,
      "learning_rate": 0.001,
      "loss": 3.1104,
      "step": 2580
    },
    {
      "epoch": 0.1476581850934474,
      "grad_norm": 0.3178844153881073,
      "learning_rate": 0.001,
      "loss": 3.1018,
      "step": 2623
    },
    {
      "epoch": 0.1500788110785859,
      "grad_norm": 0.2835923433303833,
      "learning_rate": 0.001,
      "loss": 3.1398,
      "step": 2666
    },
    {
      "epoch": 0.15249943706372437,
      "grad_norm": 0.6011675596237183,
      "learning_rate": 0.001,
      "loss": 3.1098,
      "step": 2709
    },
    {
      "epoch": 0.15492006304886285,
      "grad_norm": 0.2724011242389679,
      "learning_rate": 0.001,
      "loss": 3.1069,
      "step": 2752
    },
    {
      "epoch": 0.15734068903400136,
      "grad_norm": 0.32648223638534546,
      "learning_rate": 0.001,
      "loss": 3.1056,
      "step": 2795
    },
    {
      "epoch": 0.15976131501913984,
      "grad_norm": 0.27339959144592285,
      "learning_rate": 0.001,
      "loss": 3.0848,
      "step": 2838
    },
    {
      "epoch": 0.16218194100427832,
      "grad_norm": 0.22968047857284546,
      "learning_rate": 0.001,
      "loss": 3.0991,
      "step": 2881
    },
    {
      "epoch": 0.1646025669894168,
      "grad_norm": 0.28805041313171387,
      "learning_rate": 0.001,
      "loss": 3.0911,
      "step": 2924
    },
    {
      "epoch": 0.16702319297455528,
      "grad_norm": 0.2599535584449768,
      "learning_rate": 0.001,
      "loss": 3.1062,
      "step": 2967
    },
    {
      "epoch": 0.16944381895969376,
      "grad_norm": 0.24836072325706482,
      "learning_rate": 0.001,
      "loss": 3.0839,
      "step": 3010
    },
    {
      "epoch": 0.17186444494483225,
      "grad_norm": 0.28369227051734924,
      "learning_rate": 0.001,
      "loss": 3.0947,
      "step": 3053
    },
    {
      "epoch": 0.17428507092997073,
      "grad_norm": 0.2852259874343872,
      "learning_rate": 0.001,
      "loss": 3.0757,
      "step": 3096
    },
    {
      "epoch": 0.1767056969151092,
      "grad_norm": 0.31145989894866943,
      "learning_rate": 0.001,
      "loss": 3.0758,
      "step": 3139
    },
    {
      "epoch": 0.1791263229002477,
      "grad_norm": 0.4880489706993103,
      "learning_rate": 0.001,
      "loss": 3.092,
      "step": 3182
    },
    {
      "epoch": 0.18154694888538617,
      "grad_norm": 0.3003547787666321,
      "learning_rate": 0.001,
      "loss": 3.0877,
      "step": 3225
    },
    {
      "epoch": 0.18396757487052465,
      "grad_norm": 0.2658514082431793,
      "learning_rate": 0.001,
      "loss": 3.0756,
      "step": 3268
    },
    {
      "epoch": 0.18638820085566313,
      "grad_norm": 0.2758990526199341,
      "learning_rate": 0.001,
      "loss": 3.0827,
      "step": 3311
    },
    {
      "epoch": 0.1888088268408016,
      "grad_norm": 0.3104788362979889,
      "learning_rate": 0.001,
      "loss": 3.0814,
      "step": 3354
    },
    {
      "epoch": 0.19122945282594012,
      "grad_norm": 0.3327251672744751,
      "learning_rate": 0.001,
      "loss": 3.0744,
      "step": 3397
    },
    {
      "epoch": 0.1936500788110786,
      "grad_norm": 0.3819736838340759,
      "learning_rate": 0.001,
      "loss": 3.0878,
      "step": 3440
    },
    {
      "epoch": 0.19607070479621708,
      "grad_norm": 0.23961155116558075,
      "learning_rate": 0.001,
      "loss": 3.0708,
      "step": 3483
    },
    {
      "epoch": 0.19849133078135556,
      "grad_norm": 0.3279309868812561,
      "learning_rate": 0.001,
      "loss": 3.0613,
      "step": 3526
    },
    {
      "epoch": 0.20091195676649404,
      "grad_norm": 0.21976926922798157,
      "learning_rate": 0.001,
      "loss": 3.066,
      "step": 3569
    },
    {
      "epoch": 0.20333258275163252,
      "grad_norm": 0.3456175923347473,
      "learning_rate": 0.001,
      "loss": 3.0603,
      "step": 3612
    },
    {
      "epoch": 0.205753208736771,
      "grad_norm": 0.3243654668331146,
      "learning_rate": 0.001,
      "loss": 3.0534,
      "step": 3655
    },
    {
      "epoch": 0.20817383472190948,
      "grad_norm": 0.2562883198261261,
      "learning_rate": 0.001,
      "loss": 3.0702,
      "step": 3698
    },
    {
      "epoch": 0.21059446070704796,
      "grad_norm": 0.2658468186855316,
      "learning_rate": 0.001,
      "loss": 3.0642,
      "step": 3741
    },
    {
      "epoch": 0.21301508669218644,
      "grad_norm": 0.2608698010444641,
      "learning_rate": 0.001,
      "loss": 3.0597,
      "step": 3784
    },
    {
      "epoch": 0.21543571267732492,
      "grad_norm": 0.2657344341278076,
      "learning_rate": 0.001,
      "loss": 3.0558,
      "step": 3827
    },
    {
      "epoch": 0.2178563386624634,
      "grad_norm": 0.2623925805091858,
      "learning_rate": 0.001,
      "loss": 3.0533,
      "step": 3870
    },
    {
      "epoch": 0.22027696464760188,
      "grad_norm": 0.30456966161727905,
      "learning_rate": 0.001,
      "loss": 3.0549,
      "step": 3913
    },
    {
      "epoch": 0.22269759063274036,
      "grad_norm": 0.3162463903427124,
      "learning_rate": 0.001,
      "loss": 3.0522,
      "step": 3956
    },
    {
      "epoch": 0.22511821661787887,
      "grad_norm": 0.25987708568573,
      "learning_rate": 0.001,
      "loss": 3.0629,
      "step": 3999
    },
    {
      "epoch": 0.22753884260301735,
      "grad_norm": 0.25829339027404785,
      "learning_rate": 0.001,
      "loss": 3.0583,
      "step": 4042
    },
    {
      "epoch": 0.22995946858815583,
      "grad_norm": 0.30163493752479553,
      "learning_rate": 0.001,
      "loss": 3.0556,
      "step": 4085
    },
    {
      "epoch": 0.2323800945732943,
      "grad_norm": 0.26830950379371643,
      "learning_rate": 0.001,
      "loss": 3.0474,
      "step": 4128
    },
    {
      "epoch": 0.2348007205584328,
      "grad_norm": 0.2521868944168091,
      "learning_rate": 0.001,
      "loss": 3.0497,
      "step": 4171
    },
    {
      "epoch": 0.23722134654357127,
      "grad_norm": 0.2956887185573578,
      "learning_rate": 0.001,
      "loss": 3.0407,
      "step": 4214
    },
    {
      "epoch": 0.23964197252870975,
      "grad_norm": 0.2560979425907135,
      "learning_rate": 0.001,
      "loss": 3.0399,
      "step": 4257
    },
    {
      "epoch": 0.240148615176762,
      "eval_ag_news_bleu_score": 3.5996576527895505,
      "eval_ag_news_bleu_score_sem": 0.13938536439791438,
      "eval_ag_news_emb_cos_sim": 0.9899976849555969,
      "eval_ag_news_emb_cos_sim_sem": 0.00033553823339171527,
      "eval_ag_news_emb_top1_equal": 0.5933333039283752,
      "eval_ag_news_emb_top1_equal_sem": 0.028407504362121784,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 4.137884140014648,
      "eval_ag_news_n_ngrams_match_1": 4.434,
      "eval_ag_news_n_ngrams_match_2": 0.786,
      "eval_ag_news_n_ngrams_match_3": 0.198,
      "eval_ag_news_num_pred_words": 23.69,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.22751108134833642,
      "eval_ag_news_runtime": 16.4529,
      "eval_ag_news_samples_per_second": 30.39,
      "eval_ag_news_steps_per_second": 0.122,
      "eval_ag_news_token_set_f1": 0.20914330465384395,
      "eval_ag_news_token_set_f1_sem": 0.005012451755762966,
      "eval_ag_news_token_set_precision": 0.1970916594795352,
      "eval_ag_news_token_set_recall": 0.23145823801017673,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 4266
    },
    {
      "epoch": 0.240148615176762,
      "eval_anthropic_toxic_prompts_bleu_score": 4.19593184744914,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.17295889609736684,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.99015873670578,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002991306917923213,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6166666746139526,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028117578599086417,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.372663974761963,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.102,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.128,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.408,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.814,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2338780108263081,
      "eval_anthropic_toxic_prompts_runtime": 6.5842,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.94,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.304,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.28111766929984633,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006308424027237078,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3203717547171201,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.27661048900942775,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 4266
    },
    {
      "epoch": 0.240148615176762,
      "eval_arxiv_bleu_score": 3.345837167187051,
      "eval_arxiv_bleu_score_sem": 0.109813930497769,
      "eval_arxiv_emb_cos_sim": 0.9897241592407227,
      "eval_arxiv_emb_cos_sim_sem": 0.00031241441954135083,
      "eval_arxiv_emb_top1_equal": 0.5299999713897705,
      "eval_arxiv_emb_top1_equal_sem": 0.028863650627428784,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 4.1174445152282715,
      "eval_arxiv_n_ngrams_match_1": 4.316,
      "eval_arxiv_n_ngrams_match_2": 0.59,
      "eval_arxiv_n_ngrams_match_3": 0.118,
      "eval_arxiv_num_pred_words": 18.402,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.22326769572720567,
      "eval_arxiv_runtime": 6.9163,
      "eval_arxiv_samples_per_second": 72.293,
      "eval_arxiv_steps_per_second": 0.289,
      "eval_arxiv_token_set_f1": 0.21890849632921155,
      "eval_arxiv_token_set_f1_sem": 0.0046148846829510036,
      "eval_arxiv_token_set_precision": 0.18147517493803317,
      "eval_arxiv_token_set_recall": 0.2981134752330574,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 4266
    },
    {
      "epoch": 0.240148615176762,
      "eval_python_code_alpaca_bleu_score": 4.524312963484297,
      "eval_python_code_alpaca_bleu_score_sem": 0.14727177109622144,
      "eval_python_code_alpaca_emb_cos_sim": 0.9880996942520142,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003525723602614284,
      "eval_python_code_alpaca_emb_top1_equal": 0.6833333373069763,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02690183265038281,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.5380005836486816,
      "eval_python_code_alpaca_n_ngrams_match_1": 4.93,
      "eval_python_code_alpaca_n_ngrams_match_2": 0.974,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.276,
      "eval_python_code_alpaca_num_pred_words": 19.886,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.27148189021175845,
      "eval_python_code_alpaca_runtime": 6.564,
      "eval_python_code_alpaca_samples_per_second": 76.173,
      "eval_python_code_alpaca_steps_per_second": 0.305,
      "eval_python_code_alpaca_token_set_f1": 0.2975157063627567,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005309450311016666,
      "eval_python_code_alpaca_token_set_precision": 0.2861185755668331,
      "eval_python_code_alpaca_token_set_recall": 0.3456960169771454,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 4266
    },
    {
      "epoch": 0.240148615176762,
      "eval_wikibio_bleu_score": 3.5941572484520075,
      "eval_wikibio_bleu_score_sem": 0.12889916552814795,
      "eval_wikibio_emb_cos_sim": 0.9901023507118225,
      "eval_wikibio_emb_cos_sim_sem": 0.00036892691582134833,
      "eval_wikibio_emb_top1_equal": 0.6866666674613953,
      "eval_wikibio_emb_top1_equal_sem": 0.02682505950857856,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.182645320892334,
      "eval_wikibio_n_ngrams_match_1": 3.092,
      "eval_wikibio_n_ngrams_match_2": 0.772,
      "eval_wikibio_n_ngrams_match_3": 0.17,
      "eval_wikibio_num_pred_words": 21.144,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.24133585226414536,
      "eval_wikibio_runtime": 6.5188,
      "eval_wikibio_samples_per_second": 76.702,
      "eval_wikibio_steps_per_second": 0.307,
      "eval_wikibio_token_set_f1": 0.19771841775404264,
      "eval_wikibio_token_set_f1_sem": 0.005119956038024957,
      "eval_wikibio_token_set_precision": 0.23415783493267467,
      "eval_wikibio_token_set_recall": 0.17767662655213223,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 4266
    },
    {
      "epoch": 0.240148615176762,
      "eval_nq_5round_bleu_score": 8.443468178909258,
      "eval_nq_5round_bleu_score_sem": 0.4069344902488757,
      "eval_nq_5round_emb_cos_sim": 0.9914259910583496,
      "eval_nq_5round_emb_cos_sim_sem": 0.00027128036632593534,
      "eval_nq_5round_emb_top1_equal": 0.6066666841506958,
      "eval_nq_5round_emb_top1_equal_sem": 0.028250091805185003,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.298,
      "eval_nq_5round_n_ngrams_match_2": 2.692,
      "eval_nq_5round_n_ngrams_match_3": 1.146,
      "eval_nq_5round_num_pred_words": 23.804,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.32642431341734307,
      "eval_nq_5round_token_set_f1": 0.3716270558486446,
      "eval_nq_5round_token_set_f1_sem": 0.006310634553202306,
      "eval_nq_5round_token_set_precision": 0.3344437278224889,
      "eval_nq_5round_token_set_recall": 0.43849118497534295,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.736978329171647,
      "eval_nq_bleu_score_sem": 0.4067158404250311,
      "eval_nq_emb_cos_sim": 0.9918575286865234,
      "eval_nq_emb_cos_sim_sem": 0.0002671566425540754,
      "eval_nq_emb_top1_equal": 0.6133333444595337,
      "eval_nq_emb_top1_equal_sem": 0.028163137649013424,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.791560173034668,
      "eval_nq_n_ngrams_match_1": 8.676,
      "eval_nq_n_ngrams_match_2": 2.834,
      "eval_nq_n_ngrams_match_3": 1.182,
      "eval_nq_num_pred_words": 23.9,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3419440275557251,
      "eval_nq_runtime": 18.7625,
      "eval_nq_samples_per_second": 26.649,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.38336877884109805,
      "eval_nq_token_set_f1_sem": 0.0060935396700816,
      "eval_nq_token_set_precision": 0.35133803436873184,
      "eval_nq_token_set_recall": 0.43818778624704624,
      "eval_nq_true_num_tokens": 32.0,
      "step": 4266
    },
    {
      "epoch": 0.24206259851384823,
      "grad_norm": 0.29140353202819824,
      "learning_rate": 0.001,
      "loss": 3.0409,
      "step": 4300
    },
    {
      "epoch": 0.2444832244989867,
      "grad_norm": 0.24609676003456116,
      "learning_rate": 0.001,
      "loss": 3.0371,
      "step": 4343
    },
    {
      "epoch": 0.2469038504841252,
      "grad_norm": 0.27688905596733093,
      "learning_rate": 0.001,
      "loss": 3.0356,
      "step": 4386
    },
    {
      "epoch": 0.24932447646926367,
      "grad_norm": 0.26407530903816223,
      "learning_rate": 0.001,
      "loss": 3.0333,
      "step": 4429
    },
    {
      "epoch": 0.2517451024544022,
      "grad_norm": 0.2620108723640442,
      "learning_rate": 0.001,
      "loss": 3.0345,
      "step": 4472
    },
    {
      "epoch": 0.25416572843954066,
      "grad_norm": 0.2631508409976959,
      "learning_rate": 0.001,
      "loss": 3.0432,
      "step": 4515
    },
    {
      "epoch": 0.25658635442467914,
      "grad_norm": 0.2280658632516861,
      "learning_rate": 0.001,
      "loss": 3.0437,
      "step": 4558
    },
    {
      "epoch": 0.2590069804098176,
      "grad_norm": 0.23430952429771423,
      "learning_rate": 0.001,
      "loss": 3.0341,
      "step": 4601
    },
    {
      "epoch": 0.2614276063949561,
      "grad_norm": 0.29745519161224365,
      "learning_rate": 0.001,
      "loss": 3.0211,
      "step": 4644
    },
    {
      "epoch": 0.2638482323800946,
      "grad_norm": 0.24295496940612793,
      "learning_rate": 0.001,
      "loss": 3.0242,
      "step": 4687
    },
    {
      "epoch": 0.26626885836523306,
      "grad_norm": 0.2332877814769745,
      "learning_rate": 0.001,
      "loss": 3.0217,
      "step": 4730
    },
    {
      "epoch": 0.26868948435037154,
      "grad_norm": 0.2688761353492737,
      "learning_rate": 0.001,
      "loss": 3.0229,
      "step": 4773
    },
    {
      "epoch": 0.27111011033551,
      "grad_norm": 0.2577943503856659,
      "learning_rate": 0.001,
      "loss": 3.0223,
      "step": 4816
    },
    {
      "epoch": 0.2735307363206485,
      "grad_norm": 0.3007420599460602,
      "learning_rate": 0.001,
      "loss": 3.03,
      "step": 4859
    },
    {
      "epoch": 0.275951362305787,
      "grad_norm": 0.3002866804599762,
      "learning_rate": 0.001,
      "loss": 3.0169,
      "step": 4902
    },
    {
      "epoch": 0.27837198829092547,
      "grad_norm": 0.2752511501312256,
      "learning_rate": 0.001,
      "loss": 3.0093,
      "step": 4945
    },
    {
      "epoch": 0.28079261427606395,
      "grad_norm": 0.33148136734962463,
      "learning_rate": 0.001,
      "loss": 3.0218,
      "step": 4988
    },
    {
      "epoch": 0.2832132402612024,
      "grad_norm": 0.2148127555847168,
      "learning_rate": 0.001,
      "loss": 3.0166,
      "step": 5031
    },
    {
      "epoch": 0.2856338662463409,
      "grad_norm": 0.2769113779067993,
      "learning_rate": 0.001,
      "loss": 3.022,
      "step": 5074
    },
    {
      "epoch": 0.2880544922314794,
      "grad_norm": 0.29342222213745117,
      "learning_rate": 0.001,
      "loss": 3.0319,
      "step": 5117
    },
    {
      "epoch": 0.29047511821661787,
      "grad_norm": 0.2981497347354889,
      "learning_rate": 0.001,
      "loss": 3.0042,
      "step": 5160
    },
    {
      "epoch": 0.29289574420175635,
      "grad_norm": 0.5130410194396973,
      "learning_rate": 0.001,
      "loss": 3.0189,
      "step": 5203
    },
    {
      "epoch": 0.2953163701868948,
      "grad_norm": 0.23910002410411835,
      "learning_rate": 0.001,
      "loss": 3.0195,
      "step": 5246
    },
    {
      "epoch": 0.2977369961720333,
      "grad_norm": 0.3065224289894104,
      "learning_rate": 0.001,
      "loss": 3.0176,
      "step": 5289
    },
    {
      "epoch": 0.3001576221571718,
      "grad_norm": 0.25541120767593384,
      "learning_rate": 0.001,
      "loss": 3.0089,
      "step": 5332
    },
    {
      "epoch": 0.30257824814231027,
      "grad_norm": 0.2646283209323883,
      "learning_rate": 0.001,
      "loss": 3.0163,
      "step": 5375
    },
    {
      "epoch": 0.30499887412744875,
      "grad_norm": 0.2744535505771637,
      "learning_rate": 0.001,
      "loss": 3.018,
      "step": 5418
    },
    {
      "epoch": 0.30741950011258723,
      "grad_norm": 0.29248756170272827,
      "learning_rate": 0.001,
      "loss": 3.0045,
      "step": 5461
    },
    {
      "epoch": 0.3098401260977257,
      "grad_norm": 0.3028799593448639,
      "learning_rate": 0.001,
      "loss": 3.0031,
      "step": 5504
    },
    {
      "epoch": 0.31226075208286425,
      "grad_norm": 0.23199155926704407,
      "learning_rate": 0.001,
      "loss": 3.0099,
      "step": 5547
    },
    {
      "epoch": 0.3146813780680027,
      "grad_norm": 0.24495764076709747,
      "learning_rate": 0.001,
      "loss": 3.0043,
      "step": 5590
    },
    {
      "epoch": 0.3171020040531412,
      "grad_norm": 0.2618519365787506,
      "learning_rate": 0.001,
      "loss": 3.0068,
      "step": 5633
    },
    {
      "epoch": 0.3195226300382797,
      "grad_norm": 0.42693182826042175,
      "learning_rate": 0.001,
      "loss": 2.9968,
      "step": 5676
    },
    {
      "epoch": 0.32194325602341817,
      "grad_norm": 0.29266175627708435,
      "learning_rate": 0.001,
      "loss": 3.0155,
      "step": 5719
    },
    {
      "epoch": 0.32436388200855665,
      "grad_norm": 0.4010351002216339,
      "learning_rate": 0.001,
      "loss": 2.9886,
      "step": 5762
    },
    {
      "epoch": 0.32678450799369513,
      "grad_norm": 0.2732771337032318,
      "learning_rate": 0.001,
      "loss": 3.0012,
      "step": 5805
    },
    {
      "epoch": 0.3292051339788336,
      "grad_norm": 0.3057151734828949,
      "learning_rate": 0.001,
      "loss": 2.9995,
      "step": 5848
    },
    {
      "epoch": 0.3316257599639721,
      "grad_norm": 0.2591167092323303,
      "learning_rate": 0.001,
      "loss": 2.9937,
      "step": 5891
    },
    {
      "epoch": 0.33404638594911057,
      "grad_norm": 0.2276996374130249,
      "learning_rate": 0.001,
      "loss": 2.9901,
      "step": 5934
    },
    {
      "epoch": 0.33646701193424905,
      "grad_norm": 0.25985318422317505,
      "learning_rate": 0.001,
      "loss": 2.9821,
      "step": 5977
    },
    {
      "epoch": 0.33888763791938753,
      "grad_norm": 0.3097335696220398,
      "learning_rate": 0.001,
      "loss": 2.9845,
      "step": 6020
    },
    {
      "epoch": 0.341308263904526,
      "grad_norm": 0.24783314764499664,
      "learning_rate": 0.001,
      "loss": 2.9911,
      "step": 6063
    },
    {
      "epoch": 0.3437288898896645,
      "grad_norm": 0.3052878975868225,
      "learning_rate": 0.001,
      "loss": 2.9849,
      "step": 6106
    },
    {
      "epoch": 0.34614951587480297,
      "grad_norm": 0.30991458892822266,
      "learning_rate": 0.001,
      "loss": 2.9705,
      "step": 6149
    },
    {
      "epoch": 0.34857014185994145,
      "grad_norm": 0.2614838778972626,
      "learning_rate": 0.001,
      "loss": 2.9767,
      "step": 6192
    },
    {
      "epoch": 0.35099076784507993,
      "grad_norm": 0.2360420674085617,
      "learning_rate": 0.001,
      "loss": 2.9828,
      "step": 6235
    },
    {
      "epoch": 0.3534113938302184,
      "grad_norm": 0.30047130584716797,
      "learning_rate": 0.001,
      "loss": 2.9643,
      "step": 6278
    },
    {
      "epoch": 0.3558320198153569,
      "grad_norm": 0.3277058005332947,
      "learning_rate": 0.001,
      "loss": 2.9635,
      "step": 6321
    },
    {
      "epoch": 0.3582526458004954,
      "grad_norm": 0.24637471139431,
      "learning_rate": 0.001,
      "loss": 2.9573,
      "step": 6364
    },
    {
      "epoch": 0.360222922765143,
      "eval_ag_news_bleu_score": 3.593712475717885,
      "eval_ag_news_bleu_score_sem": 0.13425332598435585,
      "eval_ag_news_emb_cos_sim": 0.9901178479194641,
      "eval_ag_news_emb_cos_sim_sem": 0.00032929081230992233,
      "eval_ag_news_emb_top1_equal": 0.6166666746139526,
      "eval_ag_news_emb_top1_equal_sem": 0.028117578599086417,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 4.109241008758545,
      "eval_ag_news_n_ngrams_match_1": 4.394,
      "eval_ag_news_n_ngrams_match_2": 0.768,
      "eval_ag_news_n_ngrams_match_3": 0.194,
      "eval_ag_news_num_pred_words": 23.54,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2267456634705723,
      "eval_ag_news_runtime": 15.697,
      "eval_ag_news_samples_per_second": 31.853,
      "eval_ag_news_steps_per_second": 0.127,
      "eval_ag_news_token_set_f1": 0.20826018173746766,
      "eval_ag_news_token_set_f1_sem": 0.00491191945517863,
      "eval_ag_news_token_set_precision": 0.197630873074377,
      "eval_ag_news_token_set_recall": 0.23268396959414298,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 6399
    },
    {
      "epoch": 0.360222922765143,
      "eval_anthropic_toxic_prompts_bleu_score": 4.213857683322143,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.17836009370510916,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9900322556495667,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00028965543469260417,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6333333253860474,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02786867456387452,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.3565516471862793,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.09,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.09,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.39,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.43,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2371539478581356,
      "eval_anthropic_toxic_prompts_runtime": 6.6233,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.491,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.302,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2776909085147973,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006204725051745472,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.32196841334772064,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2717212560106322,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 6399
    },
    {
      "epoch": 0.360222922765143,
      "eval_arxiv_bleu_score": 3.2837248302300464,
      "eval_arxiv_bleu_score_sem": 0.10519276410085954,
      "eval_arxiv_emb_cos_sim": 0.9895082712173462,
      "eval_arxiv_emb_cos_sim_sem": 0.00031775458052818483,
      "eval_arxiv_emb_top1_equal": 0.6033333539962769,
      "eval_arxiv_emb_top1_equal_sem": 0.028291497235256893,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 4.078197956085205,
      "eval_arxiv_n_ngrams_match_1": 4.27,
      "eval_arxiv_n_ngrams_match_2": 0.56,
      "eval_arxiv_n_ngrams_match_3": 0.104,
      "eval_arxiv_num_pred_words": 18.198,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2238987175794123,
      "eval_arxiv_runtime": 7.318,
      "eval_arxiv_samples_per_second": 68.324,
      "eval_arxiv_steps_per_second": 0.273,
      "eval_arxiv_token_set_f1": 0.21979648353031128,
      "eval_arxiv_token_set_f1_sem": 0.004445052406105191,
      "eval_arxiv_token_set_precision": 0.17980896152864292,
      "eval_arxiv_token_set_recall": 0.3090831291395975,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 6399
    },
    {
      "epoch": 0.360222922765143,
      "eval_python_code_alpaca_bleu_score": 4.60589806016857,
      "eval_python_code_alpaca_bleu_score_sem": 0.15504305035223456,
      "eval_python_code_alpaca_emb_cos_sim": 0.9879406094551086,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004610746561399092,
      "eval_python_code_alpaca_emb_top1_equal": 0.6133333444595337,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028163139369651306,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.4298906326293945,
      "eval_python_code_alpaca_n_ngrams_match_1": 4.91,
      "eval_python_code_alpaca_n_ngrams_match_2": 0.972,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.294,
      "eval_python_code_alpaca_num_pred_words": 19.97,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2713522393128047,
      "eval_python_code_alpaca_runtime": 6.6005,
      "eval_python_code_alpaca_samples_per_second": 75.751,
      "eval_python_code_alpaca_steps_per_second": 0.303,
      "eval_python_code_alpaca_token_set_f1": 0.29664893779173757,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005128954241052294,
      "eval_python_code_alpaca_token_set_precision": 0.285839169012165,
      "eval_python_code_alpaca_token_set_recall": 0.3428541640606897,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 6399
    },
    {
      "epoch": 0.360222922765143,
      "eval_wikibio_bleu_score": 3.533687413553569,
      "eval_wikibio_bleu_score_sem": 0.12364186410615884,
      "eval_wikibio_emb_cos_sim": 0.9905949234962463,
      "eval_wikibio_emb_cos_sim_sem": 0.00032369123827691223,
      "eval_wikibio_emb_top1_equal": 0.6600000262260437,
      "eval_wikibio_emb_top1_equal_sem": 0.02739528406568103,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.170434474945068,
      "eval_wikibio_n_ngrams_match_1": 3.034,
      "eval_wikibio_n_ngrams_match_2": 0.76,
      "eval_wikibio_n_ngrams_match_3": 0.154,
      "eval_wikibio_num_pred_words": 20.986,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.23853286137662044,
      "eval_wikibio_runtime": 6.521,
      "eval_wikibio_samples_per_second": 76.675,
      "eval_wikibio_steps_per_second": 0.307,
      "eval_wikibio_token_set_f1": 0.1927398354960495,
      "eval_wikibio_token_set_f1_sem": 0.00502019711661643,
      "eval_wikibio_token_set_precision": 0.23053137519530167,
      "eval_wikibio_token_set_recall": 0.17223044093109488,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 6399
    },
    {
      "epoch": 0.360222922765143,
      "eval_nq_5round_bleu_score": 8.424580389165813,
      "eval_nq_5round_bleu_score_sem": 0.41032467332924527,
      "eval_nq_5round_emb_cos_sim": 0.9919509887695312,
      "eval_nq_5round_emb_cos_sim_sem": 0.00025740067951435155,
      "eval_nq_5round_emb_top1_equal": 0.6033333539962769,
      "eval_nq_5round_emb_top1_equal_sem": 0.028291497235256893,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.392,
      "eval_nq_5round_n_ngrams_match_2": 2.68,
      "eval_nq_5round_n_ngrams_match_3": 1.136,
      "eval_nq_5round_num_pred_words": 23.896,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.32730513948511464,
      "eval_nq_5round_token_set_f1": 0.3703306068509083,
      "eval_nq_5round_token_set_f1_sem": 0.0062048369800834936,
      "eval_nq_5round_token_set_precision": 0.3370777854165533,
      "eval_nq_5round_token_set_recall": 0.42944884554991314,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.702254617744972,
      "eval_nq_bleu_score_sem": 0.4059025754893612,
      "eval_nq_emb_cos_sim": 0.9922280311584473,
      "eval_nq_emb_cos_sim_sem": 0.00022448092467839702,
      "eval_nq_emb_top1_equal": 0.6433333158493042,
      "eval_nq_emb_top1_equal_sem": 0.02770216499443815,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.71516489982605,
      "eval_nq_n_ngrams_match_1": 8.7,
      "eval_nq_n_ngrams_match_2": 2.802,
      "eval_nq_n_ngrams_match_3": 1.184,
      "eval_nq_num_pred_words": 23.94,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.34029861591644656,
      "eval_nq_runtime": 19.245,
      "eval_nq_samples_per_second": 25.981,
      "eval_nq_steps_per_second": 0.104,
      "eval_nq_token_set_f1": 0.38109694073077616,
      "eval_nq_token_set_f1_sem": 0.006125275540705417,
      "eval_nq_token_set_precision": 0.3512232734277178,
      "eval_nq_token_set_recall": 0.4313427125532226,
      "eval_nq_true_num_tokens": 32.0,
      "step": 6399
    },
    {
      "epoch": 0.36067327178563385,
      "grad_norm": 0.2967488765716553,
      "learning_rate": 0.001,
      "loss": 2.9605,
      "step": 6407
    },
    {
      "epoch": 0.36309389777077233,
      "grad_norm": 0.2541279196739197,
      "learning_rate": 0.001,
      "loss": 2.9544,
      "step": 6450
    },
    {
      "epoch": 0.3655145237559108,
      "grad_norm": 0.2591663599014282,
      "learning_rate": 0.001,
      "loss": 2.9677,
      "step": 6493
    },
    {
      "epoch": 0.3679351497410493,
      "grad_norm": 0.2696775794029236,
      "learning_rate": 0.001,
      "loss": 2.9589,
      "step": 6536
    },
    {
      "epoch": 0.3703557757261878,
      "grad_norm": 0.28088483214378357,
      "learning_rate": 0.001,
      "loss": 2.95,
      "step": 6579
    },
    {
      "epoch": 0.37277640171132626,
      "grad_norm": 0.2701501250267029,
      "learning_rate": 0.001,
      "loss": 2.9532,
      "step": 6622
    },
    {
      "epoch": 0.37519702769646474,
      "grad_norm": 0.3477233350276947,
      "learning_rate": 0.001,
      "loss": 2.9532,
      "step": 6665
    },
    {
      "epoch": 0.3776176536816032,
      "grad_norm": 0.2868052124977112,
      "learning_rate": 0.001,
      "loss": 2.9448,
      "step": 6708
    },
    {
      "epoch": 0.38003827966674175,
      "grad_norm": 0.3074627220630646,
      "learning_rate": 0.001,
      "loss": 2.9508,
      "step": 6751
    },
    {
      "epoch": 0.38245890565188023,
      "grad_norm": 0.2870349884033203,
      "learning_rate": 0.001,
      "loss": 2.9642,
      "step": 6794
    },
    {
      "epoch": 0.3848795316370187,
      "grad_norm": 0.23239901661872864,
      "learning_rate": 0.001,
      "loss": 2.9648,
      "step": 6837
    },
    {
      "epoch": 0.3873001576221572,
      "grad_norm": 0.27058079838752747,
      "learning_rate": 0.001,
      "loss": 2.9436,
      "step": 6880
    },
    {
      "epoch": 0.3897207836072957,
      "grad_norm": 0.22367165982723236,
      "learning_rate": 0.001,
      "loss": 2.9395,
      "step": 6923
    },
    {
      "epoch": 0.39214140959243415,
      "grad_norm": 0.26665636897087097,
      "learning_rate": 0.001,
      "loss": 2.9489,
      "step": 6966
    },
    {
      "epoch": 0.39456203557757263,
      "grad_norm": 0.2768474519252777,
      "learning_rate": 0.001,
      "loss": 2.9523,
      "step": 7009
    },
    {
      "epoch": 0.3969826615627111,
      "grad_norm": 0.26923081278800964,
      "learning_rate": 0.001,
      "loss": 2.9325,
      "step": 7052
    },
    {
      "epoch": 0.3994032875478496,
      "grad_norm": 0.28105637431144714,
      "learning_rate": 0.001,
      "loss": 2.9339,
      "step": 7095
    },
    {
      "epoch": 0.4018239135329881,
      "grad_norm": 0.29147085547447205,
      "learning_rate": 0.001,
      "loss": 2.9438,
      "step": 7138
    },
    {
      "epoch": 0.40424453951812656,
      "grad_norm": 0.26321926712989807,
      "learning_rate": 0.001,
      "loss": 2.9321,
      "step": 7181
    },
    {
      "epoch": 0.40666516550326504,
      "grad_norm": 0.28511470556259155,
      "learning_rate": 0.001,
      "loss": 2.9234,
      "step": 7224
    },
    {
      "epoch": 0.4090857914884035,
      "grad_norm": 0.2718827724456787,
      "learning_rate": 0.001,
      "loss": 2.9401,
      "step": 7267
    },
    {
      "epoch": 0.411506417473542,
      "grad_norm": 0.2701495289802551,
      "learning_rate": 0.001,
      "loss": 2.9356,
      "step": 7310
    },
    {
      "epoch": 0.4139270434586805,
      "grad_norm": 0.2793698012828827,
      "learning_rate": 0.001,
      "loss": 2.936,
      "step": 7353
    },
    {
      "epoch": 0.41634766944381896,
      "grad_norm": 0.2938210070133209,
      "learning_rate": 0.001,
      "loss": 2.9211,
      "step": 7396
    },
    {
      "epoch": 0.41876829542895744,
      "grad_norm": 0.2396402359008789,
      "learning_rate": 0.001,
      "loss": 2.9224,
      "step": 7439
    },
    {
      "epoch": 0.4211889214140959,
      "grad_norm": 0.2958645820617676,
      "learning_rate": 0.001,
      "loss": 2.9272,
      "step": 7482
    },
    {
      "epoch": 0.4236095473992344,
      "grad_norm": 0.24851560592651367,
      "learning_rate": 0.001,
      "loss": 2.92,
      "step": 7525
    },
    {
      "epoch": 0.4260301733843729,
      "grad_norm": 0.23405689001083374,
      "learning_rate": 0.001,
      "loss": 2.9214,
      "step": 7568
    },
    {
      "epoch": 0.42845079936951136,
      "grad_norm": 0.29919901490211487,
      "learning_rate": 0.001,
      "loss": 2.9105,
      "step": 7611
    },
    {
      "epoch": 0.43087142535464984,
      "grad_norm": 0.276898592710495,
      "learning_rate": 0.001,
      "loss": 2.9045,
      "step": 7654
    },
    {
      "epoch": 0.4332920513397883,
      "grad_norm": 0.2918844223022461,
      "learning_rate": 0.001,
      "loss": 2.919,
      "step": 7697
    },
    {
      "epoch": 0.4357126773249268,
      "grad_norm": 0.27978968620300293,
      "learning_rate": 0.001,
      "loss": 2.9173,
      "step": 7740
    },
    {
      "epoch": 0.4381333033100653,
      "grad_norm": 0.29411765933036804,
      "learning_rate": 0.001,
      "loss": 2.9148,
      "step": 7783
    },
    {
      "epoch": 0.44055392929520376,
      "grad_norm": 0.24493253231048584,
      "learning_rate": 0.001,
      "loss": 2.9061,
      "step": 7826
    },
    {
      "epoch": 0.44297455528034224,
      "grad_norm": 0.3043416440486908,
      "learning_rate": 0.001,
      "loss": 2.9127,
      "step": 7869
    },
    {
      "epoch": 0.4453951812654807,
      "grad_norm": 0.2920740246772766,
      "learning_rate": 0.001,
      "loss": 2.9027,
      "step": 7912
    },
    {
      "epoch": 0.4478158072506192,
      "grad_norm": 0.2851109802722931,
      "learning_rate": 0.001,
      "loss": 2.9098,
      "step": 7955
    },
    {
      "epoch": 0.45023643323575774,
      "grad_norm": 0.2831255793571472,
      "learning_rate": 0.001,
      "loss": 2.915,
      "step": 7998
    },
    {
      "epoch": 0.4526570592208962,
      "grad_norm": 0.28081315755844116,
      "learning_rate": 0.001,
      "loss": 2.9018,
      "step": 8041
    },
    {
      "epoch": 0.4550776852060347,
      "grad_norm": 0.2737744152545929,
      "learning_rate": 0.001,
      "loss": 2.8812,
      "step": 8084
    },
    {
      "epoch": 0.4574983111911732,
      "grad_norm": 0.2822270691394806,
      "learning_rate": 0.001,
      "loss": 2.8991,
      "step": 8127
    },
    {
      "epoch": 0.45991893717631166,
      "grad_norm": 0.3260841369628906,
      "learning_rate": 0.001,
      "loss": 2.9117,
      "step": 8170
    },
    {
      "epoch": 0.46233956316145014,
      "grad_norm": 0.30120620131492615,
      "learning_rate": 0.001,
      "loss": 2.8887,
      "step": 8213
    },
    {
      "epoch": 0.4647601891465886,
      "grad_norm": 0.35214346647262573,
      "learning_rate": 0.001,
      "loss": 2.8946,
      "step": 8256
    },
    {
      "epoch": 0.4671808151317271,
      "grad_norm": 0.3437255918979645,
      "learning_rate": 0.001,
      "loss": 2.8941,
      "step": 8299
    },
    {
      "epoch": 0.4696014411168656,
      "grad_norm": 0.2606765031814575,
      "learning_rate": 0.001,
      "loss": 2.8965,
      "step": 8342
    },
    {
      "epoch": 0.47202206710200406,
      "grad_norm": 0.26756149530410767,
      "learning_rate": 0.001,
      "loss": 2.8995,
      "step": 8385
    },
    {
      "epoch": 0.47444269308714254,
      "grad_norm": 0.2606211304664612,
      "learning_rate": 0.001,
      "loss": 2.8907,
      "step": 8428
    },
    {
      "epoch": 0.476863319072281,
      "grad_norm": 0.2529884874820709,
      "learning_rate": 0.001,
      "loss": 2.8626,
      "step": 8471
    },
    {
      "epoch": 0.4792839450574195,
      "grad_norm": 0.29952558875083923,
      "learning_rate": 0.001,
      "loss": 2.8858,
      "step": 8514
    },
    {
      "epoch": 0.480297230353524,
      "eval_ag_news_bleu_score": 3.5769065949885035,
      "eval_ag_news_bleu_score_sem": 0.13882763476559576,
      "eval_ag_news_emb_cos_sim": 0.9904878735542297,
      "eval_ag_news_emb_cos_sim_sem": 0.00028336929925839194,
      "eval_ag_news_emb_top1_equal": 0.596666693687439,
      "eval_ag_news_emb_top1_equal_sem": 0.028370197491492683,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 4.034976005554199,
      "eval_ag_news_n_ngrams_match_1": 4.42,
      "eval_ag_news_n_ngrams_match_2": 0.712,
      "eval_ag_news_n_ngrams_match_3": 0.19,
      "eval_ag_news_num_pred_words": 23.18,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.22976139898991654,
      "eval_ag_news_runtime": 15.8043,
      "eval_ag_news_samples_per_second": 31.637,
      "eval_ag_news_steps_per_second": 0.127,
      "eval_ag_news_token_set_f1": 0.20785220563685167,
      "eval_ag_news_token_set_f1_sem": 0.004882230473548854,
      "eval_ag_news_token_set_precision": 0.19826495978935096,
      "eval_ag_news_token_set_recall": 0.22707544315707345,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 8532
    },
    {
      "epoch": 0.480297230353524,
      "eval_anthropic_toxic_prompts_bleu_score": 4.303968388665982,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.1784214078698339,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9904258251190186,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002626157447499308,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6233333349227905,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028022260422154388,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.2738823890686035,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.156,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.116,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.41,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.284,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2401335123598238,
      "eval_anthropic_toxic_prompts_runtime": 6.5867,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.911,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.304,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2770034633751842,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0062028462313865156,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.32251063261601587,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2685599366684278,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 8532
    },
    {
      "epoch": 0.480297230353524,
      "eval_arxiv_bleu_score": 3.309602014008217,
      "eval_arxiv_bleu_score_sem": 0.10740523090384958,
      "eval_arxiv_emb_cos_sim": 0.9898874163627625,
      "eval_arxiv_emb_cos_sim_sem": 0.00032882212668094643,
      "eval_arxiv_emb_top1_equal": 0.5633333325386047,
      "eval_arxiv_emb_top1_equal_sem": 0.028682840836527678,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 4.04457950592041,
      "eval_arxiv_n_ngrams_match_1": 4.27,
      "eval_arxiv_n_ngrams_match_2": 0.578,
      "eval_arxiv_n_ngrams_match_3": 0.1,
      "eval_arxiv_num_pred_words": 17.656,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.22485763337782574,
      "eval_arxiv_runtime": 6.4861,
      "eval_arxiv_samples_per_second": 77.088,
      "eval_arxiv_steps_per_second": 0.308,
      "eval_arxiv_token_set_f1": 0.21670640994498108,
      "eval_arxiv_token_set_f1_sem": 0.004610691735314551,
      "eval_arxiv_token_set_precision": 0.17895736004113486,
      "eval_arxiv_token_set_recall": 0.2985524395693749,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 8532
    },
    {
      "epoch": 0.480297230353524,
      "eval_python_code_alpaca_bleu_score": 4.604418924750929,
      "eval_python_code_alpaca_bleu_score_sem": 0.15557957891156105,
      "eval_python_code_alpaca_emb_cos_sim": 0.9882889986038208,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004531169210001268,
      "eval_python_code_alpaca_emb_top1_equal": 0.6666666865348816,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027262027544015993,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.362619161605835,
      "eval_python_code_alpaca_n_ngrams_match_1": 4.896,
      "eval_python_code_alpaca_n_ngrams_match_2": 0.922,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.278,
      "eval_python_code_alpaca_num_pred_words": 19.47,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2739751598835251,
      "eval_python_code_alpaca_runtime": 7.2861,
      "eval_python_code_alpaca_samples_per_second": 68.624,
      "eval_python_code_alpaca_steps_per_second": 0.274,
      "eval_python_code_alpaca_token_set_f1": 0.29449538576650997,
      "eval_python_code_alpaca_token_set_f1_sem": 0.00526252365064125,
      "eval_python_code_alpaca_token_set_precision": 0.28509524356012717,
      "eval_python_code_alpaca_token_set_recall": 0.34046432172283264,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 8532
    },
    {
      "epoch": 0.480297230353524,
      "eval_wikibio_bleu_score": 3.5387744096321905,
      "eval_wikibio_bleu_score_sem": 0.12705134352479505,
      "eval_wikibio_emb_cos_sim": 0.9905848503112793,
      "eval_wikibio_emb_cos_sim_sem": 0.0002794891801721369,
      "eval_wikibio_emb_top1_equal": 0.6200000047683716,
      "eval_wikibio_emb_top1_equal_sem": 0.028070622391196547,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.148077964782715,
      "eval_wikibio_n_ngrams_match_1": 3.016,
      "eval_wikibio_n_ngrams_match_2": 0.726,
      "eval_wikibio_n_ngrams_match_3": 0.15,
      "eval_wikibio_num_pred_words": 20.688,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.23611486332727422,
      "eval_wikibio_runtime": 7.1613,
      "eval_wikibio_samples_per_second": 69.819,
      "eval_wikibio_steps_per_second": 0.279,
      "eval_wikibio_token_set_f1": 0.19405189369966466,
      "eval_wikibio_token_set_f1_sem": 0.005085219421917528,
      "eval_wikibio_token_set_precision": 0.22924282766268303,
      "eval_wikibio_token_set_recall": 0.17502098339399635,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 8532
    },
    {
      "epoch": 0.480297230353524,
      "eval_nq_5round_bleu_score": 8.534068529739828,
      "eval_nq_5round_bleu_score_sem": 0.40904790709459443,
      "eval_nq_5round_emb_cos_sim": 0.991858720779419,
      "eval_nq_5round_emb_cos_sim_sem": 0.000260727156476044,
      "eval_nq_5round_emb_top1_equal": 0.6233333349227905,
      "eval_nq_5round_emb_top1_equal_sem": 0.028022260422154388,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.296,
      "eval_nq_5round_n_ngrams_match_2": 2.698,
      "eval_nq_5round_n_ngrams_match_3": 1.144,
      "eval_nq_5round_num_pred_words": 23.386,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.32586807967670234,
      "eval_nq_5round_token_set_f1": 0.3684598695114288,
      "eval_nq_5round_token_set_f1_sem": 0.0062414349658299825,
      "eval_nq_5round_token_set_precision": 0.33704006309357654,
      "eval_nq_5round_token_set_recall": 0.42209659034496394,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.666528574026724,
      "eval_nq_bleu_score_sem": 0.40475303082797176,
      "eval_nq_emb_cos_sim": 0.9919183850288391,
      "eval_nq_emb_cos_sim_sem": 0.0002626448073992122,
      "eval_nq_emb_top1_equal": 0.6433333158493042,
      "eval_nq_emb_top1_equal_sem": 0.02770216499443815,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.6212146282196045,
      "eval_nq_n_ngrams_match_1": 8.646,
      "eval_nq_n_ngrams_match_2": 2.76,
      "eval_nq_n_ngrams_match_3": 1.158,
      "eval_nq_num_pred_words": 23.736,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.34145642038200014,
      "eval_nq_runtime": 19.1922,
      "eval_nq_samples_per_second": 26.052,
      "eval_nq_steps_per_second": 0.104,
      "eval_nq_token_set_f1": 0.37869187204837307,
      "eval_nq_token_set_f1_sem": 0.005984381054878401,
      "eval_nq_token_set_precision": 0.35029116038800573,
      "eval_nq_token_set_recall": 0.4241027671583591,
      "eval_nq_true_num_tokens": 32.0,
      "step": 8532
    },
    {
      "epoch": 0.481704571042558,
      "grad_norm": 0.30499598383903503,
      "learning_rate": 0.001,
      "loss": 2.8853,
      "step": 8557
    },
    {
      "epoch": 0.48412519702769646,
      "grad_norm": 0.25930219888687134,
      "learning_rate": 0.001,
      "loss": 2.8775,
      "step": 8600
    },
    {
      "epoch": 0.48654582301283494,
      "grad_norm": 0.24619758129119873,
      "learning_rate": 0.001,
      "loss": 2.8819,
      "step": 8643
    },
    {
      "epoch": 0.4889664489979734,
      "grad_norm": 0.2929568290710449,
      "learning_rate": 0.001,
      "loss": 2.8769,
      "step": 8686
    },
    {
      "epoch": 0.4913870749831119,
      "grad_norm": 0.2775774598121643,
      "learning_rate": 0.001,
      "loss": 2.8904,
      "step": 8729
    },
    {
      "epoch": 0.4938077009682504,
      "grad_norm": 0.27790728211402893,
      "learning_rate": 0.001,
      "loss": 2.8817,
      "step": 8772
    },
    {
      "epoch": 0.49622832695338887,
      "grad_norm": 0.3122977614402771,
      "learning_rate": 0.001,
      "loss": 2.8789,
      "step": 8815
    },
    {
      "epoch": 0.49864895293852735,
      "grad_norm": 0.34443268179893494,
      "learning_rate": 0.001,
      "loss": 2.8839,
      "step": 8858
    },
    {
      "epoch": 0.5010695789236659,
      "grad_norm": 0.33139607310295105,
      "learning_rate": 0.001,
      "loss": 2.8795,
      "step": 8901
    },
    {
      "epoch": 0.5034902049088044,
      "grad_norm": 0.26959049701690674,
      "learning_rate": 0.001,
      "loss": 2.8753,
      "step": 8944
    },
    {
      "epoch": 0.5059108308939428,
      "grad_norm": 0.329349547624588,
      "learning_rate": 0.001,
      "loss": 2.8755,
      "step": 8987
    },
    {
      "epoch": 0.5083314568790813,
      "grad_norm": 0.27014046907424927,
      "learning_rate": 0.001,
      "loss": 2.865,
      "step": 9030
    },
    {
      "epoch": 0.5107520828642198,
      "grad_norm": 0.29060497879981995,
      "learning_rate": 0.001,
      "loss": 2.8646,
      "step": 9073
    },
    {
      "epoch": 0.5131727088493583,
      "grad_norm": 0.29821956157684326,
      "learning_rate": 0.001,
      "loss": 2.8657,
      "step": 9116
    },
    {
      "epoch": 0.5155933348344968,
      "grad_norm": 0.3179447054862976,
      "learning_rate": 0.001,
      "loss": 2.8716,
      "step": 9159
    },
    {
      "epoch": 0.5180139608196352,
      "grad_norm": 0.3506690561771393,
      "learning_rate": 0.001,
      "loss": 2.8631,
      "step": 9202
    },
    {
      "epoch": 0.5204345868047737,
      "grad_norm": 0.3335030972957611,
      "learning_rate": 0.001,
      "loss": 2.8726,
      "step": 9245
    },
    {
      "epoch": 0.5228552127899122,
      "grad_norm": 0.2654228210449219,
      "learning_rate": 0.001,
      "loss": 2.8601,
      "step": 9288
    },
    {
      "epoch": 0.5252758387750507,
      "grad_norm": 0.3020453453063965,
      "learning_rate": 0.001,
      "loss": 2.8755,
      "step": 9331
    },
    {
      "epoch": 0.5276964647601892,
      "grad_norm": 0.286029577255249,
      "learning_rate": 0.001,
      "loss": 2.8707,
      "step": 9374
    },
    {
      "epoch": 0.5301170907453276,
      "grad_norm": 0.3444445729255676,
      "learning_rate": 0.001,
      "loss": 2.8667,
      "step": 9417
    },
    {
      "epoch": 0.5325377167304661,
      "grad_norm": 0.2904354929924011,
      "learning_rate": 0.001,
      "loss": 2.8595,
      "step": 9460
    },
    {
      "epoch": 0.5349583427156046,
      "grad_norm": 0.27774274349212646,
      "learning_rate": 0.001,
      "loss": 2.8562,
      "step": 9503
    },
    {
      "epoch": 0.5373789687007431,
      "grad_norm": 0.29608452320098877,
      "learning_rate": 0.001,
      "loss": 2.851,
      "step": 9546
    },
    {
      "epoch": 0.5397995946858816,
      "grad_norm": 0.3411881923675537,
      "learning_rate": 0.001,
      "loss": 2.8735,
      "step": 9589
    },
    {
      "epoch": 0.54222022067102,
      "grad_norm": 0.3129940629005432,
      "learning_rate": 0.001,
      "loss": 2.8555,
      "step": 9632
    },
    {
      "epoch": 0.5446408466561585,
      "grad_norm": 0.27610525488853455,
      "learning_rate": 0.001,
      "loss": 2.8433,
      "step": 9675
    },
    {
      "epoch": 0.547061472641297,
      "grad_norm": 0.3172253966331482,
      "learning_rate": 0.001,
      "loss": 2.8651,
      "step": 9718
    },
    {
      "epoch": 0.5494820986264355,
      "grad_norm": 0.357971727848053,
      "learning_rate": 0.001,
      "loss": 2.8498,
      "step": 9761
    },
    {
      "epoch": 0.551902724611574,
      "grad_norm": 0.3187626898288727,
      "learning_rate": 0.001,
      "loss": 2.8533,
      "step": 9804
    },
    {
      "epoch": 0.5543233505967124,
      "grad_norm": 0.37656545639038086,
      "learning_rate": 0.001,
      "loss": 2.8549,
      "step": 9847
    },
    {
      "epoch": 0.5567439765818509,
      "grad_norm": 0.3399082124233246,
      "learning_rate": 0.001,
      "loss": 2.8423,
      "step": 9890
    },
    {
      "epoch": 0.5591646025669894,
      "grad_norm": 0.39030832052230835,
      "learning_rate": 0.001,
      "loss": 2.8498,
      "step": 9933
    },
    {
      "epoch": 0.5615852285521279,
      "grad_norm": 0.792371392250061,
      "learning_rate": 0.001,
      "loss": 2.8476,
      "step": 9976
    },
    {
      "epoch": 0.5640058545372664,
      "grad_norm": 0.31796830892562866,
      "learning_rate": 0.001,
      "loss": 2.8467,
      "step": 10019
    },
    {
      "epoch": 0.5664264805224049,
      "grad_norm": 0.27938956022262573,
      "learning_rate": 0.001,
      "loss": 2.8481,
      "step": 10062
    },
    {
      "epoch": 0.5688471065075433,
      "grad_norm": 0.26249992847442627,
      "learning_rate": 0.001,
      "loss": 2.8447,
      "step": 10105
    },
    {
      "epoch": 0.5712677324926818,
      "grad_norm": 0.31493791937828064,
      "learning_rate": 0.001,
      "loss": 2.8361,
      "step": 10148
    },
    {
      "epoch": 0.5736883584778203,
      "grad_norm": 0.37958547472953796,
      "learning_rate": 0.001,
      "loss": 2.8241,
      "step": 10191
    },
    {
      "epoch": 0.5761089844629588,
      "grad_norm": 0.2744309902191162,
      "learning_rate": 0.001,
      "loss": 2.8301,
      "step": 10234
    },
    {
      "epoch": 0.5785296104480973,
      "grad_norm": 0.3143821060657501,
      "learning_rate": 0.001,
      "loss": 2.8373,
      "step": 10277
    },
    {
      "epoch": 0.5809502364332357,
      "grad_norm": 0.2859244644641876,
      "learning_rate": 0.001,
      "loss": 2.8242,
      "step": 10320
    },
    {
      "epoch": 0.5833708624183742,
      "grad_norm": 0.3199649155139923,
      "learning_rate": 0.001,
      "loss": 2.834,
      "step": 10363
    },
    {
      "epoch": 0.5857914884035127,
      "grad_norm": 0.2949970066547394,
      "learning_rate": 0.001,
      "loss": 2.8322,
      "step": 10406
    },
    {
      "epoch": 0.5882121143886512,
      "grad_norm": 0.33437350392341614,
      "learning_rate": 0.001,
      "loss": 2.831,
      "step": 10449
    },
    {
      "epoch": 0.5906327403737897,
      "grad_norm": 0.28598418831825256,
      "learning_rate": 0.001,
      "loss": 2.8268,
      "step": 10492
    },
    {
      "epoch": 0.5930533663589281,
      "grad_norm": 0.244672030210495,
      "learning_rate": 0.001,
      "loss": 2.8242,
      "step": 10535
    },
    {
      "epoch": 0.5954739923440666,
      "grad_norm": 0.28037846088409424,
      "learning_rate": 0.001,
      "loss": 2.8195,
      "step": 10578
    },
    {
      "epoch": 0.5978946183292051,
      "grad_norm": 0.6620934009552002,
      "learning_rate": 0.001,
      "loss": 2.8274,
      "step": 10621
    },
    {
      "epoch": 0.6003152443143436,
      "grad_norm": 0.29259413480758667,
      "learning_rate": 0.001,
      "loss": 2.8331,
      "step": 10664
    },
    {
      "epoch": 0.600371537941905,
      "eval_ag_news_bleu_score": 3.5397659293076176,
      "eval_ag_news_bleu_score_sem": 0.13709091953657368,
      "eval_ag_news_emb_cos_sim": 0.9902831315994263,
      "eval_ag_news_emb_cos_sim_sem": 0.0003212146757756577,
      "eval_ag_news_emb_top1_equal": 0.5666666626930237,
      "eval_ag_news_emb_top1_equal_sem": 0.02865756638663049,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.9731087684631348,
      "eval_ag_news_n_ngrams_match_1": 4.304,
      "eval_ag_news_n_ngrams_match_2": 0.718,
      "eval_ag_news_n_ngrams_match_3": 0.184,
      "eval_ag_news_num_pred_words": 22.878,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.22758939713643217,
      "eval_ag_news_runtime": 36.7004,
      "eval_ag_news_samples_per_second": 13.624,
      "eval_ag_news_steps_per_second": 0.054,
      "eval_ag_news_token_set_f1": 0.2026746164389096,
      "eval_ag_news_token_set_f1_sem": 0.004964256918011628,
      "eval_ag_news_token_set_precision": 0.19397840503143404,
      "eval_ag_news_token_set_recall": 0.22176614032697087,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 10665
    },
    {
      "epoch": 0.600371537941905,
      "eval_anthropic_toxic_prompts_bleu_score": 4.322315948377128,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.18582892350996044,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9901501536369324,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00029724688904255304,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.5833333134651184,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02851131216637989,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.2043468952178955,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.026,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.116,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.418,
      "eval_anthropic_toxic_prompts_num_pred_words": 23.93,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.23693083671508267,
      "eval_anthropic_toxic_prompts_runtime": 7.2055,
      "eval_anthropic_toxic_prompts_samples_per_second": 69.392,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.278,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2735100800779503,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006409000586714652,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.318562720246009,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2653497774481396,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 10665
    },
    {
      "epoch": 0.600371537941905,
      "eval_arxiv_bleu_score": 3.221930914205215,
      "eval_arxiv_bleu_score_sem": 0.10210189964092369,
      "eval_arxiv_emb_cos_sim": 0.9896777868270874,
      "eval_arxiv_emb_cos_sim_sem": 0.000326864470926945,
      "eval_arxiv_emb_top1_equal": 0.5766666531562805,
      "eval_arxiv_emb_top1_equal_sem": 0.028573804013736142,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.9882383346557617,
      "eval_arxiv_n_ngrams_match_1": 4.174,
      "eval_arxiv_n_ngrams_match_2": 0.566,
      "eval_arxiv_n_ngrams_match_3": 0.092,
      "eval_arxiv_num_pred_words": 17.204,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.22336538727246408,
      "eval_arxiv_runtime": 7.1168,
      "eval_arxiv_samples_per_second": 70.257,
      "eval_arxiv_steps_per_second": 0.281,
      "eval_arxiv_token_set_f1": 0.21254153552869048,
      "eval_arxiv_token_set_f1_sem": 0.004604413789592912,
      "eval_arxiv_token_set_precision": 0.17469027209774451,
      "eval_arxiv_token_set_recall": 0.29864769861373597,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 10665
    },
    {
      "epoch": 0.600371537941905,
      "eval_python_code_alpaca_bleu_score": 4.814017933086643,
      "eval_python_code_alpaca_bleu_score_sem": 0.17343270627759008,
      "eval_python_code_alpaca_emb_cos_sim": 0.9878620505332947,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00043902177806334297,
      "eval_python_code_alpaca_emb_top1_equal": 0.6100000143051147,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02820730814416699,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.3150548934936523,
      "eval_python_code_alpaca_n_ngrams_match_1": 4.752,
      "eval_python_code_alpaca_n_ngrams_match_2": 0.94,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.286,
      "eval_python_code_alpaca_num_pred_words": 18.064,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2770785278470168,
      "eval_python_code_alpaca_runtime": 8.5003,
      "eval_python_code_alpaca_samples_per_second": 58.821,
      "eval_python_code_alpaca_steps_per_second": 0.235,
      "eval_python_code_alpaca_token_set_f1": 0.29126351903310005,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005396295262657392,
      "eval_python_code_alpaca_token_set_precision": 0.2760389275071111,
      "eval_python_code_alpaca_token_set_recall": 0.34329128123338687,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 10665
    },
    {
      "epoch": 0.600371537941905,
      "eval_wikibio_bleu_score": 3.506141883012896,
      "eval_wikibio_bleu_score_sem": 0.12013193644909957,
      "eval_wikibio_emb_cos_sim": 0.9905354976654053,
      "eval_wikibio_emb_cos_sim_sem": 0.00032098736338002925,
      "eval_wikibio_emb_top1_equal": 0.6600000262260437,
      "eval_wikibio_emb_top1_equal_sem": 0.027395285786318915,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.089732646942139,
      "eval_wikibio_n_ngrams_match_1": 2.964,
      "eval_wikibio_n_ngrams_match_2": 0.718,
      "eval_wikibio_n_ngrams_match_3": 0.136,
      "eval_wikibio_num_pred_words": 20.556,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2350685759453553,
      "eval_wikibio_runtime": 6.8179,
      "eval_wikibio_samples_per_second": 73.336,
      "eval_wikibio_steps_per_second": 0.293,
      "eval_wikibio_token_set_f1": 0.19211254199071162,
      "eval_wikibio_token_set_f1_sem": 0.004996769377418635,
      "eval_wikibio_token_set_precision": 0.2258496102385523,
      "eval_wikibio_token_set_recall": 0.17453582172243676,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 10665
    },
    {
      "epoch": 0.600371537941905,
      "eval_nq_5round_bleu_score": 8.520213668191467,
      "eval_nq_5round_bleu_score_sem": 0.4007298182509701,
      "eval_nq_5round_emb_cos_sim": 0.9918636083602905,
      "eval_nq_5round_emb_cos_sim_sem": 0.00032253986268197614,
      "eval_nq_5round_emb_top1_equal": 0.6133333444595337,
      "eval_nq_5round_emb_top1_equal_sem": 0.028163137649013424,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.292,
      "eval_nq_5round_n_ngrams_match_2": 2.688,
      "eval_nq_5round_n_ngrams_match_3": 1.134,
      "eval_nq_5round_num_pred_words": 23.306,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.32870707327950477,
      "eval_nq_5round_token_set_f1": 0.3658401861343948,
      "eval_nq_5round_token_set_f1_sem": 0.006359649915128281,
      "eval_nq_5round_token_set_precision": 0.3352722178602435,
      "eval_nq_5round_token_set_recall": 0.4181330710976257,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.782020410045078,
      "eval_nq_bleu_score_sem": 0.4002638423868162,
      "eval_nq_emb_cos_sim": 0.9925541281700134,
      "eval_nq_emb_cos_sim_sem": 0.00021886948093247503,
      "eval_nq_emb_top1_equal": 0.6033333539962769,
      "eval_nq_emb_top1_equal_sem": 0.028291498955894778,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.5697455406188965,
      "eval_nq_n_ngrams_match_1": 8.688,
      "eval_nq_n_ngrams_match_2": 2.842,
      "eval_nq_n_ngrams_match_3": 1.174,
      "eval_nq_num_pred_words": 23.466,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.34546607530496054,
      "eval_nq_runtime": 19.8753,
      "eval_nq_samples_per_second": 25.157,
      "eval_nq_steps_per_second": 0.101,
      "eval_nq_token_set_f1": 0.3811523165958929,
      "eval_nq_token_set_f1_sem": 0.006145189074784452,
      "eval_nq_token_set_precision": 0.3520031738822799,
      "eval_nq_token_set_recall": 0.4290789027538924,
      "eval_nq_true_num_tokens": 32.0,
      "step": 10665
    },
    {
      "epoch": 0.6027358702994821,
      "grad_norm": 0.284180223941803,
      "learning_rate": 0.001,
      "loss": 2.8253,
      "step": 10707
    },
    {
      "epoch": 0.6051564962846205,
      "grad_norm": 0.283913254737854,
      "learning_rate": 0.001,
      "loss": 2.8176,
      "step": 10750
    },
    {
      "epoch": 0.607577122269759,
      "grad_norm": 0.3278023302555084,
      "learning_rate": 0.001,
      "loss": 2.8192,
      "step": 10793
    },
    {
      "epoch": 0.6099977482548975,
      "grad_norm": 0.39008229970932007,
      "learning_rate": 0.001,
      "loss": 2.8165,
      "step": 10836
    },
    {
      "epoch": 0.612418374240036,
      "grad_norm": 0.3141207993030548,
      "learning_rate": 0.001,
      "loss": 2.8298,
      "step": 10879
    },
    {
      "epoch": 0.6148390002251745,
      "grad_norm": 0.29817718267440796,
      "learning_rate": 0.001,
      "loss": 2.8205,
      "step": 10922
    },
    {
      "epoch": 0.6172596262103129,
      "grad_norm": 0.29102644324302673,
      "learning_rate": 0.001,
      "loss": 2.8177,
      "step": 10965
    },
    {
      "epoch": 0.6196802521954514,
      "grad_norm": 0.26628413796424866,
      "learning_rate": 0.001,
      "loss": 2.8229,
      "step": 11008
    },
    {
      "epoch": 0.62210087818059,
      "grad_norm": 0.28101006150245667,
      "learning_rate": 0.001,
      "loss": 2.8186,
      "step": 11051
    },
    {
      "epoch": 0.6245215041657285,
      "grad_norm": 0.32386457920074463,
      "learning_rate": 0.001,
      "loss": 2.8103,
      "step": 11094
    },
    {
      "epoch": 0.626942130150867,
      "grad_norm": 0.27414003014564514,
      "learning_rate": 0.001,
      "loss": 2.8234,
      "step": 11137
    },
    {
      "epoch": 0.6293627561360055,
      "grad_norm": 0.3164566159248352,
      "learning_rate": 0.001,
      "loss": 2.8134,
      "step": 11180
    },
    {
      "epoch": 0.6317833821211439,
      "grad_norm": 0.29690995812416077,
      "learning_rate": 0.001,
      "loss": 2.8116,
      "step": 11223
    },
    {
      "epoch": 0.6342040081062824,
      "grad_norm": 0.380837619304657,
      "learning_rate": 0.001,
      "loss": 2.808,
      "step": 11266
    },
    {
      "epoch": 0.6366246340914209,
      "grad_norm": 0.29184088110923767,
      "learning_rate": 0.001,
      "loss": 2.822,
      "step": 11309
    },
    {
      "epoch": 0.6390452600765594,
      "grad_norm": 0.30186617374420166,
      "learning_rate": 0.001,
      "loss": 2.8065,
      "step": 11352
    },
    {
      "epoch": 0.6414658860616979,
      "grad_norm": 0.22835130989551544,
      "learning_rate": 0.001,
      "loss": 2.8069,
      "step": 11395
    },
    {
      "epoch": 0.6438865120468363,
      "grad_norm": 0.2772247791290283,
      "learning_rate": 0.001,
      "loss": 2.8151,
      "step": 11438
    },
    {
      "epoch": 0.6463071380319748,
      "grad_norm": 0.29030343890190125,
      "learning_rate": 0.001,
      "loss": 2.7984,
      "step": 11481
    },
    {
      "epoch": 0.6487277640171133,
      "grad_norm": 0.32622942328453064,
      "learning_rate": 0.001,
      "loss": 2.8141,
      "step": 11524
    },
    {
      "epoch": 0.6511483900022518,
      "grad_norm": 0.3310418426990509,
      "learning_rate": 0.001,
      "loss": 2.8075,
      "step": 11567
    },
    {
      "epoch": 0.6535690159873903,
      "grad_norm": 0.2643769383430481,
      "learning_rate": 0.001,
      "loss": 2.8112,
      "step": 11610
    },
    {
      "epoch": 0.6559896419725287,
      "grad_norm": 0.3302895426750183,
      "learning_rate": 0.001,
      "loss": 2.8182,
      "step": 11653
    },
    {
      "epoch": 0.6584102679576672,
      "grad_norm": 0.3150666654109955,
      "learning_rate": 0.001,
      "loss": 2.8242,
      "step": 11696
    },
    {
      "epoch": 0.6608308939428057,
      "grad_norm": 0.2968304455280304,
      "learning_rate": 0.001,
      "loss": 2.8022,
      "step": 11739
    },
    {
      "epoch": 0.6632515199279442,
      "grad_norm": 0.2833590507507324,
      "learning_rate": 0.001,
      "loss": 2.8108,
      "step": 11782
    },
    {
      "epoch": 0.6656721459130827,
      "grad_norm": 0.3103300631046295,
      "learning_rate": 0.001,
      "loss": 2.81,
      "step": 11825
    },
    {
      "epoch": 0.6680927718982211,
      "grad_norm": 0.3549913465976715,
      "learning_rate": 0.001,
      "loss": 2.8046,
      "step": 11868
    },
    {
      "epoch": 0.6705133978833596,
      "grad_norm": 0.32602420449256897,
      "learning_rate": 0.001,
      "loss": 2.7868,
      "step": 11911
    },
    {
      "epoch": 0.6729340238684981,
      "grad_norm": 0.27826112508773804,
      "learning_rate": 0.001,
      "loss": 2.7985,
      "step": 11954
    },
    {
      "epoch": 0.6753546498536366,
      "grad_norm": 0.31484636664390564,
      "learning_rate": 0.001,
      "loss": 2.7955,
      "step": 11997
    },
    {
      "epoch": 0.6777752758387751,
      "grad_norm": 0.34207823872566223,
      "learning_rate": 0.001,
      "loss": 2.7918,
      "step": 12040
    },
    {
      "epoch": 0.6801959018239135,
      "grad_norm": 0.35010504722595215,
      "learning_rate": 0.001,
      "loss": 2.7926,
      "step": 12083
    },
    {
      "epoch": 0.682616527809052,
      "grad_norm": 0.3252001702785492,
      "learning_rate": 0.001,
      "loss": 2.7866,
      "step": 12126
    },
    {
      "epoch": 0.6850371537941905,
      "grad_norm": 0.32791271805763245,
      "learning_rate": 0.001,
      "loss": 2.779,
      "step": 12169
    },
    {
      "epoch": 0.687457779779329,
      "grad_norm": 0.3388611376285553,
      "learning_rate": 0.001,
      "loss": 2.7914,
      "step": 12212
    },
    {
      "epoch": 0.6898784057644675,
      "grad_norm": 0.29582807421684265,
      "learning_rate": 0.001,
      "loss": 2.7923,
      "step": 12255
    },
    {
      "epoch": 0.6922990317496059,
      "grad_norm": 0.3353387415409088,
      "learning_rate": 0.001,
      "loss": 2.7892,
      "step": 12298
    },
    {
      "epoch": 0.6947196577347444,
      "grad_norm": 0.27928057312965393,
      "learning_rate": 0.001,
      "loss": 2.7949,
      "step": 12341
    },
    {
      "epoch": 0.6971402837198829,
      "grad_norm": 0.3066178262233734,
      "learning_rate": 0.001,
      "loss": 2.7903,
      "step": 12384
    },
    {
      "epoch": 0.6995609097050214,
      "grad_norm": 0.41318798065185547,
      "learning_rate": 0.001,
      "loss": 2.803,
      "step": 12427
    },
    {
      "epoch": 0.7019815356901599,
      "grad_norm": 0.3321063816547394,
      "learning_rate": 0.001,
      "loss": 2.7856,
      "step": 12470
    },
    {
      "epoch": 0.7044021616752983,
      "grad_norm": 0.30045291781425476,
      "learning_rate": 0.001,
      "loss": 2.7954,
      "step": 12513
    },
    {
      "epoch": 0.7068227876604368,
      "grad_norm": 0.5469635725021362,
      "learning_rate": 0.001,
      "loss": 2.7961,
      "step": 12556
    },
    {
      "epoch": 0.7092434136455753,
      "grad_norm": 0.33449339866638184,
      "learning_rate": 0.001,
      "loss": 2.7821,
      "step": 12599
    },
    {
      "epoch": 0.7116640396307138,
      "grad_norm": 0.2796306014060974,
      "learning_rate": 0.001,
      "loss": 2.7927,
      "step": 12642
    },
    {
      "epoch": 0.7140846656158523,
      "grad_norm": 0.3235500454902649,
      "learning_rate": 0.001,
      "loss": 2.7876,
      "step": 12685
    },
    {
      "epoch": 0.7165052916009907,
      "grad_norm": 0.2499043345451355,
      "learning_rate": 0.001,
      "loss": 2.7891,
      "step": 12728
    },
    {
      "epoch": 0.7189259175861292,
      "grad_norm": 0.3014410734176636,
      "learning_rate": 0.001,
      "loss": 2.778,
      "step": 12771
    },
    {
      "epoch": 0.720445845530286,
      "eval_ag_news_bleu_score": 3.701128270454277,
      "eval_ag_news_bleu_score_sem": 0.142166255663979,
      "eval_ag_news_emb_cos_sim": 0.9904065132141113,
      "eval_ag_news_emb_cos_sim_sem": 0.00030119293006778867,
      "eval_ag_news_emb_top1_equal": 0.596666693687439,
      "eval_ag_news_emb_top1_equal_sem": 0.028370197491492683,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.910900354385376,
      "eval_ag_news_n_ngrams_match_1": 4.472,
      "eval_ag_news_n_ngrams_match_2": 0.77,
      "eval_ag_news_n_ngrams_match_3": 0.2,
      "eval_ag_news_num_pred_words": 22.94,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.23543439450411213,
      "eval_ag_news_runtime": 16.2248,
      "eval_ag_news_samples_per_second": 30.817,
      "eval_ag_news_steps_per_second": 0.123,
      "eval_ag_news_token_set_f1": 0.2094754862153183,
      "eval_ag_news_token_set_f1_sem": 0.004972660505551616,
      "eval_ag_news_token_set_precision": 0.19982207633201649,
      "eval_ag_news_token_set_recall": 0.23029661513371852,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 12798
    },
    {
      "epoch": 0.720445845530286,
      "eval_anthropic_toxic_prompts_bleu_score": 4.367969953696823,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.18457441584302203,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9903440475463867,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00028557139313197395,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6466666460037231,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027643749338232177,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.186903953552246,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.126,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.15,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.426,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.382,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.23954637763889075,
      "eval_anthropic_toxic_prompts_runtime": 6.8994,
      "eval_anthropic_toxic_prompts_samples_per_second": 72.47,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.29,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2788819929221169,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006359361329035184,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3249417255051937,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2703034299797062,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 12798
    },
    {
      "epoch": 0.720445845530286,
      "eval_arxiv_bleu_score": 3.2938101531131854,
      "eval_arxiv_bleu_score_sem": 0.1043276979750365,
      "eval_arxiv_emb_cos_sim": 0.990004301071167,
      "eval_arxiv_emb_cos_sim_sem": 0.00029780424129255585,
      "eval_arxiv_emb_top1_equal": 0.550000011920929,
      "eval_arxiv_emb_top1_equal_sem": 0.028770805007137933,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.9621994495391846,
      "eval_arxiv_n_ngrams_match_1": 4.222,
      "eval_arxiv_n_ngrams_match_2": 0.584,
      "eval_arxiv_n_ngrams_match_3": 0.106,
      "eval_arxiv_num_pred_words": 17.706,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.22343512324013148,
      "eval_arxiv_runtime": 7.0068,
      "eval_arxiv_samples_per_second": 71.359,
      "eval_arxiv_steps_per_second": 0.285,
      "eval_arxiv_token_set_f1": 0.2145909470460872,
      "eval_arxiv_token_set_f1_sem": 0.00473167224408577,
      "eval_arxiv_token_set_precision": 0.1782915402823801,
      "eval_arxiv_token_set_recall": 0.2943116486619078,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 12798
    },
    {
      "epoch": 0.720445845530286,
      "eval_python_code_alpaca_bleu_score": 5.033654437181438,
      "eval_python_code_alpaca_bleu_score_sem": 0.18369663691100008,
      "eval_python_code_alpaca_emb_cos_sim": 0.9882196187973022,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00042214796625159945,
      "eval_python_code_alpaca_emb_top1_equal": 0.6333333253860474,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02786867456387452,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.285463571548462,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.086,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.076,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.356,
      "eval_python_code_alpaca_num_pred_words": 19.652,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2850642025424246,
      "eval_python_code_alpaca_runtime": 6.8331,
      "eval_python_code_alpaca_samples_per_second": 73.173,
      "eval_python_code_alpaca_steps_per_second": 0.293,
      "eval_python_code_alpaca_token_set_f1": 0.3051026089958319,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005345107614202871,
      "eval_python_code_alpaca_token_set_precision": 0.2979160646386866,
      "eval_python_code_alpaca_token_set_recall": 0.34545845760706445,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 12798
    },
    {
      "epoch": 0.720445845530286,
      "eval_wikibio_bleu_score": 3.4335651475128794,
      "eval_wikibio_bleu_score_sem": 0.12306223898400077,
      "eval_wikibio_emb_cos_sim": 0.9905062913894653,
      "eval_wikibio_emb_cos_sim_sem": 0.00030943188196194996,
      "eval_wikibio_emb_top1_equal": 0.6133333444595337,
      "eval_wikibio_emb_top1_equal_sem": 0.028163139369651306,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.127765655517578,
      "eval_wikibio_n_ngrams_match_1": 2.988,
      "eval_wikibio_n_ngrams_match_2": 0.714,
      "eval_wikibio_n_ngrams_match_3": 0.146,
      "eval_wikibio_num_pred_words": 20.746,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.23592750618858807,
      "eval_wikibio_runtime": 6.6508,
      "eval_wikibio_samples_per_second": 75.178,
      "eval_wikibio_steps_per_second": 0.301,
      "eval_wikibio_token_set_f1": 0.19098669358877773,
      "eval_wikibio_token_set_f1_sem": 0.005152380089365418,
      "eval_wikibio_token_set_precision": 0.22653478744913128,
      "eval_wikibio_token_set_recall": 0.17181935068162454,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 12798
    },
    {
      "epoch": 0.720445845530286,
      "eval_nq_5round_bleu_score": 8.289047378747052,
      "eval_nq_5round_bleu_score_sem": 0.40511632851134965,
      "eval_nq_5round_emb_cos_sim": 0.9918684959411621,
      "eval_nq_5round_emb_cos_sim_sem": 0.0003082116270816679,
      "eval_nq_5round_emb_top1_equal": 0.6266666650772095,
      "eval_nq_5round_emb_top1_equal_sem": 0.027972489250684164,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.294,
      "eval_nq_5round_n_ngrams_match_2": 2.628,
      "eval_nq_5round_n_ngrams_match_3": 1.098,
      "eval_nq_5round_num_pred_words": 23.398,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3283072621977694,
      "eval_nq_5round_token_set_f1": 0.3633749188731967,
      "eval_nq_5round_token_set_f1_sem": 0.006292003665072815,
      "eval_nq_5round_token_set_precision": 0.3351697123002902,
      "eval_nq_5round_token_set_recall": 0.41308139733981675,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.026991914230447,
      "eval_nq_bleu_score_sem": 0.40761960153101234,
      "eval_nq_emb_cos_sim": 0.9922158122062683,
      "eval_nq_emb_cos_sim_sem": 0.00028406067306867955,
      "eval_nq_emb_top1_equal": 0.6433333158493042,
      "eval_nq_emb_top1_equal_sem": 0.02770216499443815,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.505817413330078,
      "eval_nq_n_ngrams_match_1": 8.836,
      "eval_nq_n_ngrams_match_2": 2.904,
      "eval_nq_n_ngrams_match_3": 1.244,
      "eval_nq_num_pred_words": 23.752,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3493100618305851,
      "eval_nq_runtime": 19.8753,
      "eval_nq_samples_per_second": 25.157,
      "eval_nq_steps_per_second": 0.101,
      "eval_nq_token_set_f1": 0.38592084449085184,
      "eval_nq_token_set_f1_sem": 0.006095817232517719,
      "eval_nq_token_set_precision": 0.35825582311869725,
      "eval_nq_token_set_recall": 0.4326123815036467,
      "eval_nq_true_num_tokens": 32.0,
      "step": 12798
    },
    {
      "epoch": 0.7213465435712677,
      "grad_norm": 0.4080064594745636,
      "learning_rate": 0.001,
      "loss": 2.7775,
      "step": 12814
    },
    {
      "epoch": 0.7237671695564062,
      "grad_norm": 0.3079150319099426,
      "learning_rate": 0.001,
      "loss": 2.7718,
      "step": 12857
    },
    {
      "epoch": 0.7261877955415447,
      "grad_norm": 0.3324522078037262,
      "learning_rate": 0.001,
      "loss": 2.7727,
      "step": 12900
    },
    {
      "epoch": 0.7286084215266831,
      "grad_norm": 0.302970290184021,
      "learning_rate": 0.001,
      "loss": 2.7813,
      "step": 12943
    },
    {
      "epoch": 0.7310290475118216,
      "grad_norm": 0.2800823450088501,
      "learning_rate": 0.001,
      "loss": 2.7884,
      "step": 12986
    },
    {
      "epoch": 0.7334496734969601,
      "grad_norm": 0.30401697754859924,
      "learning_rate": 0.001,
      "loss": 2.7831,
      "step": 13029
    },
    {
      "epoch": 0.7358702994820986,
      "grad_norm": 0.278766930103302,
      "learning_rate": 0.001,
      "loss": 2.7638,
      "step": 13072
    },
    {
      "epoch": 0.7382909254672371,
      "grad_norm": 0.307171493768692,
      "learning_rate": 0.001,
      "loss": 2.7695,
      "step": 13115
    },
    {
      "epoch": 0.7407115514523755,
      "grad_norm": 0.36417290568351746,
      "learning_rate": 0.001,
      "loss": 2.7814,
      "step": 13158
    },
    {
      "epoch": 0.743132177437514,
      "grad_norm": 0.3266918957233429,
      "learning_rate": 0.001,
      "loss": 2.7773,
      "step": 13201
    },
    {
      "epoch": 0.7455528034226525,
      "grad_norm": 0.2807528078556061,
      "learning_rate": 0.001,
      "loss": 2.7836,
      "step": 13244
    },
    {
      "epoch": 0.747973429407791,
      "grad_norm": 0.3334343433380127,
      "learning_rate": 0.001,
      "loss": 2.7791,
      "step": 13287
    },
    {
      "epoch": 0.7503940553929295,
      "grad_norm": 0.3244103491306305,
      "learning_rate": 0.001,
      "loss": 2.778,
      "step": 13330
    },
    {
      "epoch": 0.752814681378068,
      "grad_norm": 0.3411544859409332,
      "learning_rate": 0.001,
      "loss": 2.7745,
      "step": 13373
    },
    {
      "epoch": 0.7552353073632064,
      "grad_norm": 0.32141536474227905,
      "learning_rate": 0.001,
      "loss": 2.7675,
      "step": 13416
    },
    {
      "epoch": 0.7576559333483449,
      "grad_norm": 0.3332141637802124,
      "learning_rate": 0.001,
      "loss": 2.7739,
      "step": 13459
    },
    {
      "epoch": 0.7600765593334835,
      "grad_norm": 0.2830438017845154,
      "learning_rate": 0.001,
      "loss": 2.7753,
      "step": 13502
    },
    {
      "epoch": 0.762497185318622,
      "grad_norm": 0.30623412132263184,
      "learning_rate": 0.001,
      "loss": 2.7677,
      "step": 13545
    },
    {
      "epoch": 0.7649178113037605,
      "grad_norm": 0.3300396203994751,
      "learning_rate": 0.001,
      "loss": 2.7695,
      "step": 13588
    },
    {
      "epoch": 0.767338437288899,
      "grad_norm": 0.27616170048713684,
      "learning_rate": 0.001,
      "loss": 2.7736,
      "step": 13631
    },
    {
      "epoch": 0.7697590632740374,
      "grad_norm": 0.31893834471702576,
      "learning_rate": 0.001,
      "loss": 2.7703,
      "step": 13674
    },
    {
      "epoch": 0.7721796892591759,
      "grad_norm": 0.2808237671852112,
      "learning_rate": 0.001,
      "loss": 2.7743,
      "step": 13717
    },
    {
      "epoch": 0.7746003152443144,
      "grad_norm": 0.2813253700733185,
      "learning_rate": 0.001,
      "loss": 2.7604,
      "step": 13760
    },
    {
      "epoch": 0.7770209412294529,
      "grad_norm": 0.2873164415359497,
      "learning_rate": 0.001,
      "loss": 2.7705,
      "step": 13803
    },
    {
      "epoch": 0.7794415672145913,
      "grad_norm": 0.2793399691581726,
      "learning_rate": 0.001,
      "loss": 2.7697,
      "step": 13846
    },
    {
      "epoch": 0.7818621931997298,
      "grad_norm": 0.36446264386177063,
      "learning_rate": 0.001,
      "loss": 2.7575,
      "step": 13889
    },
    {
      "epoch": 0.7842828191848683,
      "grad_norm": 0.36235177516937256,
      "learning_rate": 0.001,
      "loss": 2.7517,
      "step": 13932
    },
    {
      "epoch": 0.7867034451700068,
      "grad_norm": 0.2927075922489166,
      "learning_rate": 0.001,
      "loss": 2.7579,
      "step": 13975
    },
    {
      "epoch": 0.7891240711551453,
      "grad_norm": 0.29674389958381653,
      "learning_rate": 0.001,
      "loss": 2.7606,
      "step": 14018
    },
    {
      "epoch": 0.7915446971402837,
      "grad_norm": 0.40778595209121704,
      "learning_rate": 0.001,
      "loss": 2.767,
      "step": 14061
    },
    {
      "epoch": 0.7939653231254222,
      "grad_norm": 0.3394840657711029,
      "learning_rate": 0.001,
      "loss": 2.7672,
      "step": 14104
    },
    {
      "epoch": 0.7963859491105607,
      "grad_norm": 0.32396915555000305,
      "learning_rate": 0.001,
      "loss": 2.769,
      "step": 14147
    },
    {
      "epoch": 0.7988065750956992,
      "grad_norm": 0.33721813559532166,
      "learning_rate": 0.001,
      "loss": 2.7541,
      "step": 14190
    },
    {
      "epoch": 0.8012272010808377,
      "grad_norm": 0.3539986312389374,
      "learning_rate": 0.001,
      "loss": 2.773,
      "step": 14233
    },
    {
      "epoch": 0.8036478270659762,
      "grad_norm": 0.310371071100235,
      "learning_rate": 0.001,
      "loss": 2.7625,
      "step": 14276
    },
    {
      "epoch": 0.8060684530511146,
      "grad_norm": 0.3418547511100769,
      "learning_rate": 0.001,
      "loss": 2.7619,
      "step": 14319
    },
    {
      "epoch": 0.8084890790362531,
      "grad_norm": 0.35570278763771057,
      "learning_rate": 0.001,
      "loss": 2.7558,
      "step": 14362
    },
    {
      "epoch": 0.8109097050213916,
      "grad_norm": 0.2710742950439453,
      "learning_rate": 0.001,
      "loss": 2.7624,
      "step": 14405
    },
    {
      "epoch": 0.8133303310065301,
      "grad_norm": 0.33577442169189453,
      "learning_rate": 0.001,
      "loss": 2.76,
      "step": 14448
    },
    {
      "epoch": 0.8157509569916686,
      "grad_norm": 0.3735579252243042,
      "learning_rate": 0.001,
      "loss": 2.7623,
      "step": 14491
    },
    {
      "epoch": 0.818171582976807,
      "grad_norm": 0.3267165422439575,
      "learning_rate": 0.001,
      "loss": 2.7562,
      "step": 14534
    },
    {
      "epoch": 0.8205922089619455,
      "grad_norm": 0.33206281065940857,
      "learning_rate": 0.001,
      "loss": 2.7522,
      "step": 14577
    },
    {
      "epoch": 0.823012834947084,
      "grad_norm": 0.2644685208797455,
      "learning_rate": 0.001,
      "loss": 2.7574,
      "step": 14620
    },
    {
      "epoch": 0.8254334609322225,
      "grad_norm": 0.31752756237983704,
      "learning_rate": 0.001,
      "loss": 2.7558,
      "step": 14663
    },
    {
      "epoch": 0.827854086917361,
      "grad_norm": 0.32095977663993835,
      "learning_rate": 0.001,
      "loss": 2.7546,
      "step": 14706
    },
    {
      "epoch": 0.8302747129024994,
      "grad_norm": 0.30490657687187195,
      "learning_rate": 0.001,
      "loss": 2.751,
      "step": 14749
    },
    {
      "epoch": 0.8326953388876379,
      "grad_norm": 0.29289528727531433,
      "learning_rate": 0.001,
      "loss": 2.7596,
      "step": 14792
    },
    {
      "epoch": 0.8351159648727764,
      "grad_norm": 0.2679915428161621,
      "learning_rate": 0.001,
      "loss": 2.7435,
      "step": 14835
    },
    {
      "epoch": 0.8375365908579149,
      "grad_norm": 0.30280086398124695,
      "learning_rate": 0.001,
      "loss": 2.738,
      "step": 14878
    },
    {
      "epoch": 0.8399572168430534,
      "grad_norm": 0.34722641110420227,
      "learning_rate": 0.001,
      "loss": 2.7391,
      "step": 14921
    },
    {
      "epoch": 0.840520153118667,
      "eval_ag_news_bleu_score": 3.5377604437781303,
      "eval_ag_news_bleu_score_sem": 0.12897160687231332,
      "eval_ag_news_emb_cos_sim": 0.9899711608886719,
      "eval_ag_news_emb_cos_sim_sem": 0.00035444073104016284,
      "eval_ag_news_emb_top1_equal": 0.6066666841506958,
      "eval_ag_news_emb_top1_equal_sem": 0.028250091805185003,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.923144578933716,
      "eval_ag_news_n_ngrams_match_1": 4.48,
      "eval_ag_news_n_ngrams_match_2": 0.724,
      "eval_ag_news_n_ngrams_match_3": 0.176,
      "eval_ag_news_num_pred_words": 22.958,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.23744382934222757,
      "eval_ag_news_runtime": 16.0907,
      "eval_ag_news_samples_per_second": 31.074,
      "eval_ag_news_steps_per_second": 0.124,
      "eval_ag_news_token_set_f1": 0.2102386636670784,
      "eval_ag_news_token_set_f1_sem": 0.004949977179864459,
      "eval_ag_news_token_set_precision": 0.20166052592881534,
      "eval_ag_news_token_set_recall": 0.22636360027687147,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 14931
    },
    {
      "epoch": 0.840520153118667,
      "eval_anthropic_toxic_prompts_bleu_score": 4.462669450651725,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.18377186205730286,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9903033971786499,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0003145579848418657,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6233333349227905,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028022262142792274,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.141181230545044,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.292,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.212,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.46,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.37,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2472661960561794,
      "eval_anthropic_toxic_prompts_runtime": 7.0178,
      "eval_anthropic_toxic_prompts_samples_per_second": 71.247,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.285,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2839141385983301,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006460189277455879,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3388580132457614,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.26943807578707746,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 14931
    },
    {
      "epoch": 0.840520153118667,
      "eval_arxiv_bleu_score": 3.340778516942671,
      "eval_arxiv_bleu_score_sem": 0.09726623692555353,
      "eval_arxiv_emb_cos_sim": 0.9899803996086121,
      "eval_arxiv_emb_cos_sim_sem": 0.0003364449020174012,
      "eval_arxiv_emb_top1_equal": 0.5299999713897705,
      "eval_arxiv_emb_top1_equal_sem": 0.028863650627428784,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.962177276611328,
      "eval_arxiv_n_ngrams_match_1": 4.33,
      "eval_arxiv_n_ngrams_match_2": 0.604,
      "eval_arxiv_n_ngrams_match_3": 0.102,
      "eval_arxiv_num_pred_words": 17.996,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2252235158782392,
      "eval_arxiv_runtime": 6.9966,
      "eval_arxiv_samples_per_second": 71.463,
      "eval_arxiv_steps_per_second": 0.286,
      "eval_arxiv_token_set_f1": 0.21555172187628194,
      "eval_arxiv_token_set_f1_sem": 0.004458429742817735,
      "eval_arxiv_token_set_precision": 0.1811493213493959,
      "eval_arxiv_token_set_recall": 0.2856677682229383,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 14931
    },
    {
      "epoch": 0.840520153118667,
      "eval_python_code_alpaca_bleu_score": 5.132636235804955,
      "eval_python_code_alpaca_bleu_score_sem": 0.19539714661750565,
      "eval_python_code_alpaca_emb_cos_sim": 0.9876632690429688,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004713367017976472,
      "eval_python_code_alpaca_emb_top1_equal": 0.5866666436195374,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02847805567733279,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.165525436401367,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.144,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.074,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.33,
      "eval_python_code_alpaca_num_pred_words": 19.238,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.28916644531406577,
      "eval_python_code_alpaca_runtime": 7.0494,
      "eval_python_code_alpaca_samples_per_second": 70.928,
      "eval_python_code_alpaca_steps_per_second": 0.284,
      "eval_python_code_alpaca_token_set_f1": 0.30913852887078225,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005592822897559268,
      "eval_python_code_alpaca_token_set_precision": 0.29994801142967464,
      "eval_python_code_alpaca_token_set_recall": 0.34547530190304454,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 14931
    },
    {
      "epoch": 0.840520153118667,
      "eval_wikibio_bleu_score": 3.6225134238005037,
      "eval_wikibio_bleu_score_sem": 0.12567650142372055,
      "eval_wikibio_emb_cos_sim": 0.9905586242675781,
      "eval_wikibio_emb_cos_sim_sem": 0.00028784330726474547,
      "eval_wikibio_emb_top1_equal": 0.6299999952316284,
      "eval_wikibio_emb_top1_equal_sem": 0.027921293391044915,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.099316596984863,
      "eval_wikibio_n_ngrams_match_1": 3.034,
      "eval_wikibio_n_ngrams_match_2": 0.742,
      "eval_wikibio_n_ngrams_match_3": 0.15,
      "eval_wikibio_num_pred_words": 20.528,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.241950567288387,
      "eval_wikibio_runtime": 6.8852,
      "eval_wikibio_samples_per_second": 72.619,
      "eval_wikibio_steps_per_second": 0.29,
      "eval_wikibio_token_set_f1": 0.19606182267684533,
      "eval_wikibio_token_set_f1_sem": 0.00504714466311069,
      "eval_wikibio_token_set_precision": 0.23124122812232117,
      "eval_wikibio_token_set_recall": 0.17675501214225312,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 14931
    },
    {
      "epoch": 0.840520153118667,
      "eval_nq_5round_bleu_score": 8.43652610783253,
      "eval_nq_5round_bleu_score_sem": 0.4117587559025611,
      "eval_nq_5round_emb_cos_sim": 0.9922388792037964,
      "eval_nq_5round_emb_cos_sim_sem": 0.00023403931009594774,
      "eval_nq_5round_emb_top1_equal": 0.6266666650772095,
      "eval_nq_5round_emb_top1_equal_sem": 0.027972489250684164,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.324,
      "eval_nq_5round_n_ngrams_match_2": 2.668,
      "eval_nq_5round_n_ngrams_match_3": 1.154,
      "eval_nq_5round_num_pred_words": 23.202,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.33290193096337617,
      "eval_nq_5round_token_set_f1": 0.364394619188876,
      "eval_nq_5round_token_set_f1_sem": 0.006371388418485476,
      "eval_nq_5round_token_set_precision": 0.3365218673576029,
      "eval_nq_5round_token_set_recall": 0.4127168262713495,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.794283251036768,
      "eval_nq_bleu_score_sem": 0.40438699298732345,
      "eval_nq_emb_cos_sim": 0.9925456047058105,
      "eval_nq_emb_cos_sim_sem": 0.00022823411983564416,
      "eval_nq_emb_top1_equal": 0.653333306312561,
      "eval_nq_emb_top1_equal_sem": 0.027522497707092888,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.481304168701172,
      "eval_nq_n_ngrams_match_1": 8.846,
      "eval_nq_n_ngrams_match_2": 2.884,
      "eval_nq_n_ngrams_match_3": 1.192,
      "eval_nq_num_pred_words": 23.536,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3522717008970166,
      "eval_nq_runtime": 20.1491,
      "eval_nq_samples_per_second": 24.815,
      "eval_nq_steps_per_second": 0.099,
      "eval_nq_token_set_f1": 0.38689329702494124,
      "eval_nq_token_set_f1_sem": 0.00618108585165258,
      "eval_nq_token_set_precision": 0.3603190512380785,
      "eval_nq_token_set_recall": 0.42994469001815794,
      "eval_nq_true_num_tokens": 32.0,
      "step": 14931
    },
    {
      "epoch": 0.8423778428281918,
      "grad_norm": 0.3114996552467346,
      "learning_rate": 0.001,
      "loss": 2.7481,
      "step": 14964
    },
    {
      "epoch": 0.8447984688133303,
      "grad_norm": 0.34977731108665466,
      "learning_rate": 0.001,
      "loss": 2.7559,
      "step": 15007
    },
    {
      "epoch": 0.8472190947984688,
      "grad_norm": 0.343776673078537,
      "learning_rate": 0.001,
      "loss": 2.7588,
      "step": 15050
    },
    {
      "epoch": 0.8496397207836073,
      "grad_norm": 0.3556782305240631,
      "learning_rate": 0.001,
      "loss": 2.7592,
      "step": 15093
    },
    {
      "epoch": 0.8520603467687458,
      "grad_norm": 0.33249431848526,
      "learning_rate": 0.001,
      "loss": 2.7524,
      "step": 15136
    },
    {
      "epoch": 0.8544809727538842,
      "grad_norm": 0.4269940257072449,
      "learning_rate": 0.001,
      "loss": 2.7466,
      "step": 15179
    },
    {
      "epoch": 0.8569015987390227,
      "grad_norm": 0.3385619819164276,
      "learning_rate": 0.001,
      "loss": 2.7331,
      "step": 15222
    },
    {
      "epoch": 0.8593222247241612,
      "grad_norm": 0.2910108268260956,
      "learning_rate": 0.001,
      "loss": 2.7446,
      "step": 15265
    },
    {
      "epoch": 0.8617428507092997,
      "grad_norm": 0.32539159059524536,
      "learning_rate": 0.001,
      "loss": 2.7533,
      "step": 15308
    },
    {
      "epoch": 0.8641634766944382,
      "grad_norm": 0.25635749101638794,
      "learning_rate": 0.001,
      "loss": 2.7475,
      "step": 15351
    },
    {
      "epoch": 0.8665841026795766,
      "grad_norm": 0.6902783513069153,
      "learning_rate": 0.001,
      "loss": 2.7466,
      "step": 15394
    },
    {
      "epoch": 0.8690047286647151,
      "grad_norm": 0.38919487595558167,
      "learning_rate": 0.001,
      "loss": 2.7525,
      "step": 15437
    },
    {
      "epoch": 0.8714253546498536,
      "grad_norm": 0.3098400831222534,
      "learning_rate": 0.001,
      "loss": 2.742,
      "step": 15480
    },
    {
      "epoch": 0.8738459806349921,
      "grad_norm": 0.37423354387283325,
      "learning_rate": 0.001,
      "loss": 2.741,
      "step": 15523
    },
    {
      "epoch": 0.8762666066201306,
      "grad_norm": 0.2834371328353882,
      "learning_rate": 0.001,
      "loss": 2.7531,
      "step": 15566
    },
    {
      "epoch": 0.878687232605269,
      "grad_norm": 0.33132556080818176,
      "learning_rate": 0.001,
      "loss": 2.7381,
      "step": 15609
    },
    {
      "epoch": 0.8811078585904075,
      "grad_norm": 0.33835652470588684,
      "learning_rate": 0.001,
      "loss": 2.7371,
      "step": 15652
    },
    {
      "epoch": 0.883528484575546,
      "grad_norm": 0.3738958537578583,
      "learning_rate": 0.001,
      "loss": 2.7441,
      "step": 15695
    },
    {
      "epoch": 0.8859491105606845,
      "grad_norm": 1.5895246267318726,
      "learning_rate": 0.001,
      "loss": 2.7326,
      "step": 15738
    },
    {
      "epoch": 0.888369736545823,
      "grad_norm": 0.42023199796676636,
      "learning_rate": 0.001,
      "loss": 2.7471,
      "step": 15781
    },
    {
      "epoch": 0.8907903625309614,
      "grad_norm": 0.34421205520629883,
      "learning_rate": 0.001,
      "loss": 2.7408,
      "step": 15824
    },
    {
      "epoch": 0.8932109885160999,
      "grad_norm": 0.3311266601085663,
      "learning_rate": 0.001,
      "loss": 2.7426,
      "step": 15867
    },
    {
      "epoch": 0.8956316145012384,
      "grad_norm": 0.3178233206272125,
      "learning_rate": 0.001,
      "loss": 2.7507,
      "step": 15910
    },
    {
      "epoch": 0.898052240486377,
      "grad_norm": 0.351895272731781,
      "learning_rate": 0.001,
      "loss": 2.7334,
      "step": 15953
    },
    {
      "epoch": 0.9004728664715155,
      "grad_norm": 0.2933900058269501,
      "learning_rate": 0.001,
      "loss": 2.7341,
      "step": 15996
    },
    {
      "epoch": 0.902893492456654,
      "grad_norm": 1.671920657157898,
      "learning_rate": 0.001,
      "loss": 2.7346,
      "step": 16039
    },
    {
      "epoch": 0.9053141184417924,
      "grad_norm": 0.3117462992668152,
      "learning_rate": 0.001,
      "loss": 2.7403,
      "step": 16082
    },
    {
      "epoch": 0.9077347444269309,
      "grad_norm": 0.35146528482437134,
      "learning_rate": 0.001,
      "loss": 2.7277,
      "step": 16125
    },
    {
      "epoch": 0.9101553704120694,
      "grad_norm": 0.4618609547615051,
      "learning_rate": 0.001,
      "loss": 2.7289,
      "step": 16168
    },
    {
      "epoch": 0.9125759963972079,
      "grad_norm": 0.25520193576812744,
      "learning_rate": 0.001,
      "loss": 2.7327,
      "step": 16211
    },
    {
      "epoch": 0.9149966223823464,
      "grad_norm": 0.341301828622818,
      "learning_rate": 0.001,
      "loss": 2.7213,
      "step": 16254
    },
    {
      "epoch": 0.9174172483674848,
      "grad_norm": 0.25876420736312866,
      "learning_rate": 0.001,
      "loss": 2.7282,
      "step": 16297
    },
    {
      "epoch": 0.9198378743526233,
      "grad_norm": 0.29863297939300537,
      "learning_rate": 0.001,
      "loss": 2.7138,
      "step": 16340
    },
    {
      "epoch": 0.9222585003377618,
      "grad_norm": 0.3050626218318939,
      "learning_rate": 0.001,
      "loss": 2.7314,
      "step": 16383
    },
    {
      "epoch": 0.9246791263229003,
      "grad_norm": 0.3613375425338745,
      "learning_rate": 0.001,
      "loss": 2.7286,
      "step": 16426
    },
    {
      "epoch": 0.9270997523080388,
      "grad_norm": 0.3157224953174591,
      "learning_rate": 0.001,
      "loss": 2.7403,
      "step": 16469
    },
    {
      "epoch": 0.9295203782931772,
      "grad_norm": 0.8742790222167969,
      "learning_rate": 0.001,
      "loss": 2.7184,
      "step": 16512
    },
    {
      "epoch": 0.9319410042783157,
      "grad_norm": 0.30806833505630493,
      "learning_rate": 0.001,
      "loss": 2.7194,
      "step": 16555
    },
    {
      "epoch": 0.9343616302634542,
      "grad_norm": 0.3987646698951721,
      "learning_rate": 0.001,
      "loss": 2.7169,
      "step": 16598
    },
    {
      "epoch": 0.9367822562485927,
      "grad_norm": 0.32958006858825684,
      "learning_rate": 0.001,
      "loss": 2.7255,
      "step": 16641
    },
    {
      "epoch": 0.9392028822337312,
      "grad_norm": 0.3005693256855011,
      "learning_rate": 0.001,
      "loss": 2.7326,
      "step": 16684
    },
    {
      "epoch": 0.9416235082188696,
      "grad_norm": 0.3229604661464691,
      "learning_rate": 0.001,
      "loss": 2.7271,
      "step": 16727
    },
    {
      "epoch": 0.9440441342040081,
      "grad_norm": 0.3734353184700012,
      "learning_rate": 0.001,
      "loss": 2.723,
      "step": 16770
    },
    {
      "epoch": 0.9464647601891466,
      "grad_norm": 0.3179822266101837,
      "learning_rate": 0.001,
      "loss": 2.7195,
      "step": 16813
    },
    {
      "epoch": 0.9488853861742851,
      "grad_norm": 0.3966822624206543,
      "learning_rate": 0.001,
      "loss": 2.7102,
      "step": 16856
    },
    {
      "epoch": 0.9513060121594236,
      "grad_norm": 0.3697901964187622,
      "learning_rate": 0.001,
      "loss": 2.728,
      "step": 16899
    },
    {
      "epoch": 0.953726638144562,
      "grad_norm": 0.37111711502075195,
      "learning_rate": 0.001,
      "loss": 2.7165,
      "step": 16942
    },
    {
      "epoch": 0.9561472641297005,
      "grad_norm": 0.31283438205718994,
      "learning_rate": 0.001,
      "loss": 2.7174,
      "step": 16985
    },
    {
      "epoch": 0.958567890114839,
      "grad_norm": 0.352120965719223,
      "learning_rate": 0.001,
      "loss": 2.7244,
      "step": 17028
    },
    {
      "epoch": 0.960594460707048,
      "eval_ag_news_bleu_score": 3.727698047534403,
      "eval_ag_news_bleu_score_sem": 0.13571859340620013,
      "eval_ag_news_emb_cos_sim": 0.9904396533966064,
      "eval_ag_news_emb_cos_sim_sem": 0.00029136625956496576,
      "eval_ag_news_emb_top1_equal": 0.6233333349227905,
      "eval_ag_news_emb_top1_equal_sem": 0.028022260422154388,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.8957502841949463,
      "eval_ag_news_n_ngrams_match_1": 4.518,
      "eval_ag_news_n_ngrams_match_2": 0.78,
      "eval_ag_news_n_ngrams_match_3": 0.196,
      "eval_ag_news_num_pred_words": 23.002,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2377441943203574,
      "eval_ag_news_runtime": 37.286,
      "eval_ag_news_samples_per_second": 13.41,
      "eval_ag_news_steps_per_second": 0.054,
      "eval_ag_news_token_set_f1": 0.2110801753538497,
      "eval_ag_news_token_set_f1_sem": 0.004849267391087997,
      "eval_ag_news_token_set_precision": 0.20164164518043523,
      "eval_ag_news_token_set_recall": 0.2307404525670561,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 17064
    },
    {
      "epoch": 0.960594460707048,
      "eval_anthropic_toxic_prompts_bleu_score": 4.609576666626912,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2048153347353945,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9906299114227295,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.000287210542682426,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6000000238418579,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02833152959629631,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.1257970333099365,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.36,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.22,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.484,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.922,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.25043876477834426,
      "eval_anthropic_toxic_prompts_runtime": 28.7254,
      "eval_anthropic_toxic_prompts_samples_per_second": 17.406,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.07,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2877874903912111,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006261258650256147,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3408938611172731,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2738354513293552,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 17064
    },
    {
      "epoch": 0.960594460707048,
      "eval_arxiv_bleu_score": 3.354481409416532,
      "eval_arxiv_bleu_score_sem": 0.09662613444883023,
      "eval_arxiv_emb_cos_sim": 0.9902742505073547,
      "eval_arxiv_emb_cos_sim_sem": 0.00025158427822282657,
      "eval_arxiv_emb_top1_equal": 0.54666668176651,
      "eval_arxiv_emb_top1_equal_sem": 0.028789527267967893,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.9123644828796387,
      "eval_arxiv_n_ngrams_match_1": 4.396,
      "eval_arxiv_n_ngrams_match_2": 0.614,
      "eval_arxiv_n_ngrams_match_3": 0.096,
      "eval_arxiv_num_pred_words": 17.748,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.23053318757857005,
      "eval_arxiv_runtime": 29.2639,
      "eval_arxiv_samples_per_second": 17.086,
      "eval_arxiv_steps_per_second": 0.068,
      "eval_arxiv_token_set_f1": 0.221103434421826,
      "eval_arxiv_token_set_f1_sem": 0.004367761451065217,
      "eval_arxiv_token_set_precision": 0.18496595315536898,
      "eval_arxiv_token_set_recall": 0.2975203701973982,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 17064
    },
    {
      "epoch": 0.960594460707048,
      "eval_python_code_alpaca_bleu_score": 4.9295545899884585,
      "eval_python_code_alpaca_bleu_score_sem": 0.16284664414880975,
      "eval_python_code_alpaca_emb_cos_sim": 0.9882261157035828,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004210918978645289,
      "eval_python_code_alpaca_emb_top1_equal": 0.6633333563804626,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027329419768069584,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.156618356704712,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.128,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.058,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.31,
      "eval_python_code_alpaca_num_pred_words": 19.23,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2901142252728618,
      "eval_python_code_alpaca_runtime": 33.4029,
      "eval_python_code_alpaca_samples_per_second": 14.969,
      "eval_python_code_alpaca_steps_per_second": 0.06,
      "eval_python_code_alpaca_token_set_f1": 0.3103592557521753,
      "eval_python_code_alpaca_token_set_f1_sem": 0.00564202262615087,
      "eval_python_code_alpaca_token_set_precision": 0.2982031556194514,
      "eval_python_code_alpaca_token_set_recall": 0.3571020695316264,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 17064
    },
    {
      "epoch": 0.960594460707048,
      "eval_wikibio_bleu_score": 3.6250084692466418,
      "eval_wikibio_bleu_score_sem": 0.13575064044625604,
      "eval_wikibio_emb_cos_sim": 0.9904478192329407,
      "eval_wikibio_emb_cos_sim_sem": 0.0003124584840021947,
      "eval_wikibio_emb_top1_equal": 0.6200000047683716,
      "eval_wikibio_emb_top1_equal_sem": 0.028070624111834433,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.05482816696167,
      "eval_wikibio_n_ngrams_match_1": 3.068,
      "eval_wikibio_n_ngrams_match_2": 0.768,
      "eval_wikibio_n_ngrams_match_3": 0.166,
      "eval_wikibio_num_pred_words": 20.632,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2412733300930539,
      "eval_wikibio_runtime": 28.2428,
      "eval_wikibio_samples_per_second": 17.704,
      "eval_wikibio_steps_per_second": 0.071,
      "eval_wikibio_token_set_f1": 0.19745993915580426,
      "eval_wikibio_token_set_f1_sem": 0.005258238155461177,
      "eval_wikibio_token_set_precision": 0.23318784222500408,
      "eval_wikibio_token_set_recall": 0.17777562993511822,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 17064
    },
    {
      "epoch": 0.960594460707048,
      "eval_nq_5round_bleu_score": 8.571913240901864,
      "eval_nq_5round_bleu_score_sem": 0.4228544868844374,
      "eval_nq_5round_emb_cos_sim": 0.9922306537628174,
      "eval_nq_5round_emb_cos_sim_sem": 0.00030978281143565053,
      "eval_nq_5round_emb_top1_equal": 0.653333306312561,
      "eval_nq_5round_emb_top1_equal_sem": 0.027522499427730773,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.514,
      "eval_nq_5round_n_ngrams_match_2": 2.67,
      "eval_nq_5round_n_ngrams_match_3": 1.134,
      "eval_nq_5round_num_pred_words": 23.408,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3355986473171025,
      "eval_nq_5round_token_set_f1": 0.3728641971858251,
      "eval_nq_5round_token_set_f1_sem": 0.00634927536775864,
      "eval_nq_5round_token_set_precision": 0.34612737400694726,
      "eval_nq_5round_token_set_recall": 0.41788611965404043,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.217578221471921,
      "eval_nq_bleu_score_sem": 0.41932078051753474,
      "eval_nq_emb_cos_sim": 0.9926722645759583,
      "eval_nq_emb_cos_sim_sem": 0.0002216008994780063,
      "eval_nq_emb_top1_equal": 0.6299999952316284,
      "eval_nq_emb_top1_equal_sem": 0.027921293391044915,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.4414749145507812,
      "eval_nq_n_ngrams_match_1": 8.952,
      "eval_nq_n_ngrams_match_2": 2.974,
      "eval_nq_n_ngrams_match_3": 1.274,
      "eval_nq_num_pred_words": 23.81,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.35224613207824185,
      "eval_nq_runtime": 62.3339,
      "eval_nq_samples_per_second": 8.021,
      "eval_nq_steps_per_second": 0.032,
      "eval_nq_token_set_f1": 0.3886862522496393,
      "eval_nq_token_set_f1_sem": 0.006239003615979623,
      "eval_nq_token_set_precision": 0.36297820302136763,
      "eval_nq_token_set_recall": 0.43198140010211905,
      "eval_nq_true_num_tokens": 32.0,
      "step": 17064
    },
    {
      "epoch": 0.9609885160999775,
      "grad_norm": 0.31058382987976074,
      "learning_rate": 0.001,
      "loss": 2.7184,
      "step": 17071
    },
    {
      "epoch": 0.963409142085116,
      "grad_norm": 0.31824374198913574,
      "learning_rate": 0.001,
      "loss": 2.7136,
      "step": 17114
    },
    {
      "epoch": 0.9658297680702544,
      "grad_norm": 0.3818407654762268,
      "learning_rate": 0.001,
      "loss": 2.7202,
      "step": 17157
    },
    {
      "epoch": 0.9682503940553929,
      "grad_norm": 2.440358877182007,
      "learning_rate": 0.001,
      "loss": 2.7221,
      "step": 17200
    },
    {
      "epoch": 0.9706710200405314,
      "grad_norm": 0.3153367042541504,
      "learning_rate": 0.001,
      "loss": 2.7237,
      "step": 17243
    },
    {
      "epoch": 0.9730916460256699,
      "grad_norm": 0.3164927363395691,
      "learning_rate": 0.001,
      "loss": 2.7167,
      "step": 17286
    },
    {
      "epoch": 0.9755122720108084,
      "grad_norm": 0.28781813383102417,
      "learning_rate": 0.001,
      "loss": 2.7123,
      "step": 17329
    },
    {
      "epoch": 0.9779328979959468,
      "grad_norm": 0.3262360095977783,
      "learning_rate": 0.001,
      "loss": 2.714,
      "step": 17372
    },
    {
      "epoch": 0.9803535239810853,
      "grad_norm": 0.31895849108695984,
      "learning_rate": 0.001,
      "loss": 2.714,
      "step": 17415
    },
    {
      "epoch": 0.9827741499662238,
      "grad_norm": 0.28788313269615173,
      "learning_rate": 0.001,
      "loss": 2.7029,
      "step": 17458
    },
    {
      "epoch": 0.9851947759513623,
      "grad_norm": 0.37557026743888855,
      "learning_rate": 0.001,
      "loss": 2.7037,
      "step": 17501
    },
    {
      "epoch": 0.9876154019365008,
      "grad_norm": 0.3447340428829193,
      "learning_rate": 0.001,
      "loss": 2.7085,
      "step": 17544
    },
    {
      "epoch": 0.9900360279216393,
      "grad_norm": 0.306481271982193,
      "learning_rate": 0.001,
      "loss": 2.7202,
      "step": 17587
    },
    {
      "epoch": 0.9924566539067777,
      "grad_norm": 0.3163767457008362,
      "learning_rate": 0.001,
      "loss": 2.6975,
      "step": 17630
    },
    {
      "epoch": 0.9948772798919162,
      "grad_norm": 0.3638165593147278,
      "learning_rate": 0.001,
      "loss": 2.7124,
      "step": 17673
    },
    {
      "epoch": 0.9972979058770547,
      "grad_norm": 0.3889255225658417,
      "learning_rate": 0.001,
      "loss": 2.7124,
      "step": 17716
    },
    {
      "epoch": 0.9997185318621932,
      "grad_norm": 0.3365430533885956,
      "learning_rate": 0.001,
      "loss": 2.7149,
      "step": 17759
    },
    {
      "epoch": 1.0021391578473318,
      "grad_norm": 0.29857319593429565,
      "learning_rate": 0.001,
      "loss": 2.6695,
      "step": 17802
    },
    {
      "epoch": 1.0045597838324702,
      "grad_norm": 0.3036225438117981,
      "learning_rate": 0.001,
      "loss": 2.6692,
      "step": 17845
    },
    {
      "epoch": 1.0069804098176087,
      "grad_norm": 0.3170844316482544,
      "learning_rate": 0.001,
      "loss": 2.6616,
      "step": 17888
    },
    {
      "epoch": 1.0094010358027472,
      "grad_norm": 0.28965848684310913,
      "learning_rate": 0.001,
      "loss": 2.6562,
      "step": 17931
    },
    {
      "epoch": 1.0118216617878857,
      "grad_norm": 0.3531702160835266,
      "learning_rate": 0.001,
      "loss": 2.6633,
      "step": 17974
    },
    {
      "epoch": 1.0142422877730242,
      "grad_norm": 0.29570698738098145,
      "learning_rate": 0.001,
      "loss": 2.6507,
      "step": 18017
    },
    {
      "epoch": 1.0166629137581626,
      "grad_norm": 0.31439104676246643,
      "learning_rate": 0.001,
      "loss": 2.6618,
      "step": 18060
    },
    {
      "epoch": 1.0190835397433011,
      "grad_norm": 0.2981725335121155,
      "learning_rate": 0.001,
      "loss": 2.6682,
      "step": 18103
    },
    {
      "epoch": 1.0215041657284396,
      "grad_norm": 0.27765563130378723,
      "learning_rate": 0.001,
      "loss": 2.6669,
      "step": 18146
    },
    {
      "epoch": 1.023924791713578,
      "grad_norm": 0.2958226501941681,
      "learning_rate": 0.001,
      "loss": 2.6535,
      "step": 18189
    },
    {
      "epoch": 1.0263454176987166,
      "grad_norm": 0.36111190915107727,
      "learning_rate": 0.001,
      "loss": 2.6577,
      "step": 18232
    },
    {
      "epoch": 1.028766043683855,
      "grad_norm": 0.34004342555999756,
      "learning_rate": 0.001,
      "loss": 2.6557,
      "step": 18275
    },
    {
      "epoch": 1.0311866696689935,
      "grad_norm": 0.33967259526252747,
      "learning_rate": 0.001,
      "loss": 2.6593,
      "step": 18318
    },
    {
      "epoch": 1.033607295654132,
      "grad_norm": 0.32512950897216797,
      "learning_rate": 0.001,
      "loss": 2.661,
      "step": 18361
    },
    {
      "epoch": 1.0360279216392705,
      "grad_norm": 0.2902744710445404,
      "learning_rate": 0.001,
      "loss": 2.6492,
      "step": 18404
    },
    {
      "epoch": 1.038448547624409,
      "grad_norm": 0.3198734223842621,
      "learning_rate": 0.001,
      "loss": 2.6651,
      "step": 18447
    },
    {
      "epoch": 1.0408691736095474,
      "grad_norm": 0.3395557403564453,
      "learning_rate": 0.001,
      "loss": 2.6648,
      "step": 18490
    },
    {
      "epoch": 1.043289799594686,
      "grad_norm": 0.3080294132232666,
      "learning_rate": 0.001,
      "loss": 2.6556,
      "step": 18533
    },
    {
      "epoch": 1.0457104255798244,
      "grad_norm": 0.3154349625110626,
      "learning_rate": 0.001,
      "loss": 2.6629,
      "step": 18576
    },
    {
      "epoch": 1.048131051564963,
      "grad_norm": 0.2601991593837738,
      "learning_rate": 0.001,
      "loss": 2.6657,
      "step": 18619
    },
    {
      "epoch": 1.0505516775501014,
      "grad_norm": 0.30697888135910034,
      "learning_rate": 0.001,
      "loss": 2.6502,
      "step": 18662
    },
    {
      "epoch": 1.0529723035352399,
      "grad_norm": 0.31738609075546265,
      "learning_rate": 0.001,
      "loss": 2.6552,
      "step": 18705
    },
    {
      "epoch": 1.0553929295203783,
      "grad_norm": 0.3180924355983734,
      "learning_rate": 0.001,
      "loss": 2.6513,
      "step": 18748
    },
    {
      "epoch": 1.0578135555055168,
      "grad_norm": 0.2951227128505707,
      "learning_rate": 0.001,
      "loss": 2.6633,
      "step": 18791
    },
    {
      "epoch": 1.0602341814906553,
      "grad_norm": 0.30238065123558044,
      "learning_rate": 0.001,
      "loss": 2.6578,
      "step": 18834
    },
    {
      "epoch": 1.0626548074757938,
      "grad_norm": 0.3746061623096466,
      "learning_rate": 0.001,
      "loss": 2.6543,
      "step": 18877
    },
    {
      "epoch": 1.0650754334609323,
      "grad_norm": 0.31387439370155334,
      "learning_rate": 0.001,
      "loss": 2.6494,
      "step": 18920
    },
    {
      "epoch": 1.0674960594460707,
      "grad_norm": 0.3482374846935272,
      "learning_rate": 0.001,
      "loss": 2.6628,
      "step": 18963
    },
    {
      "epoch": 1.0699166854312092,
      "grad_norm": 0.31455758213996887,
      "learning_rate": 0.001,
      "loss": 2.6616,
      "step": 19006
    },
    {
      "epoch": 1.0723373114163477,
      "grad_norm": 0.29046404361724854,
      "learning_rate": 0.001,
      "loss": 2.6473,
      "step": 19049
    },
    {
      "epoch": 1.0747579374014862,
      "grad_norm": 0.30390724539756775,
      "learning_rate": 0.001,
      "loss": 2.6572,
      "step": 19092
    },
    {
      "epoch": 1.0771785633866247,
      "grad_norm": 0.38469991087913513,
      "learning_rate": 0.001,
      "loss": 2.6497,
      "step": 19135
    },
    {
      "epoch": 1.0795991893717631,
      "grad_norm": 0.32003918290138245,
      "learning_rate": 0.001,
      "loss": 2.6628,
      "step": 19178
    },
    {
      "epoch": 1.080668768295429,
      "eval_ag_news_bleu_score": 3.7138367760062194,
      "eval_ag_news_bleu_score_sem": 0.1377581704485494,
      "eval_ag_news_emb_cos_sim": 0.9908197522163391,
      "eval_ag_news_emb_cos_sim_sem": 0.0003126263806208464,
      "eval_ag_news_emb_top1_equal": 0.6033333539962769,
      "eval_ag_news_emb_top1_equal_sem": 0.028291497235256893,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.8906426429748535,
      "eval_ag_news_n_ngrams_match_1": 4.478,
      "eval_ag_news_n_ngrams_match_2": 0.75,
      "eval_ag_news_n_ngrams_match_3": 0.184,
      "eval_ag_news_num_pred_words": 22.522,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.23932609128908852,
      "eval_ag_news_runtime": 16.2554,
      "eval_ag_news_samples_per_second": 30.759,
      "eval_ag_news_steps_per_second": 0.123,
      "eval_ag_news_token_set_f1": 0.2099657784661056,
      "eval_ag_news_token_set_f1_sem": 0.004865095513885848,
      "eval_ag_news_token_set_precision": 0.20249672099845734,
      "eval_ag_news_token_set_recall": 0.22693758836077296,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 19197
    },
    {
      "epoch": 1.080668768295429,
      "eval_anthropic_toxic_prompts_bleu_score": 4.697011678653797,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.19698047660238066,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9909192323684692,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002564416658579988,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6033333539962769,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028291498955894778,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.1204025745391846,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.402,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.27,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.5,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.4,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.25695296605186724,
      "eval_anthropic_toxic_prompts_runtime": 6.6215,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.512,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.302,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.28876574558885043,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0062791913441772994,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.34923526709046404,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.27163201995376807,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 19197
    },
    {
      "epoch": 1.080668768295429,
      "eval_arxiv_bleu_score": 3.3270489252057898,
      "eval_arxiv_bleu_score_sem": 0.09857132635679572,
      "eval_arxiv_emb_cos_sim": 0.9902717471122742,
      "eval_arxiv_emb_cos_sim_sem": 0.0002714596352856102,
      "eval_arxiv_emb_top1_equal": 0.5833333134651184,
      "eval_arxiv_emb_top1_equal_sem": 0.02851131216637989,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.895991563796997,
      "eval_arxiv_n_ngrams_match_1": 4.374,
      "eval_arxiv_n_ngrams_match_2": 0.606,
      "eval_arxiv_n_ngrams_match_3": 0.1,
      "eval_arxiv_num_pred_words": 17.544,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.23150839026327208,
      "eval_arxiv_runtime": 6.6255,
      "eval_arxiv_samples_per_second": 75.466,
      "eval_arxiv_steps_per_second": 0.302,
      "eval_arxiv_token_set_f1": 0.22075711835982126,
      "eval_arxiv_token_set_f1_sem": 0.0044715992242239325,
      "eval_arxiv_token_set_precision": 0.18642296724441418,
      "eval_arxiv_token_set_recall": 0.2916570654663536,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 19197
    },
    {
      "epoch": 1.080668768295429,
      "eval_python_code_alpaca_bleu_score": 5.030420780677619,
      "eval_python_code_alpaca_bleu_score_sem": 0.16761539781211907,
      "eval_python_code_alpaca_emb_cos_sim": 0.9881952404975891,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0005385877260047764,
      "eval_python_code_alpaca_emb_top1_equal": 0.6399999856948853,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02775911810844162,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.151674747467041,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.082,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.086,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.344,
      "eval_python_code_alpaca_num_pred_words": 19.264,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2906791635412057,
      "eval_python_code_alpaca_runtime": 6.582,
      "eval_python_code_alpaca_samples_per_second": 75.964,
      "eval_python_code_alpaca_steps_per_second": 0.304,
      "eval_python_code_alpaca_token_set_f1": 0.3029978794188524,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005519003750102323,
      "eval_python_code_alpaca_token_set_precision": 0.29686913323922026,
      "eval_python_code_alpaca_token_set_recall": 0.33629773714343014,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 19197
    },
    {
      "epoch": 1.080668768295429,
      "eval_wikibio_bleu_score": 3.628175903975373,
      "eval_wikibio_bleu_score_sem": 0.12188254015511456,
      "eval_wikibio_emb_cos_sim": 0.9906569719314575,
      "eval_wikibio_emb_cos_sim_sem": 0.0002626492703037272,
      "eval_wikibio_emb_top1_equal": 0.6266666650772095,
      "eval_wikibio_emb_top1_equal_sem": 0.027972489250684164,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.1592512130737305,
      "eval_wikibio_n_ngrams_match_1": 3.044,
      "eval_wikibio_n_ngrams_match_2": 0.792,
      "eval_wikibio_n_ngrams_match_3": 0.162,
      "eval_wikibio_num_pred_words": 20.542,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.24146415964730727,
      "eval_wikibio_runtime": 6.5329,
      "eval_wikibio_samples_per_second": 76.536,
      "eval_wikibio_steps_per_second": 0.306,
      "eval_wikibio_token_set_f1": 0.19458856253628407,
      "eval_wikibio_token_set_f1_sem": 0.0051073519220441775,
      "eval_wikibio_token_set_precision": 0.23115932210563256,
      "eval_wikibio_token_set_recall": 0.17409495921686016,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 19197
    },
    {
      "epoch": 1.080668768295429,
      "eval_nq_5round_bleu_score": 8.826293246536512,
      "eval_nq_5round_bleu_score_sem": 0.4147480571894368,
      "eval_nq_5round_emb_cos_sim": 0.9921815991401672,
      "eval_nq_5round_emb_cos_sim_sem": 0.00026215265119408214,
      "eval_nq_5round_emb_top1_equal": 0.6166666746139526,
      "eval_nq_5round_emb_top1_equal_sem": 0.028117578599086417,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.598,
      "eval_nq_5round_n_ngrams_match_2": 2.79,
      "eval_nq_5round_n_ngrams_match_3": 1.192,
      "eval_nq_5round_num_pred_words": 23.134,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.341920765022958,
      "eval_nq_5round_token_set_f1": 0.3763246488203964,
      "eval_nq_5round_token_set_f1_sem": 0.006205590646496415,
      "eval_nq_5round_token_set_precision": 0.35076148656075923,
      "eval_nq_5round_token_set_recall": 0.4178951539762597,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.955281041186938,
      "eval_nq_bleu_score_sem": 0.42040320079370913,
      "eval_nq_emb_cos_sim": 0.9924865961074829,
      "eval_nq_emb_cos_sim_sem": 0.000248017530311432,
      "eval_nq_emb_top1_equal": 0.5833333134651184,
      "eval_nq_emb_top1_equal_sem": 0.02851131216637989,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.406036853790283,
      "eval_nq_n_ngrams_match_1": 8.946,
      "eval_nq_n_ngrams_match_2": 2.892,
      "eval_nq_n_ngrams_match_3": 1.202,
      "eval_nq_num_pred_words": 23.622,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.35690371074172,
      "eval_nq_runtime": 18.902,
      "eval_nq_samples_per_second": 26.452,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.3911605547651502,
      "eval_nq_token_set_f1_sem": 0.006058391236136246,
      "eval_nq_token_set_precision": 0.3657140587606256,
      "eval_nq_token_set_recall": 0.43169339840065557,
      "eval_nq_true_num_tokens": 32.0,
      "step": 19197
    },
    {
      "epoch": 1.0820198153569016,
      "grad_norm": 0.29094913601875305,
      "learning_rate": 0.001,
      "loss": 2.6488,
      "step": 19221
    },
    {
      "epoch": 1.08444044134204,
      "grad_norm": 0.32373663783073425,
      "learning_rate": 0.001,
      "loss": 2.6475,
      "step": 19264
    },
    {
      "epoch": 1.0868610673271786,
      "grad_norm": 0.36302849650382996,
      "learning_rate": 0.001,
      "loss": 2.6591,
      "step": 19307
    },
    {
      "epoch": 1.089281693312317,
      "grad_norm": 0.29696130752563477,
      "learning_rate": 0.001,
      "loss": 2.6592,
      "step": 19350
    },
    {
      "epoch": 1.0917023192974555,
      "grad_norm": 0.3229084312915802,
      "learning_rate": 0.001,
      "loss": 2.6445,
      "step": 19393
    },
    {
      "epoch": 1.094122945282594,
      "grad_norm": 0.3338836431503296,
      "learning_rate": 0.001,
      "loss": 2.6544,
      "step": 19436
    },
    {
      "epoch": 1.0965435712677325,
      "grad_norm": 0.38707613945007324,
      "learning_rate": 0.001,
      "loss": 2.6585,
      "step": 19479
    },
    {
      "epoch": 1.098964197252871,
      "grad_norm": 0.4195190370082855,
      "learning_rate": 0.001,
      "loss": 2.6497,
      "step": 19522
    },
    {
      "epoch": 1.1013848232380095,
      "grad_norm": 0.35069480538368225,
      "learning_rate": 0.001,
      "loss": 2.6488,
      "step": 19565
    },
    {
      "epoch": 1.103805449223148,
      "grad_norm": 0.3577353060245514,
      "learning_rate": 0.001,
      "loss": 2.6559,
      "step": 19608
    },
    {
      "epoch": 1.1062260752082864,
      "grad_norm": 0.290390282869339,
      "learning_rate": 0.001,
      "loss": 2.6594,
      "step": 19651
    },
    {
      "epoch": 1.108646701193425,
      "grad_norm": 0.3018650710582733,
      "learning_rate": 0.001,
      "loss": 2.6508,
      "step": 19694
    },
    {
      "epoch": 1.1110673271785634,
      "grad_norm": 0.26733043789863586,
      "learning_rate": 0.001,
      "loss": 2.6465,
      "step": 19737
    },
    {
      "epoch": 1.1134879531637019,
      "grad_norm": 0.36778441071510315,
      "learning_rate": 0.001,
      "loss": 2.653,
      "step": 19780
    },
    {
      "epoch": 1.1159085791488403,
      "grad_norm": 0.35062047839164734,
      "learning_rate": 0.001,
      "loss": 2.6407,
      "step": 19823
    },
    {
      "epoch": 1.1183292051339788,
      "grad_norm": 0.31850960850715637,
      "learning_rate": 0.001,
      "loss": 2.6448,
      "step": 19866
    },
    {
      "epoch": 1.1207498311191173,
      "grad_norm": 0.3169715702533722,
      "learning_rate": 0.001,
      "loss": 2.6501,
      "step": 19909
    },
    {
      "epoch": 1.1231704571042558,
      "grad_norm": 0.3112129867076874,
      "learning_rate": 0.001,
      "loss": 2.6541,
      "step": 19952
    },
    {
      "epoch": 1.1255910830893943,
      "grad_norm": 0.3728781044483185,
      "learning_rate": 0.001,
      "loss": 2.6549,
      "step": 19995
    },
    {
      "epoch": 1.1280117090745327,
      "grad_norm": 0.2775253355503082,
      "learning_rate": 0.001,
      "loss": 2.6485,
      "step": 20038
    },
    {
      "epoch": 1.1304323350596712,
      "grad_norm": 0.3383483588695526,
      "learning_rate": 0.001,
      "loss": 2.6562,
      "step": 20081
    },
    {
      "epoch": 1.1328529610448097,
      "grad_norm": 0.37347638607025146,
      "learning_rate": 0.001,
      "loss": 2.6461,
      "step": 20124
    },
    {
      "epoch": 1.1352735870299482,
      "grad_norm": 0.3401934504508972,
      "learning_rate": 0.001,
      "loss": 2.6422,
      "step": 20167
    },
    {
      "epoch": 1.1376942130150867,
      "grad_norm": 0.3246845006942749,
      "learning_rate": 0.001,
      "loss": 2.6506,
      "step": 20210
    },
    {
      "epoch": 1.1401148390002251,
      "grad_norm": 0.31002897024154663,
      "learning_rate": 0.001,
      "loss": 2.6466,
      "step": 20253
    },
    {
      "epoch": 1.1425354649853636,
      "grad_norm": 0.3258046805858612,
      "learning_rate": 0.001,
      "loss": 2.6418,
      "step": 20296
    },
    {
      "epoch": 1.144956090970502,
      "grad_norm": 0.2926499843597412,
      "learning_rate": 0.001,
      "loss": 2.6433,
      "step": 20339
    },
    {
      "epoch": 1.1473767169556406,
      "grad_norm": 0.28643637895584106,
      "learning_rate": 0.001,
      "loss": 2.6425,
      "step": 20382
    },
    {
      "epoch": 1.149797342940779,
      "grad_norm": 0.28597745299339294,
      "learning_rate": 0.001,
      "loss": 2.6421,
      "step": 20425
    },
    {
      "epoch": 1.1522179689259175,
      "grad_norm": 0.3419337868690491,
      "learning_rate": 0.001,
      "loss": 2.6396,
      "step": 20468
    },
    {
      "epoch": 1.154638594911056,
      "grad_norm": 0.4002171754837036,
      "learning_rate": 0.001,
      "loss": 2.6377,
      "step": 20511
    },
    {
      "epoch": 1.1570592208961945,
      "grad_norm": 0.31176191568374634,
      "learning_rate": 0.001,
      "loss": 2.6366,
      "step": 20554
    },
    {
      "epoch": 1.159479846881333,
      "grad_norm": 0.3887931704521179,
      "learning_rate": 0.001,
      "loss": 2.6604,
      "step": 20597
    },
    {
      "epoch": 1.1619004728664715,
      "grad_norm": 0.35901227593421936,
      "learning_rate": 0.001,
      "loss": 2.6486,
      "step": 20640
    },
    {
      "epoch": 1.16432109885161,
      "grad_norm": 0.30541568994522095,
      "learning_rate": 0.001,
      "loss": 2.6543,
      "step": 20683
    },
    {
      "epoch": 1.1667417248367484,
      "grad_norm": 0.3609279692173004,
      "learning_rate": 0.001,
      "loss": 2.6618,
      "step": 20726
    },
    {
      "epoch": 1.169162350821887,
      "grad_norm": 0.4321398437023163,
      "learning_rate": 0.001,
      "loss": 2.6408,
      "step": 20769
    },
    {
      "epoch": 1.1715829768070254,
      "grad_norm": 0.37637022137641907,
      "learning_rate": 0.001,
      "loss": 2.6453,
      "step": 20812
    },
    {
      "epoch": 1.1740036027921639,
      "grad_norm": 0.3674578070640564,
      "learning_rate": 0.001,
      "loss": 2.6383,
      "step": 20855
    },
    {
      "epoch": 1.1764242287773023,
      "grad_norm": 0.36703869700431824,
      "learning_rate": 0.001,
      "loss": 2.6337,
      "step": 20898
    },
    {
      "epoch": 1.1788448547624408,
      "grad_norm": 0.29859933257102966,
      "learning_rate": 0.001,
      "loss": 2.6276,
      "step": 20941
    },
    {
      "epoch": 1.1812654807475793,
      "grad_norm": 0.3595154881477356,
      "learning_rate": 0.001,
      "loss": 2.6358,
      "step": 20984
    },
    {
      "epoch": 1.1836861067327178,
      "grad_norm": 0.31804612278938293,
      "learning_rate": 0.001,
      "loss": 2.6374,
      "step": 21027
    },
    {
      "epoch": 1.1861067327178563,
      "grad_norm": 0.31649211049079895,
      "learning_rate": 0.001,
      "loss": 2.635,
      "step": 21070
    },
    {
      "epoch": 1.1885273587029948,
      "grad_norm": 0.28527289628982544,
      "learning_rate": 0.001,
      "loss": 2.639,
      "step": 21113
    },
    {
      "epoch": 1.1909479846881332,
      "grad_norm": 0.33192378282546997,
      "learning_rate": 0.001,
      "loss": 2.6285,
      "step": 21156
    },
    {
      "epoch": 1.1933686106732717,
      "grad_norm": 0.32406723499298096,
      "learning_rate": 0.001,
      "loss": 2.6418,
      "step": 21199
    },
    {
      "epoch": 1.1957892366584102,
      "grad_norm": 0.38631027936935425,
      "learning_rate": 0.001,
      "loss": 2.6329,
      "step": 21242
    },
    {
      "epoch": 1.1982098626435487,
      "grad_norm": 0.33428463339805603,
      "learning_rate": 0.001,
      "loss": 2.6336,
      "step": 21285
    },
    {
      "epoch": 1.2006304886286872,
      "grad_norm": 0.38925930857658386,
      "learning_rate": 0.001,
      "loss": 2.6437,
      "step": 21328
    },
    {
      "epoch": 1.20074307588381,
      "eval_ag_news_bleu_score": 3.7319367325875827,
      "eval_ag_news_bleu_score_sem": 0.14291406634348947,
      "eval_ag_news_emb_cos_sim": 0.9904884099960327,
      "eval_ag_news_emb_cos_sim_sem": 0.0003133538340567914,
      "eval_ag_news_emb_top1_equal": 0.6433333158493042,
      "eval_ag_news_emb_top1_equal_sem": 0.02770216499443815,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.8489251136779785,
      "eval_ag_news_n_ngrams_match_1": 4.52,
      "eval_ag_news_n_ngrams_match_2": 0.78,
      "eval_ag_news_n_ngrams_match_3": 0.196,
      "eval_ag_news_num_pred_words": 22.672,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.24074323424301264,
      "eval_ag_news_runtime": 15.3127,
      "eval_ag_news_samples_per_second": 32.653,
      "eval_ag_news_steps_per_second": 0.131,
      "eval_ag_news_token_set_f1": 0.21304168572253335,
      "eval_ag_news_token_set_f1_sem": 0.005010574482013604,
      "eval_ag_news_token_set_precision": 0.20462650622250717,
      "eval_ag_news_token_set_recall": 0.23086240818720255,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 21330
    },
    {
      "epoch": 1.20074307588381,
      "eval_anthropic_toxic_prompts_bleu_score": 4.632366830554466,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.19614218269253608,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9907289743423462,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00029418073921591326,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.653333306312561,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027522499427730773,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.085540771484375,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.45,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.27,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.488,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.856,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.25765096899234335,
      "eval_anthropic_toxic_prompts_runtime": 6.7143,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.467,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.298,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.2880973834754475,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0062473552290272846,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.35173175555371633,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2680321499886523,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 21330
    },
    {
      "epoch": 1.20074307588381,
      "eval_arxiv_bleu_score": 3.4170621367427683,
      "eval_arxiv_bleu_score_sem": 0.11072221032334852,
      "eval_arxiv_emb_cos_sim": 0.9899513721466064,
      "eval_arxiv_emb_cos_sim_sem": 0.0002751895211765049,
      "eval_arxiv_emb_top1_equal": 0.5366666913032532,
      "eval_arxiv_emb_top1_equal_sem": 0.02883789095764794,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.8737685680389404,
      "eval_arxiv_n_ngrams_match_1": 4.378,
      "eval_arxiv_n_ngrams_match_2": 0.64,
      "eval_arxiv_n_ngrams_match_3": 0.106,
      "eval_arxiv_num_pred_words": 17.532,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.23318828666569985,
      "eval_arxiv_runtime": 6.6683,
      "eval_arxiv_samples_per_second": 74.982,
      "eval_arxiv_steps_per_second": 0.3,
      "eval_arxiv_token_set_f1": 0.22010157566219243,
      "eval_arxiv_token_set_f1_sem": 0.004620705187468491,
      "eval_arxiv_token_set_precision": 0.18385129291855543,
      "eval_arxiv_token_set_recall": 0.29958383432424407,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 21330
    },
    {
      "epoch": 1.20074307588381,
      "eval_python_code_alpaca_bleu_score": 4.923440795367529,
      "eval_python_code_alpaca_bleu_score_sem": 0.1720311318012265,
      "eval_python_code_alpaca_emb_cos_sim": 0.9877498745918274,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00047855187415005423,
      "eval_python_code_alpaca_emb_top1_equal": 0.6333333253860474,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02786867456387452,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.1467065811157227,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.052,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.102,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.318,
      "eval_python_code_alpaca_num_pred_words": 18.798,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.28732301035537233,
      "eval_python_code_alpaca_runtime": 6.7206,
      "eval_python_code_alpaca_samples_per_second": 74.398,
      "eval_python_code_alpaca_steps_per_second": 0.298,
      "eval_python_code_alpaca_token_set_f1": 0.3038303612878576,
      "eval_python_code_alpaca_token_set_f1_sem": 0.00587058971522453,
      "eval_python_code_alpaca_token_set_precision": 0.29769056935526467,
      "eval_python_code_alpaca_token_set_recall": 0.3452775061423973,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 21330
    },
    {
      "epoch": 1.20074307588381,
      "eval_wikibio_bleu_score": 3.512340185315318,
      "eval_wikibio_bleu_score_sem": 0.12666669884975978,
      "eval_wikibio_emb_cos_sim": 0.9907965064048767,
      "eval_wikibio_emb_cos_sim_sem": 0.0002755038333252089,
      "eval_wikibio_emb_top1_equal": 0.6466666460037231,
      "eval_wikibio_emb_top1_equal_sem": 0.027643749338232177,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.032404899597168,
      "eval_wikibio_n_ngrams_match_1": 3.004,
      "eval_wikibio_n_ngrams_match_2": 0.75,
      "eval_wikibio_n_ngrams_match_3": 0.162,
      "eval_wikibio_num_pred_words": 20.53,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.23731897642981498,
      "eval_wikibio_runtime": 6.865,
      "eval_wikibio_samples_per_second": 72.833,
      "eval_wikibio_steps_per_second": 0.291,
      "eval_wikibio_token_set_f1": 0.1917303307741838,
      "eval_wikibio_token_set_f1_sem": 0.0052267869508624745,
      "eval_wikibio_token_set_precision": 0.22766890065401918,
      "eval_wikibio_token_set_recall": 0.17159304902886457,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 21330
    },
    {
      "epoch": 1.20074307588381,
      "eval_nq_5round_bleu_score": 8.466421718339248,
      "eval_nq_5round_bleu_score_sem": 0.4158350746763969,
      "eval_nq_5round_emb_cos_sim": 0.9922561049461365,
      "eval_nq_5round_emb_cos_sim_sem": 0.00026972851226319546,
      "eval_nq_5round_emb_top1_equal": 0.6133333444595337,
      "eval_nq_5round_emb_top1_equal_sem": 0.028163139369651306,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.768,
      "eval_nq_5round_n_ngrams_match_2": 2.77,
      "eval_nq_5round_n_ngrams_match_3": 1.1,
      "eval_nq_5round_num_pred_words": 23.432,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.34764327725761585,
      "eval_nq_5round_token_set_f1": 0.3772491219571878,
      "eval_nq_5round_token_set_f1_sem": 0.0062284208534278745,
      "eval_nq_5round_token_set_precision": 0.35164316871312845,
      "eval_nq_5round_token_set_recall": 0.4201878821041113,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 8.784290105101041,
      "eval_nq_bleu_score_sem": 0.40217780671463615,
      "eval_nq_emb_cos_sim": 0.9928425550460815,
      "eval_nq_emb_cos_sim_sem": 0.00024077391366398356,
      "eval_nq_emb_top1_equal": 0.6366666555404663,
      "eval_nq_emb_top1_equal_sem": 0.027814619004069896,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.3742711544036865,
      "eval_nq_n_ngrams_match_1": 9.094,
      "eval_nq_n_ngrams_match_2": 2.918,
      "eval_nq_n_ngrams_match_3": 1.156,
      "eval_nq_num_pred_words": 23.642,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.36226405945242823,
      "eval_nq_runtime": 26.9171,
      "eval_nq_samples_per_second": 18.576,
      "eval_nq_steps_per_second": 0.074,
      "eval_nq_token_set_f1": 0.39547337194826804,
      "eval_nq_token_set_f1_sem": 0.0061076247053831806,
      "eval_nq_token_set_precision": 0.37000082552982255,
      "eval_nq_token_set_recall": 0.43648599765210894,
      "eval_nq_true_num_tokens": 32.0,
      "step": 21330
    },
    {
      "epoch": 1.2030511146138256,
      "grad_norm": 0.3211119472980499,
      "learning_rate": 0.001,
      "loss": 2.6396,
      "step": 21371
    },
    {
      "epoch": 1.2054717405989641,
      "grad_norm": 0.369575560092926,
      "learning_rate": 0.001,
      "loss": 2.6335,
      "step": 21414
    },
    {
      "epoch": 1.2078923665841026,
      "grad_norm": 0.33007320761680603,
      "learning_rate": 0.001,
      "loss": 2.6293,
      "step": 21457
    },
    {
      "epoch": 1.210312992569241,
      "grad_norm": 0.29791882634162903,
      "learning_rate": 0.001,
      "loss": 2.6201,
      "step": 21500
    },
    {
      "epoch": 1.2127336185543796,
      "grad_norm": 0.3866545855998993,
      "learning_rate": 0.001,
      "loss": 2.6356,
      "step": 21543
    },
    {
      "epoch": 1.215154244539518,
      "grad_norm": 0.3258555829524994,
      "learning_rate": 0.001,
      "loss": 2.634,
      "step": 21586
    },
    {
      "epoch": 1.2175748705246565,
      "grad_norm": 0.3393552303314209,
      "learning_rate": 0.001,
      "loss": 2.6363,
      "step": 21629
    },
    {
      "epoch": 1.219995496509795,
      "grad_norm": 0.2952291965484619,
      "learning_rate": 0.001,
      "loss": 2.6261,
      "step": 21672
    },
    {
      "epoch": 1.2224161224949335,
      "grad_norm": 0.2862803041934967,
      "learning_rate": 0.001,
      "loss": 2.6358,
      "step": 21715
    },
    {
      "epoch": 1.224836748480072,
      "grad_norm": 0.3590514361858368,
      "learning_rate": 0.001,
      "loss": 2.6182,
      "step": 21758
    },
    {
      "epoch": 1.2272573744652107,
      "grad_norm": 0.28851959109306335,
      "learning_rate": 0.001,
      "loss": 2.6339,
      "step": 21801
    },
    {
      "epoch": 1.229678000450349,
      "grad_norm": 0.33343735337257385,
      "learning_rate": 0.001,
      "loss": 2.628,
      "step": 21844
    },
    {
      "epoch": 1.2320986264354876,
      "grad_norm": 0.3040972650051117,
      "learning_rate": 0.001,
      "loss": 2.6311,
      "step": 21887
    },
    {
      "epoch": 1.2345192524206259,
      "grad_norm": 0.41589412093162537,
      "learning_rate": 0.001,
      "loss": 2.6165,
      "step": 21930
    },
    {
      "epoch": 1.2369398784057646,
      "grad_norm": 0.3318420648574829,
      "learning_rate": 0.001,
      "loss": 2.6172,
      "step": 21973
    },
    {
      "epoch": 1.2393605043909028,
      "grad_norm": 0.3151552975177765,
      "learning_rate": 0.001,
      "loss": 2.6175,
      "step": 22016
    },
    {
      "epoch": 1.2417811303760415,
      "grad_norm": 0.30436044931411743,
      "learning_rate": 0.001,
      "loss": 2.6167,
      "step": 22059
    },
    {
      "epoch": 1.2442017563611798,
      "grad_norm": 0.36614084243774414,
      "learning_rate": 0.001,
      "loss": 2.633,
      "step": 22102
    },
    {
      "epoch": 1.2466223823463185,
      "grad_norm": 0.31893444061279297,
      "learning_rate": 0.001,
      "loss": 2.625,
      "step": 22145
    },
    {
      "epoch": 1.2490430083314568,
      "grad_norm": 0.44134044647216797,
      "learning_rate": 0.001,
      "loss": 2.6241,
      "step": 22188
    },
    {
      "epoch": 1.2514636343165955,
      "grad_norm": 0.3080842196941376,
      "learning_rate": 0.001,
      "loss": 2.6245,
      "step": 22231
    },
    {
      "epoch": 1.2538842603017337,
      "grad_norm": 0.36590149998664856,
      "learning_rate": 0.001,
      "loss": 2.6209,
      "step": 22274
    },
    {
      "epoch": 1.2563048862868724,
      "grad_norm": 0.3123462200164795,
      "learning_rate": 0.001,
      "loss": 2.6197,
      "step": 22317
    },
    {
      "epoch": 1.2587255122720107,
      "grad_norm": 0.34022635221481323,
      "learning_rate": 0.001,
      "loss": 2.6143,
      "step": 22360
    },
    {
      "epoch": 1.2611461382571494,
      "grad_norm": 0.3313036859035492,
      "learning_rate": 0.001,
      "loss": 2.6211,
      "step": 22403
    },
    {
      "epoch": 1.2635667642422876,
      "grad_norm": 0.3097012937068939,
      "learning_rate": 0.001,
      "loss": 2.6207,
      "step": 22446
    },
    {
      "epoch": 1.2659873902274263,
      "grad_norm": 0.34989088773727417,
      "learning_rate": 0.001,
      "loss": 2.6229,
      "step": 22489
    },
    {
      "epoch": 1.2684080162125646,
      "grad_norm": 0.3552291989326477,
      "learning_rate": 0.001,
      "loss": 2.6147,
      "step": 22532
    },
    {
      "epoch": 1.2708286421977033,
      "grad_norm": 0.326415091753006,
      "learning_rate": 0.001,
      "loss": 2.613,
      "step": 22575
    },
    {
      "epoch": 1.2732492681828416,
      "grad_norm": 0.3379911184310913,
      "learning_rate": 0.001,
      "loss": 2.617,
      "step": 22618
    },
    {
      "epoch": 1.2756698941679803,
      "grad_norm": 0.3357168436050415,
      "learning_rate": 0.001,
      "loss": 2.6157,
      "step": 22661
    },
    {
      "epoch": 1.2780905201531187,
      "grad_norm": 0.3639044761657715,
      "learning_rate": 0.001,
      "loss": 2.6263,
      "step": 22704
    },
    {
      "epoch": 1.2805111461382572,
      "grad_norm": 0.2796708345413208,
      "learning_rate": 0.001,
      "loss": 2.6106,
      "step": 22747
    },
    {
      "epoch": 1.2829317721233957,
      "grad_norm": 0.3242395520210266,
      "learning_rate": 0.001,
      "loss": 2.6184,
      "step": 22790
    },
    {
      "epoch": 1.2853523981085342,
      "grad_norm": 0.28960007429122925,
      "learning_rate": 0.001,
      "loss": 2.6241,
      "step": 22833
    },
    {
      "epoch": 1.2877730240936727,
      "grad_norm": 0.3076488673686981,
      "learning_rate": 0.001,
      "loss": 2.6118,
      "step": 22876
    },
    {
      "epoch": 1.2901936500788111,
      "grad_norm": 0.32568439841270447,
      "learning_rate": 0.001,
      "loss": 2.6195,
      "step": 22919
    },
    {
      "epoch": 1.2926142760639496,
      "grad_norm": 0.37673765420913696,
      "learning_rate": 0.001,
      "loss": 2.6189,
      "step": 22962
    },
    {
      "epoch": 1.295034902049088,
      "grad_norm": 0.33716845512390137,
      "learning_rate": 0.001,
      "loss": 2.6154,
      "step": 23005
    },
    {
      "epoch": 1.2974555280342266,
      "grad_norm": 0.35505011677742004,
      "learning_rate": 0.001,
      "loss": 2.6185,
      "step": 23048
    },
    {
      "epoch": 1.299876154019365,
      "grad_norm": 0.353729248046875,
      "learning_rate": 0.001,
      "loss": 2.6164,
      "step": 23091
    },
    {
      "epoch": 1.3022967800045036,
      "grad_norm": 0.33365678787231445,
      "learning_rate": 0.001,
      "loss": 2.6135,
      "step": 23134
    },
    {
      "epoch": 1.304717405989642,
      "grad_norm": 0.3910866379737854,
      "learning_rate": 0.001,
      "loss": 2.6073,
      "step": 23177
    },
    {
      "epoch": 1.3071380319747805,
      "grad_norm": 0.31149163842201233,
      "learning_rate": 0.001,
      "loss": 2.6195,
      "step": 23220
    },
    {
      "epoch": 1.309558657959919,
      "grad_norm": 0.31374698877334595,
      "learning_rate": 0.001,
      "loss": 2.6121,
      "step": 23263
    },
    {
      "epoch": 1.3119792839450575,
      "grad_norm": 0.3365541100502014,
      "learning_rate": 0.001,
      "loss": 2.6075,
      "step": 23306
    },
    {
      "epoch": 1.314399909930196,
      "grad_norm": 0.3457179069519043,
      "learning_rate": 0.001,
      "loss": 2.6043,
      "step": 23349
    },
    {
      "epoch": 1.3168205359153344,
      "grad_norm": 0.3998713493347168,
      "learning_rate": 0.001,
      "loss": 2.6081,
      "step": 23392
    },
    {
      "epoch": 1.319241161900473,
      "grad_norm": 0.37885963916778564,
      "learning_rate": 0.001,
      "loss": 2.6188,
      "step": 23435
    },
    {
      "epoch": 1.320817383472191,
      "eval_ag_news_bleu_score": 3.803040709856379,
      "eval_ag_news_bleu_score_sem": 0.13759386499696735,
      "eval_ag_news_emb_cos_sim": 0.990809977054596,
      "eval_ag_news_emb_cos_sim_sem": 0.0002997771139378554,
      "eval_ag_news_emb_top1_equal": 0.6100000143051147,
      "eval_ag_news_emb_top1_equal_sem": 0.028207306423529106,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.81477689743042,
      "eval_ag_news_n_ngrams_match_1": 4.69,
      "eval_ag_news_n_ngrams_match_2": 0.816,
      "eval_ag_news_n_ngrams_match_3": 0.194,
      "eval_ag_news_num_pred_words": 22.608,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.248182127786222,
      "eval_ag_news_runtime": 15.9622,
      "eval_ag_news_samples_per_second": 31.324,
      "eval_ag_news_steps_per_second": 0.125,
      "eval_ag_news_token_set_f1": 0.21948514729450952,
      "eval_ag_news_token_set_f1_sem": 0.005177734702889826,
      "eval_ag_news_token_set_precision": 0.21194718156718909,
      "eval_ag_news_token_set_recall": 0.23496008605902818,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 23463
    },
    {
      "epoch": 1.320817383472191,
      "eval_anthropic_toxic_prompts_bleu_score": 4.738827453288561,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.20159278347026113,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9910520911216736,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00027858509250327894,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6200000047683716,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028070624111834433,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.0710628032684326,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.464,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.27,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.484,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.46,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.261107147314428,
      "eval_anthropic_toxic_prompts_runtime": 6.759,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.975,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.296,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.29249983559718656,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006349338296623987,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3554328088134604,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.27402872108918896,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 23463
    },
    {
      "epoch": 1.320817383472191,
      "eval_arxiv_bleu_score": 3.414876850216779,
      "eval_arxiv_bleu_score_sem": 0.09880121885353299,
      "eval_arxiv_emb_cos_sim": 0.9902545213699341,
      "eval_arxiv_emb_cos_sim_sem": 0.0003368872672637253,
      "eval_arxiv_emb_top1_equal": 0.5933333039283752,
      "eval_arxiv_emb_top1_equal_sem": 0.028407504362121784,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.854429244995117,
      "eval_arxiv_n_ngrams_match_1": 4.444,
      "eval_arxiv_n_ngrams_match_2": 0.638,
      "eval_arxiv_n_ngrams_match_3": 0.102,
      "eval_arxiv_num_pred_words": 17.652,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.23531158739322983,
      "eval_arxiv_runtime": 6.7566,
      "eval_arxiv_samples_per_second": 74.002,
      "eval_arxiv_steps_per_second": 0.296,
      "eval_arxiv_token_set_f1": 0.22152189852434356,
      "eval_arxiv_token_set_f1_sem": 0.004545185170775502,
      "eval_arxiv_token_set_precision": 0.1895613651814604,
      "eval_arxiv_token_set_recall": 0.2854632770894924,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 23463
    },
    {
      "epoch": 1.320817383472191,
      "eval_python_code_alpaca_bleu_score": 5.283400240087324,
      "eval_python_code_alpaca_bleu_score_sem": 0.1818924388376663,
      "eval_python_code_alpaca_emb_cos_sim": 0.988405704498291,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00038845787645740214,
      "eval_python_code_alpaca_emb_top1_equal": 0.596666693687439,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028370195770854797,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.1763088703155518,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.154,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.158,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.368,
      "eval_python_code_alpaca_num_pred_words": 18.924,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.29260059260968185,
      "eval_python_code_alpaca_runtime": 6.6684,
      "eval_python_code_alpaca_samples_per_second": 74.98,
      "eval_python_code_alpaca_steps_per_second": 0.3,
      "eval_python_code_alpaca_token_set_f1": 0.30718357457399564,
      "eval_python_code_alpaca_token_set_f1_sem": 0.00570235096610177,
      "eval_python_code_alpaca_token_set_precision": 0.30151892124407426,
      "eval_python_code_alpaca_token_set_recall": 0.3442218252354584,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 23463
    },
    {
      "epoch": 1.320817383472191,
      "eval_wikibio_bleu_score": 3.6107757291326807,
      "eval_wikibio_bleu_score_sem": 0.11733941964490828,
      "eval_wikibio_emb_cos_sim": 0.9908616542816162,
      "eval_wikibio_emb_cos_sim_sem": 0.00029067386412593376,
      "eval_wikibio_emb_top1_equal": 0.6866666674613953,
      "eval_wikibio_emb_top1_equal_sem": 0.02682505950857856,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.058777332305908,
      "eval_wikibio_n_ngrams_match_1": 3.072,
      "eval_wikibio_n_ngrams_match_2": 0.764,
      "eval_wikibio_n_ngrams_match_3": 0.146,
      "eval_wikibio_num_pred_words": 20.308,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.24621105098204954,
      "eval_wikibio_runtime": 6.6297,
      "eval_wikibio_samples_per_second": 75.419,
      "eval_wikibio_steps_per_second": 0.302,
      "eval_wikibio_token_set_f1": 0.19783500447570165,
      "eval_wikibio_token_set_f1_sem": 0.005172931346891216,
      "eval_wikibio_token_set_precision": 0.2334120751940661,
      "eval_wikibio_token_set_recall": 0.17747771515490424,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 23463
    },
    {
      "epoch": 1.320817383472191,
      "eval_nq_5round_bleu_score": 8.777594507688082,
      "eval_nq_5round_bleu_score_sem": 0.43381199501527035,
      "eval_nq_5round_emb_cos_sim": 0.9927240014076233,
      "eval_nq_5round_emb_cos_sim_sem": 0.0002464858668588774,
      "eval_nq_5round_emb_top1_equal": 0.6600000262260437,
      "eval_nq_5round_emb_top1_equal_sem": 0.027395285786318915,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.7,
      "eval_nq_5round_n_ngrams_match_2": 2.78,
      "eval_nq_5round_n_ngrams_match_3": 1.156,
      "eval_nq_5round_num_pred_words": 23.082,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3484632405806716,
      "eval_nq_5round_token_set_f1": 0.38107089267111843,
      "eval_nq_5round_token_set_f1_sem": 0.006249994868147882,
      "eval_nq_5round_token_set_precision": 0.35587101378654645,
      "eval_nq_5round_token_set_recall": 0.4204488845063839,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.367897721587024,
      "eval_nq_bleu_score_sem": 0.42760510135005503,
      "eval_nq_emb_cos_sim": 0.9925958514213562,
      "eval_nq_emb_cos_sim_sem": 0.00024377833560648813,
      "eval_nq_emb_top1_equal": 0.6466666460037231,
      "eval_nq_emb_top1_equal_sem": 0.027643749338232177,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.3510849475860596,
      "eval_nq_n_ngrams_match_1": 9.16,
      "eval_nq_n_ngrams_match_2": 3.024,
      "eval_nq_n_ngrams_match_3": 1.284,
      "eval_nq_num_pred_words": 23.436,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.36620002741124047,
      "eval_nq_runtime": 18.961,
      "eval_nq_samples_per_second": 26.37,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.3982147545867761,
      "eval_nq_token_set_f1_sem": 0.006213069813491642,
      "eval_nq_token_set_precision": 0.37275793768615173,
      "eval_nq_token_set_recall": 0.43971417291098946,
      "eval_nq_true_num_tokens": 32.0,
      "step": 23463
    },
    {
      "epoch": 1.3216617878856114,
      "grad_norm": 0.3131730258464813,
      "learning_rate": 0.001,
      "loss": 2.5995,
      "step": 23478
    },
    {
      "epoch": 1.3240824138707499,
      "grad_norm": 0.3895186483860016,
      "learning_rate": 0.001,
      "loss": 2.6091,
      "step": 23521
    },
    {
      "epoch": 1.3265030398558884,
      "grad_norm": 0.32967647910118103,
      "learning_rate": 0.001,
      "loss": 2.6108,
      "step": 23564
    },
    {
      "epoch": 1.3289236658410268,
      "grad_norm": 0.40792033076286316,
      "learning_rate": 0.001,
      "loss": 2.611,
      "step": 23607
    },
    {
      "epoch": 1.3313442918261653,
      "grad_norm": 0.3644319474697113,
      "learning_rate": 0.001,
      "loss": 2.6045,
      "step": 23650
    },
    {
      "epoch": 1.3337649178113038,
      "grad_norm": 0.3093305826187134,
      "learning_rate": 0.001,
      "loss": 2.6104,
      "step": 23693
    },
    {
      "epoch": 1.3361855437964423,
      "grad_norm": 0.3653559684753418,
      "learning_rate": 0.001,
      "loss": 2.6032,
      "step": 23736
    },
    {
      "epoch": 1.3386061697815808,
      "grad_norm": 0.2907729148864746,
      "learning_rate": 0.001,
      "loss": 2.6017,
      "step": 23779
    },
    {
      "epoch": 1.3410267957667192,
      "grad_norm": 0.3605106770992279,
      "learning_rate": 0.001,
      "loss": 2.5939,
      "step": 23822
    },
    {
      "epoch": 1.3434474217518577,
      "grad_norm": 0.34191691875457764,
      "learning_rate": 0.001,
      "loss": 2.6084,
      "step": 23865
    },
    {
      "epoch": 1.3458680477369962,
      "grad_norm": 0.39918696880340576,
      "learning_rate": 0.001,
      "loss": 2.6072,
      "step": 23908
    },
    {
      "epoch": 1.3482886737221347,
      "grad_norm": 0.3399095833301544,
      "learning_rate": 0.001,
      "loss": 2.6057,
      "step": 23951
    },
    {
      "epoch": 1.3507092997072732,
      "grad_norm": 0.3300797939300537,
      "learning_rate": 0.001,
      "loss": 2.6111,
      "step": 23994
    },
    {
      "epoch": 1.3531299256924116,
      "grad_norm": 0.34920474886894226,
      "learning_rate": 0.001,
      "loss": 2.6173,
      "step": 24037
    },
    {
      "epoch": 1.3555505516775501,
      "grad_norm": 0.36294296383857727,
      "learning_rate": 0.001,
      "loss": 2.5916,
      "step": 24080
    },
    {
      "epoch": 1.3579711776626886,
      "grad_norm": 0.3643760681152344,
      "learning_rate": 0.001,
      "loss": 2.6051,
      "step": 24123
    },
    {
      "epoch": 1.360391803647827,
      "grad_norm": 0.29400601983070374,
      "learning_rate": 0.001,
      "loss": 2.6017,
      "step": 24166
    },
    {
      "epoch": 1.3628124296329656,
      "grad_norm": 0.3654016852378845,
      "learning_rate": 0.001,
      "loss": 2.6093,
      "step": 24209
    },
    {
      "epoch": 1.365233055618104,
      "grad_norm": 0.3357471525669098,
      "learning_rate": 0.001,
      "loss": 2.6019,
      "step": 24252
    },
    {
      "epoch": 1.3676536816032425,
      "grad_norm": 0.3329485058784485,
      "learning_rate": 0.001,
      "loss": 2.6023,
      "step": 24295
    },
    {
      "epoch": 1.370074307588381,
      "grad_norm": 0.3371709883213043,
      "learning_rate": 0.001,
      "loss": 2.6029,
      "step": 24338
    },
    {
      "epoch": 1.3724949335735195,
      "grad_norm": 0.45280519127845764,
      "learning_rate": 0.001,
      "loss": 2.6006,
      "step": 24381
    },
    {
      "epoch": 1.374915559558658,
      "grad_norm": 0.3184469938278198,
      "learning_rate": 0.001,
      "loss": 2.6002,
      "step": 24424
    },
    {
      "epoch": 1.3773361855437964,
      "grad_norm": 0.33607250452041626,
      "learning_rate": 0.001,
      "loss": 2.5832,
      "step": 24467
    },
    {
      "epoch": 1.379756811528935,
      "grad_norm": 0.3124711513519287,
      "learning_rate": 0.001,
      "loss": 2.604,
      "step": 24510
    },
    {
      "epoch": 1.3821774375140734,
      "grad_norm": 0.30336201190948486,
      "learning_rate": 0.001,
      "loss": 2.595,
      "step": 24553
    },
    {
      "epoch": 1.3845980634992119,
      "grad_norm": 0.3968513607978821,
      "learning_rate": 0.001,
      "loss": 2.5953,
      "step": 24596
    },
    {
      "epoch": 1.3870186894843504,
      "grad_norm": 0.30765408277511597,
      "learning_rate": 0.001,
      "loss": 2.605,
      "step": 24639
    },
    {
      "epoch": 1.3894393154694888,
      "grad_norm": 0.3227773904800415,
      "learning_rate": 0.001,
      "loss": 2.5994,
      "step": 24682
    },
    {
      "epoch": 1.3918599414546273,
      "grad_norm": 0.2955887019634247,
      "learning_rate": 0.001,
      "loss": 2.5905,
      "step": 24725
    },
    {
      "epoch": 1.3942805674397658,
      "grad_norm": 0.29220566153526306,
      "learning_rate": 0.001,
      "loss": 2.5891,
      "step": 24768
    },
    {
      "epoch": 1.3967011934249043,
      "grad_norm": 0.3334503173828125,
      "learning_rate": 0.001,
      "loss": 2.5993,
      "step": 24811
    },
    {
      "epoch": 1.3991218194100428,
      "grad_norm": 0.2966783046722412,
      "learning_rate": 0.001,
      "loss": 2.6001,
      "step": 24854
    },
    {
      "epoch": 1.4015424453951812,
      "grad_norm": 0.3414221704006195,
      "learning_rate": 0.001,
      "loss": 2.6041,
      "step": 24897
    },
    {
      "epoch": 1.4039630713803197,
      "grad_norm": 0.3064633309841156,
      "learning_rate": 0.001,
      "loss": 2.5869,
      "step": 24940
    },
    {
      "epoch": 1.4063836973654582,
      "grad_norm": 0.3376450538635254,
      "learning_rate": 0.001,
      "loss": 2.5902,
      "step": 24983
    },
    {
      "epoch": 1.4088043233505967,
      "grad_norm": 0.31322216987609863,
      "learning_rate": 0.001,
      "loss": 2.5876,
      "step": 25026
    },
    {
      "epoch": 1.4112249493357352,
      "grad_norm": 0.3850945234298706,
      "learning_rate": 0.001,
      "loss": 2.5959,
      "step": 25069
    },
    {
      "epoch": 1.4136455753208736,
      "grad_norm": 0.31558579206466675,
      "learning_rate": 0.001,
      "loss": 2.5961,
      "step": 25112
    },
    {
      "epoch": 1.4160662013060121,
      "grad_norm": 0.3538668155670166,
      "learning_rate": 0.001,
      "loss": 2.596,
      "step": 25155
    },
    {
      "epoch": 1.4184868272911506,
      "grad_norm": 0.2993002235889435,
      "learning_rate": 0.001,
      "loss": 2.5886,
      "step": 25198
    },
    {
      "epoch": 1.420907453276289,
      "grad_norm": 0.36599335074424744,
      "learning_rate": 0.001,
      "loss": 2.5857,
      "step": 25241
    },
    {
      "epoch": 1.4233280792614276,
      "grad_norm": 0.314736932516098,
      "learning_rate": 0.001,
      "loss": 2.5901,
      "step": 25284
    },
    {
      "epoch": 1.425748705246566,
      "grad_norm": 0.3434794545173645,
      "learning_rate": 0.001,
      "loss": 2.5917,
      "step": 25327
    },
    {
      "epoch": 1.4281693312317045,
      "grad_norm": 0.41864892840385437,
      "learning_rate": 0.001,
      "loss": 2.5873,
      "step": 25370
    },
    {
      "epoch": 1.430589957216843,
      "grad_norm": 0.32243841886520386,
      "learning_rate": 0.001,
      "loss": 2.5928,
      "step": 25413
    },
    {
      "epoch": 1.4330105832019815,
      "grad_norm": 0.3722641170024872,
      "learning_rate": 0.001,
      "loss": 2.5934,
      "step": 25456
    },
    {
      "epoch": 1.43543120918712,
      "grad_norm": 0.31123247742652893,
      "learning_rate": 0.001,
      "loss": 2.5787,
      "step": 25499
    },
    {
      "epoch": 1.4378518351722585,
      "grad_norm": 0.3269544541835785,
      "learning_rate": 0.001,
      "loss": 2.5913,
      "step": 25542
    },
    {
      "epoch": 1.440272461157397,
      "grad_norm": 0.3256840407848358,
      "learning_rate": 0.001,
      "loss": 2.5835,
      "step": 25585
    },
    {
      "epoch": 1.440891691060572,
      "eval_ag_news_bleu_score": 3.716122739866257,
      "eval_ag_news_bleu_score_sem": 0.13432331437189057,
      "eval_ag_news_emb_cos_sim": 0.9905405640602112,
      "eval_ag_news_emb_cos_sim_sem": 0.00034029789417199576,
      "eval_ag_news_emb_top1_equal": 0.6133333444595337,
      "eval_ag_news_emb_top1_equal_sem": 0.028163139369651306,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.8056228160858154,
      "eval_ag_news_n_ngrams_match_1": 4.636,
      "eval_ag_news_n_ngrams_match_2": 0.776,
      "eval_ag_news_n_ngrams_match_3": 0.202,
      "eval_ag_news_num_pred_words": 22.918,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.24675122944538141,
      "eval_ag_news_runtime": 15.3569,
      "eval_ag_news_samples_per_second": 32.559,
      "eval_ag_news_steps_per_second": 0.13,
      "eval_ag_news_token_set_f1": 0.21580984570619113,
      "eval_ag_news_token_set_f1_sem": 0.005045773731396865,
      "eval_ag_news_token_set_precision": 0.2082900028769648,
      "eval_ag_news_token_set_recall": 0.23317462767590938,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 25596
    },
    {
      "epoch": 1.440891691060572,
      "eval_anthropic_toxic_prompts_bleu_score": 4.750670047452916,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.17779981422053243,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9910176992416382,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00025476251147671353,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6166666746139526,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028117578599086417,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.0310702323913574,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.562,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.346,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.488,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.326,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.26951874076460397,
      "eval_anthropic_toxic_prompts_runtime": 6.6106,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.637,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.303,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3002934707137857,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006377095763985128,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3612667567843973,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2842576049832224,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 25596
    },
    {
      "epoch": 1.440891691060572,
      "eval_arxiv_bleu_score": 3.431355035996916,
      "eval_arxiv_bleu_score_sem": 0.09797404277147521,
      "eval_arxiv_emb_cos_sim": 0.9898726344108582,
      "eval_arxiv_emb_cos_sim_sem": 0.00038886295225455577,
      "eval_arxiv_emb_top1_equal": 0.5733333230018616,
      "eval_arxiv_emb_top1_equal_sem": 0.028603051416510494,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.8203601837158203,
      "eval_arxiv_n_ngrams_match_1": 4.456,
      "eval_arxiv_n_ngrams_match_2": 0.668,
      "eval_arxiv_n_ngrams_match_3": 0.106,
      "eval_arxiv_num_pred_words": 17.874,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.23401551767716483,
      "eval_arxiv_runtime": 6.5492,
      "eval_arxiv_samples_per_second": 76.345,
      "eval_arxiv_steps_per_second": 0.305,
      "eval_arxiv_token_set_f1": 0.22287694905177147,
      "eval_arxiv_token_set_f1_sem": 0.004606144542697508,
      "eval_arxiv_token_set_precision": 0.18519280399956836,
      "eval_arxiv_token_set_recall": 0.3057083019062126,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 25596
    },
    {
      "epoch": 1.440891691060572,
      "eval_python_code_alpaca_bleu_score": 5.193038031326035,
      "eval_python_code_alpaca_bleu_score_sem": 0.1872736595603848,
      "eval_python_code_alpaca_emb_cos_sim": 0.9885314106941223,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00042417735109744433,
      "eval_python_code_alpaca_emb_top1_equal": 0.6499999761581421,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027583864257272155,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.136096715927124,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.212,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.178,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.35,
      "eval_python_code_alpaca_num_pred_words": 19.082,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2947223297744834,
      "eval_python_code_alpaca_runtime": 6.5003,
      "eval_python_code_alpaca_samples_per_second": 76.919,
      "eval_python_code_alpaca_steps_per_second": 0.308,
      "eval_python_code_alpaca_token_set_f1": 0.31539624914237246,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005776658700468536,
      "eval_python_code_alpaca_token_set_precision": 0.3057380937607338,
      "eval_python_code_alpaca_token_set_recall": 0.36289249575188176,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 25596
    },
    {
      "epoch": 1.440891691060572,
      "eval_wikibio_bleu_score": 3.5955396702200977,
      "eval_wikibio_bleu_score_sem": 0.12194113886263454,
      "eval_wikibio_emb_cos_sim": 0.9908457398414612,
      "eval_wikibio_emb_cos_sim_sem": 0.00027669951534569146,
      "eval_wikibio_emb_top1_equal": 0.6133333444595337,
      "eval_wikibio_emb_top1_equal_sem": 0.028163139369651306,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.038972854614258,
      "eval_wikibio_n_ngrams_match_1": 3.06,
      "eval_wikibio_n_ngrams_match_2": 0.75,
      "eval_wikibio_n_ngrams_match_3": 0.15,
      "eval_wikibio_num_pred_words": 20.456,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2464700914572922,
      "eval_wikibio_runtime": 6.4788,
      "eval_wikibio_samples_per_second": 77.175,
      "eval_wikibio_steps_per_second": 0.309,
      "eval_wikibio_token_set_f1": 0.19871518529058282,
      "eval_wikibio_token_set_f1_sem": 0.005100482139048974,
      "eval_wikibio_token_set_precision": 0.2333915722226653,
      "eval_wikibio_token_set_recall": 0.18052384206471736,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 25596
    },
    {
      "epoch": 1.440891691060572,
      "eval_nq_5round_bleu_score": 9.210381157089552,
      "eval_nq_5round_bleu_score_sem": 0.4552743703355639,
      "eval_nq_5round_emb_cos_sim": 0.9927846789360046,
      "eval_nq_5round_emb_cos_sim_sem": 0.00027714300976062784,
      "eval_nq_5round_emb_top1_equal": 0.6899999976158142,
      "eval_nq_5round_emb_top1_equal_sem": 0.026746674129075776,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.934,
      "eval_nq_5round_n_ngrams_match_2": 2.938,
      "eval_nq_5round_n_ngrams_match_3": 1.274,
      "eval_nq_5round_num_pred_words": 23.642,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3564128331667331,
      "eval_nq_5round_token_set_f1": 0.38628060731009634,
      "eval_nq_5round_token_set_f1_sem": 0.00646941789960575,
      "eval_nq_5round_token_set_precision": 0.36195048921191586,
      "eval_nq_5round_token_set_recall": 0.4262570596356479,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.457297377920911,
      "eval_nq_bleu_score_sem": 0.43026476733515245,
      "eval_nq_emb_cos_sim": 0.9926289916038513,
      "eval_nq_emb_cos_sim_sem": 0.00023524805821037182,
      "eval_nq_emb_top1_equal": 0.6333333253860474,
      "eval_nq_emb_top1_equal_sem": 0.02786867456387452,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.326998472213745,
      "eval_nq_n_ngrams_match_1": 9.2,
      "eval_nq_n_ngrams_match_2": 3.034,
      "eval_nq_n_ngrams_match_3": 1.284,
      "eval_nq_num_pred_words": 23.814,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3664782967944617,
      "eval_nq_runtime": 18.9632,
      "eval_nq_samples_per_second": 26.367,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.3989302654848708,
      "eval_nq_token_set_f1_sem": 0.006043670991199448,
      "eval_nq_token_set_precision": 0.37388466223279615,
      "eval_nq_token_set_recall": 0.4381787998640057,
      "eval_nq_true_num_tokens": 32.0,
      "step": 25596
    },
    {
      "epoch": 1.4426930871425354,
      "grad_norm": 0.39198532700538635,
      "learning_rate": 0.001,
      "loss": 2.5996,
      "step": 25628
    },
    {
      "epoch": 1.445113713127674,
      "grad_norm": 0.36448290944099426,
      "learning_rate": 0.001,
      "loss": 2.587,
      "step": 25671
    },
    {
      "epoch": 1.4475343391128124,
      "grad_norm": 0.37067651748657227,
      "learning_rate": 0.001,
      "loss": 2.574,
      "step": 25714
    },
    {
      "epoch": 1.4499549650979509,
      "grad_norm": 0.4051074683666229,
      "learning_rate": 0.001,
      "loss": 2.5779,
      "step": 25757
    },
    {
      "epoch": 1.4523755910830893,
      "grad_norm": 0.3189801573753357,
      "learning_rate": 0.001,
      "loss": 2.577,
      "step": 25800
    },
    {
      "epoch": 1.4547962170682278,
      "grad_norm": 0.3151501417160034,
      "learning_rate": 0.001,
      "loss": 2.5857,
      "step": 25843
    },
    {
      "epoch": 1.4572168430533663,
      "grad_norm": 0.32819801568984985,
      "learning_rate": 0.001,
      "loss": 2.5802,
      "step": 25886
    },
    {
      "epoch": 1.4596374690385048,
      "grad_norm": 0.3272273540496826,
      "learning_rate": 0.001,
      "loss": 2.5711,
      "step": 25929
    },
    {
      "epoch": 1.4620580950236433,
      "grad_norm": 0.3226931691169739,
      "learning_rate": 0.001,
      "loss": 2.5968,
      "step": 25972
    },
    {
      "epoch": 1.4644787210087817,
      "grad_norm": 0.35533058643341064,
      "learning_rate": 0.001,
      "loss": 2.5854,
      "step": 26015
    },
    {
      "epoch": 1.4668993469939202,
      "grad_norm": 0.3328491151332855,
      "learning_rate": 0.001,
      "loss": 2.5838,
      "step": 26058
    },
    {
      "epoch": 1.4693199729790587,
      "grad_norm": 0.3544239103794098,
      "learning_rate": 0.001,
      "loss": 2.5794,
      "step": 26101
    },
    {
      "epoch": 1.4717405989641972,
      "grad_norm": 0.34569498896598816,
      "learning_rate": 0.001,
      "loss": 2.5857,
      "step": 26144
    },
    {
      "epoch": 1.4741612249493357,
      "grad_norm": 0.36936092376708984,
      "learning_rate": 0.001,
      "loss": 2.5843,
      "step": 26187
    },
    {
      "epoch": 1.4765818509344741,
      "grad_norm": 0.3368487060070038,
      "learning_rate": 0.001,
      "loss": 2.5776,
      "step": 26230
    },
    {
      "epoch": 1.4790024769196126,
      "grad_norm": 0.34614938497543335,
      "learning_rate": 0.001,
      "loss": 2.5827,
      "step": 26273
    },
    {
      "epoch": 1.481423102904751,
      "grad_norm": 0.3885331153869629,
      "learning_rate": 0.001,
      "loss": 2.566,
      "step": 26316
    },
    {
      "epoch": 1.4838437288898896,
      "grad_norm": 0.34162846207618713,
      "learning_rate": 0.001,
      "loss": 2.5868,
      "step": 26359
    },
    {
      "epoch": 1.4862643548750283,
      "grad_norm": 0.3429296314716339,
      "learning_rate": 0.001,
      "loss": 2.5857,
      "step": 26402
    },
    {
      "epoch": 1.4886849808601665,
      "grad_norm": 0.3596552014350891,
      "learning_rate": 0.001,
      "loss": 2.5856,
      "step": 26445
    },
    {
      "epoch": 1.4911056068453052,
      "grad_norm": 0.3357451260089874,
      "learning_rate": 0.001,
      "loss": 2.5778,
      "step": 26488
    },
    {
      "epoch": 1.4935262328304435,
      "grad_norm": 0.3357361853122711,
      "learning_rate": 0.001,
      "loss": 2.5889,
      "step": 26531
    },
    {
      "epoch": 1.4959468588155822,
      "grad_norm": 0.34575578570365906,
      "learning_rate": 0.001,
      "loss": 2.5696,
      "step": 26574
    },
    {
      "epoch": 1.4983674848007205,
      "grad_norm": 0.29873916506767273,
      "learning_rate": 0.001,
      "loss": 2.5786,
      "step": 26617
    },
    {
      "epoch": 1.5007881107858592,
      "grad_norm": 0.3526674509048462,
      "learning_rate": 0.001,
      "loss": 2.5719,
      "step": 26660
    },
    {
      "epoch": 1.5032087367709974,
      "grad_norm": 0.3181329071521759,
      "learning_rate": 0.001,
      "loss": 2.5832,
      "step": 26703
    },
    {
      "epoch": 1.5056293627561361,
      "grad_norm": 0.32514214515686035,
      "learning_rate": 0.001,
      "loss": 2.5686,
      "step": 26746
    },
    {
      "epoch": 1.5080499887412744,
      "grad_norm": 0.33324673771858215,
      "learning_rate": 0.001,
      "loss": 2.5669,
      "step": 26789
    },
    {
      "epoch": 1.510470614726413,
      "grad_norm": 0.3169005215167999,
      "learning_rate": 0.001,
      "loss": 2.5886,
      "step": 26832
    },
    {
      "epoch": 1.5128912407115513,
      "grad_norm": 0.31450915336608887,
      "learning_rate": 0.001,
      "loss": 2.5663,
      "step": 26875
    },
    {
      "epoch": 1.51531186669669,
      "grad_norm": 0.3062874674797058,
      "learning_rate": 0.001,
      "loss": 2.582,
      "step": 26918
    },
    {
      "epoch": 1.5177324926818283,
      "grad_norm": 0.3170241713523865,
      "learning_rate": 0.001,
      "loss": 2.5912,
      "step": 26961
    },
    {
      "epoch": 1.520153118666967,
      "grad_norm": 0.33269351720809937,
      "learning_rate": 0.001,
      "loss": 2.5735,
      "step": 27004
    },
    {
      "epoch": 1.5225737446521053,
      "grad_norm": 0.3486937880516052,
      "learning_rate": 0.001,
      "loss": 2.5767,
      "step": 27047
    },
    {
      "epoch": 1.524994370637244,
      "grad_norm": 0.3906833827495575,
      "learning_rate": 0.001,
      "loss": 2.578,
      "step": 27090
    },
    {
      "epoch": 1.5274149966223822,
      "grad_norm": 0.2900930643081665,
      "learning_rate": 0.001,
      "loss": 2.5819,
      "step": 27133
    },
    {
      "epoch": 1.529835622607521,
      "grad_norm": 0.3287878930568695,
      "learning_rate": 0.001,
      "loss": 2.5736,
      "step": 27176
    },
    {
      "epoch": 1.5322562485926592,
      "grad_norm": 0.3648246228694916,
      "learning_rate": 0.001,
      "loss": 2.5705,
      "step": 27219
    },
    {
      "epoch": 1.534676874577798,
      "grad_norm": 0.3586522936820984,
      "learning_rate": 0.001,
      "loss": 2.5743,
      "step": 27262
    },
    {
      "epoch": 1.5370975005629361,
      "grad_norm": 0.35216784477233887,
      "learning_rate": 0.001,
      "loss": 2.5651,
      "step": 27305
    },
    {
      "epoch": 1.5395181265480749,
      "grad_norm": 0.3033185303211212,
      "learning_rate": 0.001,
      "loss": 2.5695,
      "step": 27348
    },
    {
      "epoch": 1.541938752533213,
      "grad_norm": 0.3198002874851227,
      "learning_rate": 0.001,
      "loss": 2.5742,
      "step": 27391
    },
    {
      "epoch": 1.5443593785183518,
      "grad_norm": 0.38163867592811584,
      "learning_rate": 0.001,
      "loss": 2.5692,
      "step": 27434
    },
    {
      "epoch": 1.54678000450349,
      "grad_norm": 0.3662133812904358,
      "learning_rate": 0.001,
      "loss": 2.5821,
      "step": 27477
    },
    {
      "epoch": 1.5492006304886288,
      "grad_norm": 0.3756394386291504,
      "learning_rate": 0.001,
      "loss": 2.5819,
      "step": 27520
    },
    {
      "epoch": 1.551621256473767,
      "grad_norm": 0.4120120704174042,
      "learning_rate": 0.001,
      "loss": 2.5556,
      "step": 27563
    },
    {
      "epoch": 1.5540418824589057,
      "grad_norm": 0.3079850673675537,
      "learning_rate": 0.001,
      "loss": 2.558,
      "step": 27606
    },
    {
      "epoch": 1.556462508444044,
      "grad_norm": 0.3087233901023865,
      "learning_rate": 0.001,
      "loss": 2.5724,
      "step": 27649
    },
    {
      "epoch": 1.5588831344291827,
      "grad_norm": 0.34643685817718506,
      "learning_rate": 0.001,
      "loss": 2.5771,
      "step": 27692
    },
    {
      "epoch": 1.5609659986489528,
      "eval_ag_news_bleu_score": 3.845697889677681,
      "eval_ag_news_bleu_score_sem": 0.14207184895121333,
      "eval_ag_news_emb_cos_sim": 0.9907991886138916,
      "eval_ag_news_emb_cos_sim_sem": 0.00032400030785705947,
      "eval_ag_news_emb_top1_equal": 0.6100000143051147,
      "eval_ag_news_emb_top1_equal_sem": 0.02820730814416699,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.755255699157715,
      "eval_ag_news_n_ngrams_match_1": 4.62,
      "eval_ag_news_n_ngrams_match_2": 0.84,
      "eval_ag_news_n_ngrams_match_3": 0.21,
      "eval_ag_news_num_pred_words": 22.498,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.24981528910313575,
      "eval_ag_news_runtime": 16.0007,
      "eval_ag_news_samples_per_second": 31.249,
      "eval_ag_news_steps_per_second": 0.125,
      "eval_ag_news_token_set_f1": 0.2177090970037736,
      "eval_ag_news_token_set_f1_sem": 0.005141731163310442,
      "eval_ag_news_token_set_precision": 0.20932240725718337,
      "eval_ag_news_token_set_recall": 0.235903699702057,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 27729
    },
    {
      "epoch": 1.5609659986489528,
      "eval_anthropic_toxic_prompts_bleu_score": 4.817324215010732,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.20615165613499215,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9911608099937439,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002886941626990269,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6433333158493042,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027702163273800266,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.0208563804626465,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.596,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.32,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.5,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.526,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.26765396490580023,
      "eval_anthropic_toxic_prompts_runtime": 6.726,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.339,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.297,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.299599182338913,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006331666547060622,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.36276119743774266,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.27906384603889667,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 27729
    },
    {
      "epoch": 1.5609659986489528,
      "eval_arxiv_bleu_score": 3.4395886104611626,
      "eval_arxiv_bleu_score_sem": 0.11585969222772027,
      "eval_arxiv_emb_cos_sim": 0.9899406433105469,
      "eval_arxiv_emb_cos_sim_sem": 0.000286655852044152,
      "eval_arxiv_emb_top1_equal": 0.5766666531562805,
      "eval_arxiv_emb_top1_equal_sem": 0.028573804013736142,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.7899527549743652,
      "eval_arxiv_n_ngrams_match_1": 4.392,
      "eval_arxiv_n_ngrams_match_2": 0.666,
      "eval_arxiv_n_ngrams_match_3": 0.12,
      "eval_arxiv_num_pred_words": 17.224,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.23687140618032448,
      "eval_arxiv_runtime": 6.7258,
      "eval_arxiv_samples_per_second": 74.341,
      "eval_arxiv_steps_per_second": 0.297,
      "eval_arxiv_token_set_f1": 0.22153284809698306,
      "eval_arxiv_token_set_f1_sem": 0.004906847785568647,
      "eval_arxiv_token_set_precision": 0.1864509907923353,
      "eval_arxiv_token_set_recall": 0.30094447379432904,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 27729
    },
    {
      "epoch": 1.5609659986489528,
      "eval_python_code_alpaca_bleu_score": 5.103998158847488,
      "eval_python_code_alpaca_bleu_score_sem": 0.19800115367103394,
      "eval_python_code_alpaca_emb_cos_sim": 0.9883846044540405,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00043209405977764903,
      "eval_python_code_alpaca_emb_top1_equal": 0.6166666746139526,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028117578599086417,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.0995583534240723,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.244,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.162,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.354,
      "eval_python_code_alpaca_num_pred_words": 18.952,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2958125910728323,
      "eval_python_code_alpaca_runtime": 10.0323,
      "eval_python_code_alpaca_samples_per_second": 49.839,
      "eval_python_code_alpaca_steps_per_second": 0.199,
      "eval_python_code_alpaca_token_set_f1": 0.3144728548160174,
      "eval_python_code_alpaca_token_set_f1_sem": 0.0058890774342825695,
      "eval_python_code_alpaca_token_set_precision": 0.3078261544332788,
      "eval_python_code_alpaca_token_set_recall": 0.35757910528322484,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 27729
    },
    {
      "epoch": 1.5609659986489528,
      "eval_wikibio_bleu_score": 3.64588403685935,
      "eval_wikibio_bleu_score_sem": 0.12527715646976453,
      "eval_wikibio_emb_cos_sim": 0.9903876185417175,
      "eval_wikibio_emb_cos_sim_sem": 0.0003100774706735083,
      "eval_wikibio_emb_top1_equal": 0.6399999856948853,
      "eval_wikibio_emb_top1_equal_sem": 0.02775911810844162,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.072665691375732,
      "eval_wikibio_n_ngrams_match_1": 3.064,
      "eval_wikibio_n_ngrams_match_2": 0.762,
      "eval_wikibio_n_ngrams_match_3": 0.152,
      "eval_wikibio_num_pred_words": 19.996,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2462997228842188,
      "eval_wikibio_runtime": 6.7717,
      "eval_wikibio_samples_per_second": 73.837,
      "eval_wikibio_steps_per_second": 0.295,
      "eval_wikibio_token_set_f1": 0.19724698123734596,
      "eval_wikibio_token_set_f1_sem": 0.00518032600563407,
      "eval_wikibio_token_set_precision": 0.2325707472242391,
      "eval_wikibio_token_set_recall": 0.17737356707709534,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 27729
    },
    {
      "epoch": 1.5609659986489528,
      "eval_nq_5round_bleu_score": 9.35810170604886,
      "eval_nq_5round_bleu_score_sem": 0.45350170045868954,
      "eval_nq_5round_emb_cos_sim": 0.9926011562347412,
      "eval_nq_5round_emb_cos_sim_sem": 0.0002617137809934625,
      "eval_nq_5round_emb_top1_equal": 0.6333333253860474,
      "eval_nq_5round_emb_top1_equal_sem": 0.02786867456387452,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 8.946,
      "eval_nq_5round_n_ngrams_match_2": 3.034,
      "eval_nq_5round_n_ngrams_match_3": 1.304,
      "eval_nq_5round_num_pred_words": 23.546,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3568994565772782,
      "eval_nq_5round_token_set_f1": 0.38864664446719965,
      "eval_nq_5round_token_set_f1_sem": 0.006346364696935003,
      "eval_nq_5round_token_set_precision": 0.36355263884331124,
      "eval_nq_5round_token_set_recall": 0.4326043839762792,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.37949485197667,
      "eval_nq_bleu_score_sem": 0.4279979677788854,
      "eval_nq_emb_cos_sim": 0.9930351972579956,
      "eval_nq_emb_cos_sim_sem": 0.000209200020273938,
      "eval_nq_emb_top1_equal": 0.6833333373069763,
      "eval_nq_emb_top1_equal_sem": 0.02690183092974493,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.2910776138305664,
      "eval_nq_n_ngrams_match_1": 9.188,
      "eval_nq_n_ngrams_match_2": 3.034,
      "eval_nq_n_ngrams_match_3": 1.282,
      "eval_nq_num_pred_words": 23.812,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3639789700247893,
      "eval_nq_runtime": 19.0718,
      "eval_nq_samples_per_second": 26.217,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.3971573393939799,
      "eval_nq_token_set_f1_sem": 0.0059470537006912115,
      "eval_nq_token_set_precision": 0.37216311928817414,
      "eval_nq_token_set_recall": 0.4399523242317331,
      "eval_nq_true_num_tokens": 32.0,
      "step": 27729
    },
    {
      "epoch": 1.561303760414321,
      "grad_norm": 0.36055082082748413,
      "learning_rate": 0.001,
      "loss": 2.5603,
      "step": 27735
    },
    {
      "epoch": 1.5637243863994597,
      "grad_norm": 0.314640074968338,
      "learning_rate": 0.001,
      "loss": 2.5673,
      "step": 27778
    },
    {
      "epoch": 1.566145012384598,
      "grad_norm": 0.3926333785057068,
      "learning_rate": 0.001,
      "loss": 2.573,
      "step": 27821
    },
    {
      "epoch": 1.5685656383697366,
      "grad_norm": 0.34943607449531555,
      "learning_rate": 0.001,
      "loss": 2.5638,
      "step": 27864
    },
    {
      "epoch": 1.5709862643548749,
      "grad_norm": 0.36166220903396606,
      "learning_rate": 0.001,
      "loss": 2.5581,
      "step": 27907
    },
    {
      "epoch": 1.5734068903400136,
      "grad_norm": 0.3463379144668579,
      "learning_rate": 0.001,
      "loss": 2.5646,
      "step": 27950
    },
    {
      "epoch": 1.5758275163251518,
      "grad_norm": 0.3920922875404358,
      "learning_rate": 0.001,
      "loss": 2.5639,
      "step": 27993
    },
    {
      "epoch": 1.5782481423102905,
      "grad_norm": 0.33013153076171875,
      "learning_rate": 0.001,
      "loss": 2.5731,
      "step": 28036
    },
    {
      "epoch": 1.5806687682954288,
      "grad_norm": 0.31808432936668396,
      "learning_rate": 0.001,
      "loss": 2.5512,
      "step": 28079
    },
    {
      "epoch": 1.5830893942805675,
      "grad_norm": 0.3477192223072052,
      "learning_rate": 0.001,
      "loss": 2.558,
      "step": 28122
    },
    {
      "epoch": 1.585510020265706,
      "grad_norm": 0.31803596019744873,
      "learning_rate": 0.001,
      "loss": 2.556,
      "step": 28165
    },
    {
      "epoch": 1.5879306462508445,
      "grad_norm": 0.315228670835495,
      "learning_rate": 0.001,
      "loss": 2.5672,
      "step": 28208
    },
    {
      "epoch": 1.590351272235983,
      "grad_norm": 0.3228311538696289,
      "learning_rate": 0.001,
      "loss": 2.5512,
      "step": 28251
    },
    {
      "epoch": 1.5927718982211214,
      "grad_norm": 0.39499005675315857,
      "learning_rate": 0.001,
      "loss": 2.5566,
      "step": 28294
    },
    {
      "epoch": 1.59519252420626,
      "grad_norm": 0.3063789904117584,
      "learning_rate": 0.001,
      "loss": 2.5555,
      "step": 28337
    },
    {
      "epoch": 1.5976131501913984,
      "grad_norm": 0.3168655335903168,
      "learning_rate": 0.001,
      "loss": 2.5657,
      "step": 28380
    },
    {
      "epoch": 1.6000337761765369,
      "grad_norm": 0.30188482999801636,
      "learning_rate": 0.001,
      "loss": 2.5695,
      "step": 28423
    },
    {
      "epoch": 1.6024544021616753,
      "grad_norm": 0.3197798728942871,
      "learning_rate": 0.001,
      "loss": 2.5544,
      "step": 28466
    },
    {
      "epoch": 1.6048750281468138,
      "grad_norm": 0.3701176643371582,
      "learning_rate": 0.001,
      "loss": 2.5504,
      "step": 28509
    },
    {
      "epoch": 1.6072956541319523,
      "grad_norm": 0.32933440804481506,
      "learning_rate": 0.001,
      "loss": 2.5724,
      "step": 28552
    },
    {
      "epoch": 1.6097162801170908,
      "grad_norm": 0.29522940516471863,
      "learning_rate": 0.001,
      "loss": 2.5542,
      "step": 28595
    },
    {
      "epoch": 1.6121369061022293,
      "grad_norm": 0.3630578815937042,
      "learning_rate": 0.001,
      "loss": 2.5525,
      "step": 28638
    },
    {
      "epoch": 1.6145575320873677,
      "grad_norm": 0.3430267870426178,
      "learning_rate": 0.001,
      "loss": 2.543,
      "step": 28681
    },
    {
      "epoch": 1.6169781580725062,
      "grad_norm": 0.40477126836776733,
      "learning_rate": 0.001,
      "loss": 2.5612,
      "step": 28724
    },
    {
      "epoch": 1.6193987840576447,
      "grad_norm": 0.32684680819511414,
      "learning_rate": 0.001,
      "loss": 2.562,
      "step": 28767
    },
    {
      "epoch": 1.6218194100427832,
      "grad_norm": 0.3561391532421112,
      "learning_rate": 0.001,
      "loss": 2.559,
      "step": 28810
    },
    {
      "epoch": 1.6242400360279217,
      "grad_norm": 0.3796881139278412,
      "learning_rate": 0.001,
      "loss": 2.5585,
      "step": 28853
    },
    {
      "epoch": 1.6266606620130601,
      "grad_norm": 0.3193817138671875,
      "learning_rate": 0.001,
      "loss": 2.5481,
      "step": 28896
    },
    {
      "epoch": 1.6290812879981986,
      "grad_norm": 0.33970242738723755,
      "learning_rate": 0.001,
      "loss": 2.5629,
      "step": 28939
    },
    {
      "epoch": 1.631501913983337,
      "grad_norm": 0.36553746461868286,
      "learning_rate": 0.001,
      "loss": 2.5528,
      "step": 28982
    },
    {
      "epoch": 1.6339225399684756,
      "grad_norm": 0.3388936221599579,
      "learning_rate": 0.001,
      "loss": 2.5481,
      "step": 29025
    },
    {
      "epoch": 1.636343165953614,
      "grad_norm": 0.32142627239227295,
      "learning_rate": 0.001,
      "loss": 2.5498,
      "step": 29068
    },
    {
      "epoch": 1.6387637919387525,
      "grad_norm": 0.3467428982257843,
      "learning_rate": 0.001,
      "loss": 2.5447,
      "step": 29111
    },
    {
      "epoch": 1.641184417923891,
      "grad_norm": 0.3176419734954834,
      "learning_rate": 0.001,
      "loss": 2.5464,
      "step": 29154
    },
    {
      "epoch": 1.6436050439090295,
      "grad_norm": 0.36646729707717896,
      "learning_rate": 0.001,
      "loss": 2.5443,
      "step": 29197
    },
    {
      "epoch": 1.646025669894168,
      "grad_norm": 0.35340917110443115,
      "learning_rate": 0.001,
      "loss": 2.5433,
      "step": 29240
    },
    {
      "epoch": 1.6484462958793065,
      "grad_norm": 0.3294951319694519,
      "learning_rate": 0.001,
      "loss": 2.547,
      "step": 29283
    },
    {
      "epoch": 1.650866921864445,
      "grad_norm": 0.36375692486763,
      "learning_rate": 0.001,
      "loss": 2.5474,
      "step": 29326
    },
    {
      "epoch": 1.6532875478495834,
      "grad_norm": 0.3293711543083191,
      "learning_rate": 0.001,
      "loss": 2.5483,
      "step": 29369
    },
    {
      "epoch": 1.655708173834722,
      "grad_norm": 0.33201128244400024,
      "learning_rate": 0.001,
      "loss": 2.5492,
      "step": 29412
    },
    {
      "epoch": 1.6581287998198604,
      "grad_norm": 0.3814696669578552,
      "learning_rate": 0.001,
      "loss": 2.5361,
      "step": 29455
    },
    {
      "epoch": 1.6605494258049989,
      "grad_norm": 0.311596155166626,
      "learning_rate": 0.001,
      "loss": 2.5558,
      "step": 29498
    },
    {
      "epoch": 1.6629700517901373,
      "grad_norm": 0.35451462864875793,
      "learning_rate": 0.001,
      "loss": 2.5494,
      "step": 29541
    },
    {
      "epoch": 1.6653906777752758,
      "grad_norm": 0.2953123152256012,
      "learning_rate": 0.001,
      "loss": 2.5422,
      "step": 29584
    },
    {
      "epoch": 1.6678113037604143,
      "grad_norm": 0.3245306611061096,
      "learning_rate": 0.001,
      "loss": 2.5378,
      "step": 29627
    },
    {
      "epoch": 1.6702319297455528,
      "grad_norm": 0.3472495377063751,
      "learning_rate": 0.001,
      "loss": 2.5408,
      "step": 29670
    },
    {
      "epoch": 1.6726525557306913,
      "grad_norm": 0.3620971739292145,
      "learning_rate": 0.001,
      "loss": 2.546,
      "step": 29713
    },
    {
      "epoch": 1.6750731817158298,
      "grad_norm": 0.3200334310531616,
      "learning_rate": 0.001,
      "loss": 2.5532,
      "step": 29756
    },
    {
      "epoch": 1.6774938077009682,
      "grad_norm": 0.3324032723903656,
      "learning_rate": 0.001,
      "loss": 2.54,
      "step": 29799
    },
    {
      "epoch": 1.6799144336861067,
      "grad_norm": 0.3374853730201721,
      "learning_rate": 0.001,
      "loss": 2.5506,
      "step": 29842
    },
    {
      "epoch": 1.681040306237334,
      "eval_ag_news_bleu_score": 3.896334064492728,
      "eval_ag_news_bleu_score_sem": 0.14659939619600168,
      "eval_ag_news_emb_cos_sim": 0.9917250871658325,
      "eval_ag_news_emb_cos_sim_sem": 0.0002371674297714253,
      "eval_ag_news_emb_top1_equal": 0.6133333444595337,
      "eval_ag_news_emb_top1_equal_sem": 0.028163139369651306,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.747325897216797,
      "eval_ag_news_n_ngrams_match_1": 4.68,
      "eval_ag_news_n_ngrams_match_2": 0.856,
      "eval_ag_news_n_ngrams_match_3": 0.23,
      "eval_ag_news_num_pred_words": 22.598,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.25176961332628667,
      "eval_ag_news_runtime": 15.3696,
      "eval_ag_news_samples_per_second": 32.532,
      "eval_ag_news_steps_per_second": 0.13,
      "eval_ag_news_token_set_f1": 0.2218716931809878,
      "eval_ag_news_token_set_f1_sem": 0.005128545431168542,
      "eval_ag_news_token_set_precision": 0.2125292783871526,
      "eval_ag_news_token_set_recall": 0.240299549138811,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 29862
    },
    {
      "epoch": 1.681040306237334,
      "eval_anthropic_toxic_prompts_bleu_score": 4.8852351886478065,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.20654275811796327,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9910754561424255,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00028764857944906994,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6600000262260437,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.0273952875069568,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 3.014017105102539,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.656,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.36,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.514,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.328,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.27361696727581447,
      "eval_anthropic_toxic_prompts_runtime": 6.7325,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.267,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.297,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.30239248864609997,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006391089572408326,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.36919183315737225,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.28291881137742236,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 29862
    },
    {
      "epoch": 1.681040306237334,
      "eval_arxiv_bleu_score": 3.6018437014853784,
      "eval_arxiv_bleu_score_sem": 0.11607438869599608,
      "eval_arxiv_emb_cos_sim": 0.9903132319450378,
      "eval_arxiv_emb_cos_sim_sem": 0.00026585874388921916,
      "eval_arxiv_emb_top1_equal": 0.5666666626930237,
      "eval_arxiv_emb_top1_equal_sem": 0.02865756638663049,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.793177843093872,
      "eval_arxiv_n_ngrams_match_1": 4.552,
      "eval_arxiv_n_ngrams_match_2": 0.676,
      "eval_arxiv_n_ngrams_match_3": 0.122,
      "eval_arxiv_num_pred_words": 17.61,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.24314520622013752,
      "eval_arxiv_runtime": 6.7017,
      "eval_arxiv_samples_per_second": 74.608,
      "eval_arxiv_steps_per_second": 0.298,
      "eval_arxiv_token_set_f1": 0.22867657903089852,
      "eval_arxiv_token_set_f1_sem": 0.004592964811806698,
      "eval_arxiv_token_set_precision": 0.19347479590520172,
      "eval_arxiv_token_set_recall": 0.2997395708343349,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 29862
    },
    {
      "epoch": 1.681040306237334,
      "eval_python_code_alpaca_bleu_score": 5.392375325167191,
      "eval_python_code_alpaca_bleu_score_sem": 0.20883856825049046,
      "eval_python_code_alpaca_emb_cos_sim": 0.9885626435279846,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004551268411298943,
      "eval_python_code_alpaca_emb_top1_equal": 0.5566666722297668,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028729442593013178,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.0999677181243896,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.08,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.152,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.38,
      "eval_python_code_alpaca_num_pred_words": 18.498,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.29164074047731503,
      "eval_python_code_alpaca_runtime": 7.0778,
      "eval_python_code_alpaca_samples_per_second": 70.643,
      "eval_python_code_alpaca_steps_per_second": 0.283,
      "eval_python_code_alpaca_token_set_f1": 0.3128965537459449,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005795368630083134,
      "eval_python_code_alpaca_token_set_precision": 0.3010884417580598,
      "eval_python_code_alpaca_token_set_recall": 0.36404740538958885,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 29862
    },
    {
      "epoch": 1.681040306237334,
      "eval_wikibio_bleu_score": 3.6055492293686315,
      "eval_wikibio_bleu_score_sem": 0.12655652228724473,
      "eval_wikibio_emb_cos_sim": 0.9909404516220093,
      "eval_wikibio_emb_cos_sim_sem": 0.00026094309653064933,
      "eval_wikibio_emb_top1_equal": 0.6399999856948853,
      "eval_wikibio_emb_top1_equal_sem": 0.02775911810844162,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.0505757331848145,
      "eval_wikibio_n_ngrams_match_1": 3.014,
      "eval_wikibio_n_ngrams_match_2": 0.75,
      "eval_wikibio_n_ngrams_match_3": 0.156,
      "eval_wikibio_num_pred_words": 19.992,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.24365621589580577,
      "eval_wikibio_runtime": 6.5116,
      "eval_wikibio_samples_per_second": 76.786,
      "eval_wikibio_steps_per_second": 0.307,
      "eval_wikibio_token_set_f1": 0.19528821248109665,
      "eval_wikibio_token_set_f1_sem": 0.005268419125368062,
      "eval_wikibio_token_set_precision": 0.22897742797636988,
      "eval_wikibio_token_set_recall": 0.1769831253763911,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 29862
    },
    {
      "epoch": 1.681040306237334,
      "eval_nq_5round_bleu_score": 9.633335624358354,
      "eval_nq_5round_bleu_score_sem": 0.469771159019,
      "eval_nq_5round_emb_cos_sim": 0.9930838346481323,
      "eval_nq_5round_emb_cos_sim_sem": 0.00024550316754663477,
      "eval_nq_5round_emb_top1_equal": 0.6433333158493042,
      "eval_nq_5round_emb_top1_equal_sem": 0.027702163273800266,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 9.202,
      "eval_nq_5round_n_ngrams_match_2": 3.102,
      "eval_nq_5round_n_ngrams_match_3": 1.33,
      "eval_nq_5round_num_pred_words": 23.368,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.3675532684439868,
      "eval_nq_5round_token_set_f1": 0.39844244736873297,
      "eval_nq_5round_token_set_f1_sem": 0.006685057904941863,
      "eval_nq_5round_token_set_precision": 0.3749828026732941,
      "eval_nq_5round_token_set_recall": 0.4357601531812497,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.3992000534323,
      "eval_nq_bleu_score_sem": 0.43567844286730695,
      "eval_nq_emb_cos_sim": 0.9928473234176636,
      "eval_nq_emb_cos_sim_sem": 0.00022136663731841925,
      "eval_nq_emb_top1_equal": 0.6700000166893005,
      "eval_nq_emb_top1_equal_sem": 0.0271931022316066,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.2687196731567383,
      "eval_nq_n_ngrams_match_1": 9.238,
      "eval_nq_n_ngrams_match_2": 3.03,
      "eval_nq_n_ngrams_match_3": 1.28,
      "eval_nq_num_pred_words": 23.698,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.36842138099839894,
      "eval_nq_runtime": 18.7231,
      "eval_nq_samples_per_second": 26.705,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.40442870629297284,
      "eval_nq_token_set_f1_sem": 0.006053105916465227,
      "eval_nq_token_set_precision": 0.37901621883379427,
      "eval_nq_token_set_recall": 0.4441219684094393,
      "eval_nq_true_num_tokens": 32.0,
      "step": 29862
    },
    {
      "epoch": 1.6823350596712452,
      "grad_norm": 0.36387988924980164,
      "learning_rate": 0.001,
      "loss": 2.5464,
      "step": 29885
    },
    {
      "epoch": 1.6847556856563837,
      "grad_norm": 0.32876670360565186,
      "learning_rate": 0.001,
      "loss": 2.5457,
      "step": 29928
    },
    {
      "epoch": 1.6871763116415222,
      "grad_norm": 0.3839133381843567,
      "learning_rate": 0.001,
      "loss": 2.5281,
      "step": 29971
    },
    {
      "epoch": 1.6895969376266606,
      "grad_norm": 0.3589411973953247,
      "learning_rate": 0.001,
      "loss": 2.5358,
      "step": 30014
    },
    {
      "epoch": 1.6920175636117991,
      "grad_norm": 0.44234898686408997,
      "learning_rate": 0.001,
      "loss": 2.5295,
      "step": 30057
    },
    {
      "epoch": 1.6944381895969376,
      "grad_norm": 0.312084823846817,
      "learning_rate": 0.001,
      "loss": 2.5467,
      "step": 30100
    },
    {
      "epoch": 1.696858815582076,
      "grad_norm": 0.36416760087013245,
      "learning_rate": 0.001,
      "loss": 2.5398,
      "step": 30143
    },
    {
      "epoch": 1.6992794415672146,
      "grad_norm": 0.403786301612854,
      "learning_rate": 0.001,
      "loss": 2.5453,
      "step": 30186
    },
    {
      "epoch": 1.701700067552353,
      "grad_norm": 0.34267759323120117,
      "learning_rate": 0.001,
      "loss": 2.5409,
      "step": 30229
    },
    {
      "epoch": 1.7041206935374915,
      "grad_norm": 0.3161322772502899,
      "learning_rate": 0.001,
      "loss": 2.5458,
      "step": 30272
    },
    {
      "epoch": 1.70654131952263,
      "grad_norm": 0.3275354504585266,
      "learning_rate": 0.001,
      "loss": 2.5334,
      "step": 30315
    },
    {
      "epoch": 1.7089619455077685,
      "grad_norm": 0.3565298914909363,
      "learning_rate": 0.001,
      "loss": 2.5277,
      "step": 30358
    },
    {
      "epoch": 1.711382571492907,
      "grad_norm": 0.36023104190826416,
      "learning_rate": 0.001,
      "loss": 2.5332,
      "step": 30401
    },
    {
      "epoch": 1.7138031974780454,
      "grad_norm": 0.3294886350631714,
      "learning_rate": 0.001,
      "loss": 2.5444,
      "step": 30444
    },
    {
      "epoch": 1.716223823463184,
      "grad_norm": 0.3600999414920807,
      "learning_rate": 0.001,
      "loss": 2.5325,
      "step": 30487
    },
    {
      "epoch": 1.7186444494483224,
      "grad_norm": 0.36945220828056335,
      "learning_rate": 0.001,
      "loss": 2.5381,
      "step": 30530
    },
    {
      "epoch": 1.7210650754334609,
      "grad_norm": 0.41342002153396606,
      "learning_rate": 0.001,
      "loss": 2.5438,
      "step": 30573
    },
    {
      "epoch": 1.7234857014185994,
      "grad_norm": 0.3123305141925812,
      "learning_rate": 0.001,
      "loss": 2.5375,
      "step": 30616
    },
    {
      "epoch": 1.725906327403738,
      "grad_norm": 0.41879498958587646,
      "learning_rate": 0.001,
      "loss": 2.5344,
      "step": 30659
    },
    {
      "epoch": 1.7283269533888763,
      "grad_norm": 0.35645821690559387,
      "learning_rate": 0.001,
      "loss": 2.5204,
      "step": 30702
    },
    {
      "epoch": 1.730747579374015,
      "grad_norm": 0.3465394377708435,
      "learning_rate": 0.001,
      "loss": 2.5303,
      "step": 30745
    },
    {
      "epoch": 1.7331682053591533,
      "grad_norm": 0.3368297219276428,
      "learning_rate": 0.001,
      "loss": 2.5333,
      "step": 30788
    },
    {
      "epoch": 1.735588831344292,
      "grad_norm": 0.43187782168388367,
      "learning_rate": 0.001,
      "loss": 2.5392,
      "step": 30831
    },
    {
      "epoch": 1.7380094573294302,
      "grad_norm": 0.3638305068016052,
      "learning_rate": 0.001,
      "loss": 2.549,
      "step": 30874
    },
    {
      "epoch": 1.740430083314569,
      "grad_norm": 0.29354238510131836,
      "learning_rate": 0.001,
      "loss": 2.53,
      "step": 30917
    },
    {
      "epoch": 1.7428507092997072,
      "grad_norm": 0.4030610918998718,
      "learning_rate": 0.001,
      "loss": 2.5276,
      "step": 30960
    },
    {
      "epoch": 1.745271335284846,
      "grad_norm": 0.3239133059978485,
      "learning_rate": 0.001,
      "loss": 2.5272,
      "step": 31003
    },
    {
      "epoch": 1.7476919612699842,
      "grad_norm": 0.38051638007164,
      "learning_rate": 0.001,
      "loss": 2.5301,
      "step": 31046
    },
    {
      "epoch": 1.7501125872551229,
      "grad_norm": 0.36661601066589355,
      "learning_rate": 0.001,
      "loss": 2.5292,
      "step": 31089
    },
    {
      "epoch": 1.7525332132402611,
      "grad_norm": 0.3409247100353241,
      "learning_rate": 0.001,
      "loss": 2.5231,
      "step": 31132
    },
    {
      "epoch": 1.7549538392253998,
      "grad_norm": 0.3360842764377594,
      "learning_rate": 0.001,
      "loss": 2.5278,
      "step": 31175
    },
    {
      "epoch": 1.757374465210538,
      "grad_norm": 0.3958543837070465,
      "learning_rate": 0.001,
      "loss": 2.525,
      "step": 31218
    },
    {
      "epoch": 1.7597950911956768,
      "grad_norm": 0.39974021911621094,
      "learning_rate": 0.001,
      "loss": 2.5221,
      "step": 31261
    },
    {
      "epoch": 1.762215717180815,
      "grad_norm": 0.3530482053756714,
      "learning_rate": 0.001,
      "loss": 2.5316,
      "step": 31304
    },
    {
      "epoch": 1.7646363431659537,
      "grad_norm": 0.336683988571167,
      "learning_rate": 0.001,
      "loss": 2.5195,
      "step": 31347
    },
    {
      "epoch": 1.767056969151092,
      "grad_norm": 0.3105461895465851,
      "learning_rate": 0.001,
      "loss": 2.5294,
      "step": 31390
    },
    {
      "epoch": 1.7694775951362307,
      "grad_norm": 0.3529985845088959,
      "learning_rate": 0.001,
      "loss": 2.5322,
      "step": 31433
    },
    {
      "epoch": 1.771898221121369,
      "grad_norm": 0.3589237928390503,
      "learning_rate": 0.001,
      "loss": 2.5236,
      "step": 31476
    },
    {
      "epoch": 1.7743188471065077,
      "grad_norm": 0.38137444853782654,
      "learning_rate": 0.001,
      "loss": 2.5336,
      "step": 31519
    },
    {
      "epoch": 1.776739473091646,
      "grad_norm": 0.3179773688316345,
      "learning_rate": 0.001,
      "loss": 2.5241,
      "step": 31562
    },
    {
      "epoch": 1.7791600990767846,
      "grad_norm": 0.39458972215652466,
      "learning_rate": 0.001,
      "loss": 2.523,
      "step": 31605
    },
    {
      "epoch": 1.7815807250619229,
      "grad_norm": 0.3063313066959381,
      "learning_rate": 0.001,
      "loss": 2.5153,
      "step": 31648
    },
    {
      "epoch": 1.7840013510470616,
      "grad_norm": 0.38892364501953125,
      "learning_rate": 0.001,
      "loss": 2.5151,
      "step": 31691
    },
    {
      "epoch": 1.7864219770321998,
      "grad_norm": 0.3597552180290222,
      "learning_rate": 0.001,
      "loss": 2.5295,
      "step": 31734
    },
    {
      "epoch": 1.7888426030173386,
      "grad_norm": 0.38583481311798096,
      "learning_rate": 0.001,
      "loss": 2.5269,
      "step": 31777
    },
    {
      "epoch": 1.7912632290024768,
      "grad_norm": 0.29191845655441284,
      "learning_rate": 0.001,
      "loss": 2.5279,
      "step": 31820
    },
    {
      "epoch": 1.7936838549876155,
      "grad_norm": 0.3864074647426605,
      "learning_rate": 0.001,
      "loss": 2.5268,
      "step": 31863
    },
    {
      "epoch": 1.7961044809727538,
      "grad_norm": 0.33010604977607727,
      "learning_rate": 0.001,
      "loss": 2.5128,
      "step": 31906
    },
    {
      "epoch": 1.7985251069578925,
      "grad_norm": 0.35602349042892456,
      "learning_rate": 0.001,
      "loss": 2.5155,
      "step": 31949
    },
    {
      "epoch": 1.8009457329430307,
      "grad_norm": 0.4098254442214966,
      "learning_rate": 0.001,
      "loss": 2.52,
      "step": 31992
    },
    {
      "epoch": 1.8011146138257148,
      "eval_ag_news_bleu_score": 3.985792890811219,
      "eval_ag_news_bleu_score_sem": 0.1508758477851492,
      "eval_ag_news_emb_cos_sim": 0.9910244941711426,
      "eval_ag_news_emb_cos_sim_sem": 0.00030440302200752066,
      "eval_ag_news_emb_top1_equal": 0.6333333253860474,
      "eval_ag_news_emb_top1_equal_sem": 0.02786867456387452,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.747101068496704,
      "eval_ag_news_n_ngrams_match_1": 4.776,
      "eval_ag_news_n_ngrams_match_2": 0.912,
      "eval_ag_news_n_ngrams_match_3": 0.248,
      "eval_ag_news_num_pred_words": 22.634,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.25911222332485956,
      "eval_ag_news_runtime": 16.4359,
      "eval_ag_news_samples_per_second": 30.421,
      "eval_ag_news_steps_per_second": 0.122,
      "eval_ag_news_token_set_f1": 0.22236527712850346,
      "eval_ag_news_token_set_f1_sem": 0.005248746412566901,
      "eval_ag_news_token_set_precision": 0.21456681196496083,
      "eval_ag_news_token_set_recall": 0.23993640236931152,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 31995
    },
    {
      "epoch": 1.8011146138257148,
      "eval_anthropic_toxic_prompts_bleu_score": 5.036127766898744,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.21611371784097988,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9913012981414795,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00026660378009355455,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6399999856948853,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02775911810844162,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.9676125049591064,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.628,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.38,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.574,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.522,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.27174696243047247,
      "eval_anthropic_toxic_prompts_runtime": 6.9231,
      "eval_anthropic_toxic_prompts_samples_per_second": 72.222,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.289,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3016061766597257,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0064638798660420224,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.36431134673867005,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2815461467258758,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 31995
    },
    {
      "epoch": 1.8011146138257148,
      "eval_arxiv_bleu_score": 3.627951396275821,
      "eval_arxiv_bleu_score_sem": 0.10950309835003853,
      "eval_arxiv_emb_cos_sim": 0.9907819032669067,
      "eval_arxiv_emb_cos_sim_sem": 0.00027500374605482627,
      "eval_arxiv_emb_top1_equal": 0.54666668176651,
      "eval_arxiv_emb_top1_equal_sem": 0.028789527267967893,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.7694060802459717,
      "eval_arxiv_n_ngrams_match_1": 4.662,
      "eval_arxiv_n_ngrams_match_2": 0.708,
      "eval_arxiv_n_ngrams_match_3": 0.13,
      "eval_arxiv_num_pred_words": 18.074,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2446655873264233,
      "eval_arxiv_runtime": 6.8181,
      "eval_arxiv_samples_per_second": 73.334,
      "eval_arxiv_steps_per_second": 0.293,
      "eval_arxiv_token_set_f1": 0.23105014637725982,
      "eval_arxiv_token_set_f1_sem": 0.004698807823619965,
      "eval_arxiv_token_set_precision": 0.19707768987472488,
      "eval_arxiv_token_set_recall": 0.29771805420242003,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 31995
    },
    {
      "epoch": 1.8011146138257148,
      "eval_python_code_alpaca_bleu_score": 5.3011266643068895,
      "eval_python_code_alpaca_bleu_score_sem": 0.19815741461927558,
      "eval_python_code_alpaca_emb_cos_sim": 0.9883567690849304,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004413172434271695,
      "eval_python_code_alpaca_emb_top1_equal": 0.6133333444595337,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028163139369651306,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.0627853870391846,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.312,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.188,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.388,
      "eval_python_code_alpaca_num_pred_words": 19.538,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.2973219434870408,
      "eval_python_code_alpaca_runtime": 6.8162,
      "eval_python_code_alpaca_samples_per_second": 73.355,
      "eval_python_code_alpaca_steps_per_second": 0.293,
      "eval_python_code_alpaca_token_set_f1": 0.32392700345244124,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005671488493048895,
      "eval_python_code_alpaca_token_set_precision": 0.3162843793512247,
      "eval_python_code_alpaca_token_set_recall": 0.36921469372732585,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 31995
    },
    {
      "epoch": 1.8011146138257148,
      "eval_wikibio_bleu_score": 3.8329318706996425,
      "eval_wikibio_bleu_score_sem": 0.13467209872683478,
      "eval_wikibio_emb_cos_sim": 0.9909757971763611,
      "eval_wikibio_emb_cos_sim_sem": 0.00025879079361587317,
      "eval_wikibio_emb_top1_equal": 0.6366666555404663,
      "eval_wikibio_emb_top1_equal_sem": 0.027814619004069896,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.011568069458008,
      "eval_wikibio_n_ngrams_match_1": 3.09,
      "eval_wikibio_n_ngrams_match_2": 0.816,
      "eval_wikibio_n_ngrams_match_3": 0.178,
      "eval_wikibio_num_pred_words": 20.048,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.255365088420089,
      "eval_wikibio_runtime": 6.7279,
      "eval_wikibio_samples_per_second": 74.317,
      "eval_wikibio_steps_per_second": 0.297,
      "eval_wikibio_token_set_f1": 0.19913352333932693,
      "eval_wikibio_token_set_f1_sem": 0.005355667983812081,
      "eval_wikibio_token_set_precision": 0.23373587758833356,
      "eval_wikibio_token_set_recall": 0.1801386724348627,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 31995
    },
    {
      "epoch": 1.8011146138257148,
      "eval_nq_5round_bleu_score": 10.500943044158078,
      "eval_nq_5round_bleu_score_sem": 0.5001956901174734,
      "eval_nq_5round_emb_cos_sim": 0.9934210181236267,
      "eval_nq_5round_emb_cos_sim_sem": 0.00019950836607517732,
      "eval_nq_5round_emb_top1_equal": 0.6166666746139526,
      "eval_nq_5round_emb_top1_equal_sem": 0.028117578599086417,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 9.372,
      "eval_nq_5round_n_ngrams_match_2": 3.326,
      "eval_nq_5round_n_ngrams_match_3": 1.514,
      "eval_nq_5round_num_pred_words": 23.566,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.37493270321336347,
      "eval_nq_5round_token_set_f1": 0.4063793511696614,
      "eval_nq_5round_token_set_f1_sem": 0.006844605943177319,
      "eval_nq_5round_token_set_precision": 0.38330962496745186,
      "eval_nq_5round_token_set_recall": 0.4426589490344935,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.967602027760408,
      "eval_nq_bleu_score_sem": 0.44030631245562535,
      "eval_nq_emb_cos_sim": 0.99321049451828,
      "eval_nq_emb_cos_sim_sem": 0.0002052281293529724,
      "eval_nq_emb_top1_equal": 0.6600000262260437,
      "eval_nq_emb_top1_equal_sem": 0.027395285786318915,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.245584011077881,
      "eval_nq_n_ngrams_match_1": 9.51,
      "eval_nq_n_ngrams_match_2": 3.252,
      "eval_nq_n_ngrams_match_3": 1.402,
      "eval_nq_num_pred_words": 23.958,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.37888777304621174,
      "eval_nq_runtime": 19.3965,
      "eval_nq_samples_per_second": 25.778,
      "eval_nq_steps_per_second": 0.103,
      "eval_nq_token_set_f1": 0.4094752203173055,
      "eval_nq_token_set_f1_sem": 0.006209353956608763,
      "eval_nq_token_set_precision": 0.3871840959258876,
      "eval_nq_token_set_recall": 0.4462311303201273,
      "eval_nq_true_num_tokens": 32.0,
      "step": 31995
    },
    {
      "epoch": 1.8033663589281694,
      "grad_norm": 0.31878602504730225,
      "learning_rate": 0.001,
      "loss": 2.525,
      "step": 32035
    },
    {
      "epoch": 1.8057869849133077,
      "grad_norm": 0.2991315722465515,
      "learning_rate": 0.001,
      "loss": 2.5302,
      "step": 32078
    },
    {
      "epoch": 1.8082076108984464,
      "grad_norm": 0.3256450891494751,
      "learning_rate": 0.001,
      "loss": 2.5206,
      "step": 32121
    },
    {
      "epoch": 1.8106282368835847,
      "grad_norm": 0.34265610575675964,
      "learning_rate": 0.001,
      "loss": 2.5149,
      "step": 32164
    },
    {
      "epoch": 1.8130488628687234,
      "grad_norm": 0.3274678587913513,
      "learning_rate": 0.001,
      "loss": 2.5255,
      "step": 32207
    },
    {
      "epoch": 1.8154694888538616,
      "grad_norm": 0.3641345202922821,
      "learning_rate": 0.001,
      "loss": 2.5114,
      "step": 32250
    },
    {
      "epoch": 1.8178901148390003,
      "grad_norm": 0.3295283615589142,
      "learning_rate": 0.001,
      "loss": 2.5197,
      "step": 32293
    },
    {
      "epoch": 1.8203107408241386,
      "grad_norm": 0.3402029573917389,
      "learning_rate": 0.001,
      "loss": 2.5232,
      "step": 32336
    },
    {
      "epoch": 1.8227313668092773,
      "grad_norm": 0.36125051975250244,
      "learning_rate": 0.001,
      "loss": 2.5166,
      "step": 32379
    },
    {
      "epoch": 1.8251519927944155,
      "grad_norm": 0.3663066625595093,
      "learning_rate": 0.001,
      "loss": 2.5128,
      "step": 32422
    },
    {
      "epoch": 1.8275726187795542,
      "grad_norm": 0.3956541419029236,
      "learning_rate": 0.001,
      "loss": 2.5111,
      "step": 32465
    },
    {
      "epoch": 1.8299932447646925,
      "grad_norm": 0.3189980387687683,
      "learning_rate": 0.001,
      "loss": 2.5193,
      "step": 32508
    },
    {
      "epoch": 1.8324138707498312,
      "grad_norm": 0.3713264763355255,
      "learning_rate": 0.001,
      "loss": 2.5201,
      "step": 32551
    },
    {
      "epoch": 1.8348344967349695,
      "grad_norm": 0.3474820554256439,
      "learning_rate": 0.001,
      "loss": 2.5107,
      "step": 32594
    },
    {
      "epoch": 1.8372551227201082,
      "grad_norm": 0.32275933027267456,
      "learning_rate": 0.001,
      "loss": 2.5128,
      "step": 32637
    },
    {
      "epoch": 1.8396757487052464,
      "grad_norm": 0.3821226954460144,
      "learning_rate": 0.001,
      "loss": 2.5089,
      "step": 32680
    },
    {
      "epoch": 1.8420963746903851,
      "grad_norm": 0.366891473531723,
      "learning_rate": 0.001,
      "loss": 2.5129,
      "step": 32723
    },
    {
      "epoch": 1.8445170006755234,
      "grad_norm": 0.37658044695854187,
      "learning_rate": 0.001,
      "loss": 2.5204,
      "step": 32766
    },
    {
      "epoch": 1.846937626660662,
      "grad_norm": 0.3629218637943268,
      "learning_rate": 0.001,
      "loss": 2.4998,
      "step": 32809
    },
    {
      "epoch": 1.8493582526458003,
      "grad_norm": 0.3707180321216583,
      "learning_rate": 0.001,
      "loss": 2.5212,
      "step": 32852
    },
    {
      "epoch": 1.851778878630939,
      "grad_norm": 0.3844511806964874,
      "learning_rate": 0.001,
      "loss": 2.5106,
      "step": 32895
    },
    {
      "epoch": 1.8541995046160773,
      "grad_norm": 0.354907363653183,
      "learning_rate": 0.001,
      "loss": 2.506,
      "step": 32938
    },
    {
      "epoch": 1.856620130601216,
      "grad_norm": 0.30533674359321594,
      "learning_rate": 0.001,
      "loss": 2.4996,
      "step": 32981
    },
    {
      "epoch": 1.8590407565863543,
      "grad_norm": 0.344473272562027,
      "learning_rate": 0.001,
      "loss": 2.5185,
      "step": 33024
    },
    {
      "epoch": 1.861461382571493,
      "grad_norm": 0.33623430132865906,
      "learning_rate": 0.001,
      "loss": 2.5168,
      "step": 33067
    },
    {
      "epoch": 1.8638820085566314,
      "grad_norm": 0.3608880937099457,
      "learning_rate": 0.001,
      "loss": 2.5166,
      "step": 33110
    },
    {
      "epoch": 1.86630263454177,
      "grad_norm": 0.32814183831214905,
      "learning_rate": 0.001,
      "loss": 2.5011,
      "step": 33153
    },
    {
      "epoch": 1.8687232605269084,
      "grad_norm": 0.3604729175567627,
      "learning_rate": 0.001,
      "loss": 2.5013,
      "step": 33196
    },
    {
      "epoch": 1.8711438865120469,
      "grad_norm": 0.3216707110404968,
      "learning_rate": 0.001,
      "loss": 2.5157,
      "step": 33239
    },
    {
      "epoch": 1.8735645124971854,
      "grad_norm": 0.3299499452114105,
      "learning_rate": 0.001,
      "loss": 2.5062,
      "step": 33282
    },
    {
      "epoch": 1.8759851384823238,
      "grad_norm": 0.3168102204799652,
      "learning_rate": 0.001,
      "loss": 2.5159,
      "step": 33325
    },
    {
      "epoch": 1.8784057644674623,
      "grad_norm": 0.36540526151657104,
      "learning_rate": 0.001,
      "loss": 2.5075,
      "step": 33368
    },
    {
      "epoch": 1.8808263904526008,
      "grad_norm": 0.3469158411026001,
      "learning_rate": 0.001,
      "loss": 2.4988,
      "step": 33411
    },
    {
      "epoch": 1.8832470164377393,
      "grad_norm": 0.38187262415885925,
      "learning_rate": 0.001,
      "loss": 2.5147,
      "step": 33454
    },
    {
      "epoch": 1.8856676424228778,
      "grad_norm": 0.31114667654037476,
      "learning_rate": 0.001,
      "loss": 2.5105,
      "step": 33497
    },
    {
      "epoch": 1.8880882684080162,
      "grad_norm": 0.37502163648605347,
      "learning_rate": 0.001,
      "loss": 2.4981,
      "step": 33540
    },
    {
      "epoch": 1.8905088943931547,
      "grad_norm": 0.3471217751502991,
      "learning_rate": 0.001,
      "loss": 2.5064,
      "step": 33583
    },
    {
      "epoch": 1.8929295203782932,
      "grad_norm": 0.3307737112045288,
      "learning_rate": 0.001,
      "loss": 2.502,
      "step": 33626
    },
    {
      "epoch": 1.8953501463634317,
      "grad_norm": 0.36846327781677246,
      "learning_rate": 0.001,
      "loss": 2.5032,
      "step": 33669
    },
    {
      "epoch": 1.8977707723485702,
      "grad_norm": 0.3516838848590851,
      "learning_rate": 0.001,
      "loss": 2.5008,
      "step": 33712
    },
    {
      "epoch": 1.9001913983337086,
      "grad_norm": 0.368330180644989,
      "learning_rate": 0.001,
      "loss": 2.508,
      "step": 33755
    },
    {
      "epoch": 1.9026120243188471,
      "grad_norm": 0.32705095410346985,
      "learning_rate": 0.001,
      "loss": 2.4985,
      "step": 33798
    },
    {
      "epoch": 1.9050326503039856,
      "grad_norm": 0.33294761180877686,
      "learning_rate": 0.001,
      "loss": 2.5077,
      "step": 33841
    },
    {
      "epoch": 1.907453276289124,
      "grad_norm": 0.3405856788158417,
      "learning_rate": 0.001,
      "loss": 2.5029,
      "step": 33884
    },
    {
      "epoch": 1.9098739022742626,
      "grad_norm": 0.3602564334869385,
      "learning_rate": 0.001,
      "loss": 2.4991,
      "step": 33927
    },
    {
      "epoch": 1.912294528259401,
      "grad_norm": 0.3845968246459961,
      "learning_rate": 0.001,
      "loss": 2.5025,
      "step": 33970
    },
    {
      "epoch": 1.9147151542445395,
      "grad_norm": 0.3993507921695709,
      "learning_rate": 0.001,
      "loss": 2.487,
      "step": 34013
    },
    {
      "epoch": 1.917135780229678,
      "grad_norm": 0.34146180748939514,
      "learning_rate": 0.001,
      "loss": 2.4986,
      "step": 34056
    },
    {
      "epoch": 1.9195564062148165,
      "grad_norm": 0.3280141055583954,
      "learning_rate": 0.001,
      "loss": 2.4972,
      "step": 34099
    },
    {
      "epoch": 1.921188921414096,
      "eval_ag_news_bleu_score": 3.8992489011597042,
      "eval_ag_news_bleu_score_sem": 0.14299762887523565,
      "eval_ag_news_emb_cos_sim": 0.9914406538009644,
      "eval_ag_news_emb_cos_sim_sem": 0.0002793587880823914,
      "eval_ag_news_emb_top1_equal": 0.6000000238418579,
      "eval_ag_news_emb_top1_equal_sem": 0.02833152959629631,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.702524423599243,
      "eval_ag_news_n_ngrams_match_1": 4.816,
      "eval_ag_news_n_ngrams_match_2": 0.88,
      "eval_ag_news_n_ngrams_match_3": 0.202,
      "eval_ag_news_num_pred_words": 23.072,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2597234848732578,
      "eval_ag_news_runtime": 15.8197,
      "eval_ag_news_samples_per_second": 31.606,
      "eval_ag_news_steps_per_second": 0.126,
      "eval_ag_news_token_set_f1": 0.22339333321333424,
      "eval_ag_news_token_set_f1_sem": 0.005231088818524288,
      "eval_ag_news_token_set_precision": 0.2163652887597767,
      "eval_ag_news_token_set_recall": 0.23873518151065684,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 34128
    },
    {
      "epoch": 1.921188921414096,
      "eval_anthropic_toxic_prompts_bleu_score": 5.354523845308751,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.23838933195825746,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9913492202758789,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00025971985741903497,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6200000047683716,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028070624111834433,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.9522225856781006,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.808,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.518,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.624,
      "eval_anthropic_toxic_prompts_num_pred_words": 25.018,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2783897475181325,
      "eval_anthropic_toxic_prompts_runtime": 7.5623,
      "eval_anthropic_toxic_prompts_samples_per_second": 66.117,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.264,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.31373057384085923,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006477324027859322,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3811509769625965,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2915704856924235,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 34128
    },
    {
      "epoch": 1.921188921414096,
      "eval_arxiv_bleu_score": 3.7341462352405306,
      "eval_arxiv_bleu_score_sem": 0.11880348033469494,
      "eval_arxiv_emb_cos_sim": 0.9902634024620056,
      "eval_arxiv_emb_cos_sim_sem": 0.0003547296637800586,
      "eval_arxiv_emb_top1_equal": 0.5600000023841858,
      "eval_arxiv_emb_top1_equal_sem": 0.028706797277804726,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.716184377670288,
      "eval_arxiv_n_ngrams_match_1": 4.862,
      "eval_arxiv_n_ngrams_match_2": 0.754,
      "eval_arxiv_n_ngrams_match_3": 0.144,
      "eval_arxiv_num_pred_words": 18.532,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2534355317064464,
      "eval_arxiv_runtime": 10.3456,
      "eval_arxiv_samples_per_second": 48.33,
      "eval_arxiv_steps_per_second": 0.193,
      "eval_arxiv_token_set_f1": 0.23921238740595432,
      "eval_arxiv_token_set_f1_sem": 0.004677069593334324,
      "eval_arxiv_token_set_precision": 0.2075816265141313,
      "eval_arxiv_token_set_recall": 0.2998876478918852,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 34128
    },
    {
      "epoch": 1.921188921414096,
      "eval_python_code_alpaca_bleu_score": 5.707442826185933,
      "eval_python_code_alpaca_bleu_score_sem": 0.21075643354046453,
      "eval_python_code_alpaca_emb_cos_sim": 0.9888718724250793,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.000422725213377151,
      "eval_python_code_alpaca_emb_top1_equal": 0.6299999952316284,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027921293391044915,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 3.0001654624938965,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.746,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.366,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.486,
      "eval_python_code_alpaca_num_pred_words": 20.54,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.31383820231357273,
      "eval_python_code_alpaca_runtime": 6.8559,
      "eval_python_code_alpaca_samples_per_second": 72.93,
      "eval_python_code_alpaca_steps_per_second": 0.292,
      "eval_python_code_alpaca_token_set_f1": 0.3379368613624758,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006009790853887427,
      "eval_python_code_alpaca_token_set_precision": 0.33818925616866835,
      "eval_python_code_alpaca_token_set_recall": 0.3635196748462497,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 34128
    },
    {
      "epoch": 1.921188921414096,
      "eval_wikibio_bleu_score": 3.7024562641586534,
      "eval_wikibio_bleu_score_sem": 0.12605794301066292,
      "eval_wikibio_emb_cos_sim": 0.9908324480056763,
      "eval_wikibio_emb_cos_sim_sem": 0.0002769243543243598,
      "eval_wikibio_emb_top1_equal": 0.653333306312561,
      "eval_wikibio_emb_top1_equal_sem": 0.027522495986455002,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.01900053024292,
      "eval_wikibio_n_ngrams_match_1": 3.076,
      "eval_wikibio_n_ngrams_match_2": 0.79,
      "eval_wikibio_n_ngrams_match_3": 0.162,
      "eval_wikibio_num_pred_words": 20.422,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2549346938359512,
      "eval_wikibio_runtime": 6.7298,
      "eval_wikibio_samples_per_second": 74.296,
      "eval_wikibio_steps_per_second": 0.297,
      "eval_wikibio_token_set_f1": 0.2010142695243562,
      "eval_wikibio_token_set_f1_sem": 0.005165468177262999,
      "eval_wikibio_token_set_precision": 0.23502051357154527,
      "eval_wikibio_token_set_recall": 0.18358248591793516,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 34128
    },
    {
      "epoch": 1.921188921414096,
      "eval_nq_5round_bleu_score": 10.556181675605872,
      "eval_nq_5round_bleu_score_sem": 0.5088798313376877,
      "eval_nq_5round_emb_cos_sim": 0.9932403564453125,
      "eval_nq_5round_emb_cos_sim_sem": 0.00023211356679772526,
      "eval_nq_5round_emb_top1_equal": 0.6333333253860474,
      "eval_nq_5round_emb_top1_equal_sem": 0.02786867456387452,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 9.678,
      "eval_nq_5round_n_ngrams_match_2": 3.352,
      "eval_nq_5round_n_ngrams_match_3": 1.57,
      "eval_nq_5round_num_pred_words": 24.188,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.384287675213531,
      "eval_nq_5round_token_set_f1": 0.41417825919941664,
      "eval_nq_5round_token_set_f1_sem": 0.006517906024122984,
      "eval_nq_5round_token_set_precision": 0.393857424566636,
      "eval_nq_5round_token_set_recall": 0.44721145639444543,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 9.863062350532772,
      "eval_nq_bleu_score_sem": 0.4481622703359639,
      "eval_nq_emb_cos_sim": 0.9932145476341248,
      "eval_nq_emb_cos_sim_sem": 0.00021032593580516502,
      "eval_nq_emb_top1_equal": 0.6833333373069763,
      "eval_nq_emb_top1_equal_sem": 0.02690183092974493,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.2175791263580322,
      "eval_nq_n_ngrams_match_1": 9.588,
      "eval_nq_n_ngrams_match_2": 3.252,
      "eval_nq_n_ngrams_match_3": 1.382,
      "eval_nq_num_pred_words": 24.268,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3785190090250887,
      "eval_nq_runtime": 19.2673,
      "eval_nq_samples_per_second": 25.951,
      "eval_nq_steps_per_second": 0.104,
      "eval_nq_token_set_f1": 0.41220417360726314,
      "eval_nq_token_set_f1_sem": 0.0060996505511145506,
      "eval_nq_token_set_precision": 0.3909332737253262,
      "eval_nq_token_set_recall": 0.44535874864154124,
      "eval_nq_true_num_tokens": 32.0,
      "step": 34128
    },
    {
      "epoch": 1.921977032199955,
      "grad_norm": 0.35673293471336365,
      "learning_rate": 0.001,
      "loss": 2.5062,
      "step": 34142
    },
    {
      "epoch": 1.9243976581850935,
      "grad_norm": 0.35093992948532104,
      "learning_rate": 0.001,
      "loss": 2.4934,
      "step": 34185
    },
    {
      "epoch": 1.926818284170232,
      "grad_norm": 0.3692317605018616,
      "learning_rate": 0.001,
      "loss": 2.5016,
      "step": 34228
    },
    {
      "epoch": 1.9292389101553704,
      "grad_norm": 0.3168604373931885,
      "learning_rate": 0.001,
      "loss": 2.4955,
      "step": 34271
    },
    {
      "epoch": 1.931659536140509,
      "grad_norm": 0.39041802287101746,
      "learning_rate": 0.001,
      "loss": 2.4993,
      "step": 34314
    },
    {
      "epoch": 1.9340801621256474,
      "grad_norm": 0.33618035912513733,
      "learning_rate": 0.001,
      "loss": 2.492,
      "step": 34357
    },
    {
      "epoch": 1.9365007881107859,
      "grad_norm": 0.38058915734291077,
      "learning_rate": 0.001,
      "loss": 2.5053,
      "step": 34400
    },
    {
      "epoch": 1.9389214140959243,
      "grad_norm": 0.33193570375442505,
      "learning_rate": 0.001,
      "loss": 2.4993,
      "step": 34443
    },
    {
      "epoch": 1.9413420400810628,
      "grad_norm": 0.36967089772224426,
      "learning_rate": 0.001,
      "loss": 2.4893,
      "step": 34486
    },
    {
      "epoch": 1.9437626660662013,
      "grad_norm": 0.3781641721725464,
      "learning_rate": 0.001,
      "loss": 2.5067,
      "step": 34529
    },
    {
      "epoch": 1.9461832920513398,
      "grad_norm": 0.3502192497253418,
      "learning_rate": 0.001,
      "loss": 2.4819,
      "step": 34572
    },
    {
      "epoch": 1.9486039180364783,
      "grad_norm": 0.3458910882472992,
      "learning_rate": 0.001,
      "loss": 2.4997,
      "step": 34615
    },
    {
      "epoch": 1.9510245440216167,
      "grad_norm": 0.3472043573856354,
      "learning_rate": 0.001,
      "loss": 2.4963,
      "step": 34658
    },
    {
      "epoch": 1.9534451700067552,
      "grad_norm": 0.3556750416755676,
      "learning_rate": 0.001,
      "loss": 2.4899,
      "step": 34701
    },
    {
      "epoch": 1.9558657959918937,
      "grad_norm": 0.3635278046131134,
      "learning_rate": 0.001,
      "loss": 2.491,
      "step": 34744
    },
    {
      "epoch": 1.9582864219770322,
      "grad_norm": 0.4354323446750641,
      "learning_rate": 0.001,
      "loss": 2.4899,
      "step": 34787
    },
    {
      "epoch": 1.9607070479621707,
      "grad_norm": 0.4155554175376892,
      "learning_rate": 0.001,
      "loss": 2.4917,
      "step": 34830
    },
    {
      "epoch": 1.9631276739473091,
      "grad_norm": 0.2993789315223694,
      "learning_rate": 0.001,
      "loss": 2.493,
      "step": 34873
    },
    {
      "epoch": 1.9655482999324476,
      "grad_norm": 0.3832486867904663,
      "learning_rate": 0.001,
      "loss": 2.4966,
      "step": 34916
    },
    {
      "epoch": 1.967968925917586,
      "grad_norm": 0.3438056707382202,
      "learning_rate": 0.001,
      "loss": 2.4897,
      "step": 34959
    },
    {
      "epoch": 1.9703895519027246,
      "grad_norm": 0.31326138973236084,
      "learning_rate": 0.001,
      "loss": 2.4927,
      "step": 35002
    },
    {
      "epoch": 1.972810177887863,
      "grad_norm": 0.3530556261539459,
      "learning_rate": 0.001,
      "loss": 2.4964,
      "step": 35045
    },
    {
      "epoch": 1.9752308038730015,
      "grad_norm": 0.3480226397514343,
      "learning_rate": 0.001,
      "loss": 2.4899,
      "step": 35088
    },
    {
      "epoch": 1.97765142985814,
      "grad_norm": 0.452766090631485,
      "learning_rate": 0.001,
      "loss": 2.4889,
      "step": 35131
    },
    {
      "epoch": 1.9800720558432785,
      "grad_norm": 0.38537096977233887,
      "learning_rate": 0.001,
      "loss": 2.4815,
      "step": 35174
    },
    {
      "epoch": 1.982492681828417,
      "grad_norm": 0.37283650040626526,
      "learning_rate": 0.001,
      "loss": 2.4854,
      "step": 35217
    },
    {
      "epoch": 1.9849133078135555,
      "grad_norm": 0.3386501967906952,
      "learning_rate": 0.001,
      "loss": 2.4841,
      "step": 35260
    },
    {
      "epoch": 1.987333933798694,
      "grad_norm": 0.3585225045681,
      "learning_rate": 0.001,
      "loss": 2.4838,
      "step": 35303
    },
    {
      "epoch": 1.9897545597838324,
      "grad_norm": 0.44612357020378113,
      "learning_rate": 0.001,
      "loss": 2.4868,
      "step": 35346
    },
    {
      "epoch": 1.992175185768971,
      "grad_norm": 0.3686203956604004,
      "learning_rate": 0.001,
      "loss": 2.4888,
      "step": 35389
    },
    {
      "epoch": 1.9945958117541094,
      "grad_norm": 0.4426267743110657,
      "learning_rate": 0.001,
      "loss": 2.4843,
      "step": 35432
    },
    {
      "epoch": 1.9970164377392479,
      "grad_norm": 0.33500829339027405,
      "learning_rate": 0.001,
      "loss": 2.4916,
      "step": 35475
    },
    {
      "epoch": 1.9994370637243866,
      "grad_norm": 0.39306217432022095,
      "learning_rate": 0.001,
      "loss": 2.492,
      "step": 35518
    },
    {
      "epoch": 2.001857689709525,
      "grad_norm": 0.37825801968574524,
      "learning_rate": 0.001,
      "loss": 2.4506,
      "step": 35561
    },
    {
      "epoch": 2.0042783156946635,
      "grad_norm": 0.36101701855659485,
      "learning_rate": 0.001,
      "loss": 2.4385,
      "step": 35604
    },
    {
      "epoch": 2.006698941679802,
      "grad_norm": 0.3814053535461426,
      "learning_rate": 0.001,
      "loss": 2.4426,
      "step": 35647
    },
    {
      "epoch": 2.0091195676649405,
      "grad_norm": 0.3472451865673065,
      "learning_rate": 0.001,
      "loss": 2.4271,
      "step": 35690
    },
    {
      "epoch": 2.0115401936500787,
      "grad_norm": 0.43625107407569885,
      "learning_rate": 0.001,
      "loss": 2.4308,
      "step": 35733
    },
    {
      "epoch": 2.0139608196352174,
      "grad_norm": 0.34636393189430237,
      "learning_rate": 0.001,
      "loss": 2.4279,
      "step": 35776
    },
    {
      "epoch": 2.0163814456203557,
      "grad_norm": 0.33633214235305786,
      "learning_rate": 0.001,
      "loss": 2.4376,
      "step": 35819
    },
    {
      "epoch": 2.0188020716054944,
      "grad_norm": 0.37691423296928406,
      "learning_rate": 0.001,
      "loss": 2.4411,
      "step": 35862
    },
    {
      "epoch": 2.0212226975906327,
      "grad_norm": 0.32358798384666443,
      "learning_rate": 0.001,
      "loss": 2.44,
      "step": 35905
    },
    {
      "epoch": 2.0236433235757714,
      "grad_norm": 0.40867629647254944,
      "learning_rate": 0.001,
      "loss": 2.4514,
      "step": 35948
    },
    {
      "epoch": 2.0260639495609096,
      "grad_norm": 0.3627355396747589,
      "learning_rate": 0.001,
      "loss": 2.4412,
      "step": 35991
    },
    {
      "epoch": 2.0284845755460483,
      "grad_norm": 0.34763094782829285,
      "learning_rate": 0.001,
      "loss": 2.4258,
      "step": 36034
    },
    {
      "epoch": 2.0309052015311866,
      "grad_norm": 0.3623834252357483,
      "learning_rate": 0.001,
      "loss": 2.4315,
      "step": 36077
    },
    {
      "epoch": 2.0333258275163253,
      "grad_norm": 0.44355446100234985,
      "learning_rate": 0.001,
      "loss": 2.4379,
      "step": 36120
    },
    {
      "epoch": 2.0357464535014635,
      "grad_norm": 0.3973741829395294,
      "learning_rate": 0.001,
      "loss": 2.4212,
      "step": 36163
    },
    {
      "epoch": 2.0381670794866023,
      "grad_norm": 0.38143348693847656,
      "learning_rate": 0.001,
      "loss": 2.4325,
      "step": 36206
    },
    {
      "epoch": 2.0405877054717405,
      "grad_norm": 0.3521195650100708,
      "learning_rate": 0.001,
      "loss": 2.4436,
      "step": 36249
    },
    {
      "epoch": 2.041263229002477,
      "eval_ag_news_bleu_score": 3.992557505343619,
      "eval_ag_news_bleu_score_sem": 0.15582622888796832,
      "eval_ag_news_emb_cos_sim": 0.9914652705192566,
      "eval_ag_news_emb_cos_sim_sem": 0.0003133167328023896,
      "eval_ag_news_emb_top1_equal": 0.6333333253860474,
      "eval_ag_news_emb_top1_equal_sem": 0.02786867456387452,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.678422451019287,
      "eval_ag_news_n_ngrams_match_1": 4.836,
      "eval_ag_news_n_ngrams_match_2": 0.93,
      "eval_ag_news_n_ngrams_match_3": 0.256,
      "eval_ag_news_num_pred_words": 22.784,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2643772846919053,
      "eval_ag_news_runtime": 17.7057,
      "eval_ag_news_samples_per_second": 28.239,
      "eval_ag_news_steps_per_second": 0.113,
      "eval_ag_news_token_set_f1": 0.22383410060698597,
      "eval_ag_news_token_set_f1_sem": 0.005289180875306968,
      "eval_ag_news_token_set_precision": 0.21712679931188542,
      "eval_ag_news_token_set_recall": 0.2370716906285363,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 36261
    },
    {
      "epoch": 2.041263229002477,
      "eval_anthropic_toxic_prompts_bleu_score": 5.295790658894808,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.22160365319160072,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9912022352218628,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00027690550796252234,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6399999856948853,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027759119829079505,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.9741389751434326,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.84,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.486,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.6,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.476,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.286137149111809,
      "eval_anthropic_toxic_prompts_runtime": 6.804,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.487,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.294,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.31033701008486275,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006291420430090267,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3815737172202749,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.28346959003261735,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 36261
    },
    {
      "epoch": 2.041263229002477,
      "eval_arxiv_bleu_score": 3.5969457101598348,
      "eval_arxiv_bleu_score_sem": 0.11214615689852925,
      "eval_arxiv_emb_cos_sim": 0.9906370043754578,
      "eval_arxiv_emb_cos_sim_sem": 0.00028622929515839906,
      "eval_arxiv_emb_top1_equal": 0.5799999833106995,
      "eval_arxiv_emb_top1_equal_sem": 0.02854322483723857,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.7542378902435303,
      "eval_arxiv_n_ngrams_match_1": 4.658,
      "eval_arxiv_n_ngrams_match_2": 0.738,
      "eval_arxiv_n_ngrams_match_3": 0.136,
      "eval_arxiv_num_pred_words": 17.792,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.24976124672141903,
      "eval_arxiv_runtime": 6.6264,
      "eval_arxiv_samples_per_second": 75.456,
      "eval_arxiv_steps_per_second": 0.302,
      "eval_arxiv_token_set_f1": 0.22970193366145,
      "eval_arxiv_token_set_f1_sem": 0.004699561661290975,
      "eval_arxiv_token_set_precision": 0.19791038060661612,
      "eval_arxiv_token_set_recall": 0.2942454517032152,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 36261
    },
    {
      "epoch": 2.041263229002477,
      "eval_python_code_alpaca_bleu_score": 5.720899671870486,
      "eval_python_code_alpaca_bleu_score_sem": 0.22699218293241083,
      "eval_python_code_alpaca_emb_cos_sim": 0.9890552759170532,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004292305688319374,
      "eval_python_code_alpaca_emb_top1_equal": 0.6066666841506958,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028250091805185003,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.9900290966033936,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.466,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.28,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.46,
      "eval_python_code_alpaca_num_pred_words": 19.632,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.31087472934641425,
      "eval_python_code_alpaca_runtime": 6.6043,
      "eval_python_code_alpaca_samples_per_second": 75.708,
      "eval_python_code_alpaca_steps_per_second": 0.303,
      "eval_python_code_alpaca_token_set_f1": 0.32829944857251414,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005621609181218405,
      "eval_python_code_alpaca_token_set_precision": 0.32317977692820155,
      "eval_python_code_alpaca_token_set_recall": 0.364346290392054,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 36261
    },
    {
      "epoch": 2.041263229002477,
      "eval_wikibio_bleu_score": 3.7876973265733525,
      "eval_wikibio_bleu_score_sem": 0.13281538052474498,
      "eval_wikibio_emb_cos_sim": 0.991131603717804,
      "eval_wikibio_emb_cos_sim_sem": 0.00028681541432304673,
      "eval_wikibio_emb_top1_equal": 0.6066666841506958,
      "eval_wikibio_emb_top1_equal_sem": 0.028250091805185003,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 4.035548210144043,
      "eval_wikibio_n_ngrams_match_1": 3.15,
      "eval_wikibio_n_ngrams_match_2": 0.822,
      "eval_wikibio_n_ngrams_match_3": 0.176,
      "eval_wikibio_num_pred_words": 20.336,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.26108348217412064,
      "eval_wikibio_runtime": 6.5529,
      "eval_wikibio_samples_per_second": 76.302,
      "eval_wikibio_steps_per_second": 0.305,
      "eval_wikibio_token_set_f1": 0.20294060892207938,
      "eval_wikibio_token_set_f1_sem": 0.005406938676403592,
      "eval_wikibio_token_set_precision": 0.23901384109324791,
      "eval_wikibio_token_set_recall": 0.18259227050502935,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 36261
    },
    {
      "epoch": 2.041263229002477,
      "eval_nq_5round_bleu_score": 10.840253670098107,
      "eval_nq_5round_bleu_score_sem": 0.5118977315306759,
      "eval_nq_5round_emb_cos_sim": 0.9929274916648865,
      "eval_nq_5round_emb_cos_sim_sem": 0.0003416832496343659,
      "eval_nq_5round_emb_top1_equal": 0.6733333468437195,
      "eval_nq_5round_emb_top1_equal_sem": 0.027122635227651973,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 9.798,
      "eval_nq_5round_n_ngrams_match_2": 3.5,
      "eval_nq_5round_n_ngrams_match_3": 1.584,
      "eval_nq_5round_num_pred_words": 23.73,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.39870461746583985,
      "eval_nq_5round_token_set_f1": 0.42339166429642877,
      "eval_nq_5round_token_set_f1_sem": 0.006743774469057885,
      "eval_nq_5round_token_set_precision": 0.40210104633234295,
      "eval_nq_5round_token_set_recall": 0.45676017516130824,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 10.011207478473164,
      "eval_nq_bleu_score_sem": 0.4511501213039603,
      "eval_nq_emb_cos_sim": 0.9929502606391907,
      "eval_nq_emb_cos_sim_sem": 0.00034649353480742796,
      "eval_nq_emb_top1_equal": 0.6566666960716248,
      "eval_nq_emb_top1_equal_sem": 0.027459642805142834,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.1911509037017822,
      "eval_nq_n_ngrams_match_1": 9.564,
      "eval_nq_n_ngrams_match_2": 3.258,
      "eval_nq_n_ngrams_match_3": 1.428,
      "eval_nq_num_pred_words": 24.076,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3825135080056513,
      "eval_nq_runtime": 18.7882,
      "eval_nq_samples_per_second": 26.612,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.4098526374565034,
      "eval_nq_token_set_f1_sem": 0.006119465979654705,
      "eval_nq_token_set_precision": 0.38902351527110474,
      "eval_nq_token_set_recall": 0.4423183039393219,
      "eval_nq_true_num_tokens": 32.0,
      "step": 36261
    },
    {
      "epoch": 2.043008331456879,
      "grad_norm": 0.3444273769855499,
      "learning_rate": 0.001,
      "loss": 2.4346,
      "step": 36292
    },
    {
      "epoch": 2.0454289574420175,
      "grad_norm": 0.341897189617157,
      "learning_rate": 0.001,
      "loss": 2.437,
      "step": 36335
    },
    {
      "epoch": 2.047849583427156,
      "grad_norm": 0.3755689859390259,
      "learning_rate": 0.001,
      "loss": 2.4294,
      "step": 36378
    },
    {
      "epoch": 2.0502702094122944,
      "grad_norm": 0.4129968285560608,
      "learning_rate": 0.001,
      "loss": 2.4415,
      "step": 36421
    },
    {
      "epoch": 2.052690835397433,
      "grad_norm": 0.37119561433792114,
      "learning_rate": 0.001,
      "loss": 2.4452,
      "step": 36464
    },
    {
      "epoch": 2.0551114613825714,
      "grad_norm": 0.31138062477111816,
      "learning_rate": 0.001,
      "loss": 2.4392,
      "step": 36507
    },
    {
      "epoch": 2.05753208736771,
      "grad_norm": 0.3494478166103363,
      "learning_rate": 0.001,
      "loss": 2.4347,
      "step": 36550
    },
    {
      "epoch": 2.0599527133528484,
      "grad_norm": 0.34699490666389465,
      "learning_rate": 0.001,
      "loss": 2.4354,
      "step": 36593
    },
    {
      "epoch": 2.062373339337987,
      "grad_norm": 0.4091903567314148,
      "learning_rate": 0.001,
      "loss": 2.4329,
      "step": 36636
    },
    {
      "epoch": 2.0647939653231253,
      "grad_norm": 0.363101989030838,
      "learning_rate": 0.001,
      "loss": 2.4271,
      "step": 36679
    },
    {
      "epoch": 2.067214591308264,
      "grad_norm": 0.3200317919254303,
      "learning_rate": 0.001,
      "loss": 2.431,
      "step": 36722
    },
    {
      "epoch": 2.0696352172934023,
      "grad_norm": 0.3649059534072876,
      "learning_rate": 0.001,
      "loss": 2.4387,
      "step": 36765
    },
    {
      "epoch": 2.072055843278541,
      "grad_norm": 0.3455341160297394,
      "learning_rate": 0.001,
      "loss": 2.4298,
      "step": 36808
    },
    {
      "epoch": 2.0744764692636792,
      "grad_norm": 0.38921815156936646,
      "learning_rate": 0.001,
      "loss": 2.4314,
      "step": 36851
    },
    {
      "epoch": 2.076897095248818,
      "grad_norm": 0.38815197348594666,
      "learning_rate": 0.001,
      "loss": 2.4417,
      "step": 36894
    },
    {
      "epoch": 2.079317721233956,
      "grad_norm": 0.4552236795425415,
      "learning_rate": 0.001,
      "loss": 2.4512,
      "step": 36937
    },
    {
      "epoch": 2.081738347219095,
      "grad_norm": 0.4007509648799896,
      "learning_rate": 0.001,
      "loss": 2.432,
      "step": 36980
    },
    {
      "epoch": 2.084158973204233,
      "grad_norm": 0.37881866097450256,
      "learning_rate": 0.001,
      "loss": 2.424,
      "step": 37023
    },
    {
      "epoch": 2.086579599189372,
      "grad_norm": 0.40189164876937866,
      "learning_rate": 0.001,
      "loss": 2.4367,
      "step": 37066
    },
    {
      "epoch": 2.08900022517451,
      "grad_norm": 0.3644411861896515,
      "learning_rate": 0.001,
      "loss": 2.431,
      "step": 37109
    },
    {
      "epoch": 2.091420851159649,
      "grad_norm": 0.4166003465652466,
      "learning_rate": 0.001,
      "loss": 2.4338,
      "step": 37152
    },
    {
      "epoch": 2.093841477144787,
      "grad_norm": 0.3362993896007538,
      "learning_rate": 0.001,
      "loss": 2.4208,
      "step": 37195
    },
    {
      "epoch": 2.096262103129926,
      "grad_norm": 0.3972605764865875,
      "learning_rate": 0.001,
      "loss": 2.436,
      "step": 37238
    },
    {
      "epoch": 2.098682729115064,
      "grad_norm": 0.3782831132411957,
      "learning_rate": 0.001,
      "loss": 2.4262,
      "step": 37281
    },
    {
      "epoch": 2.1011033551002027,
      "grad_norm": 0.3895215392112732,
      "learning_rate": 0.001,
      "loss": 2.4309,
      "step": 37324
    },
    {
      "epoch": 2.103523981085341,
      "grad_norm": 0.373237282037735,
      "learning_rate": 0.001,
      "loss": 2.4344,
      "step": 37367
    },
    {
      "epoch": 2.1059446070704797,
      "grad_norm": 0.3701364994049072,
      "learning_rate": 0.001,
      "loss": 2.4356,
      "step": 37410
    },
    {
      "epoch": 2.108365233055618,
      "grad_norm": 0.37699952721595764,
      "learning_rate": 0.001,
      "loss": 2.4225,
      "step": 37453
    },
    {
      "epoch": 2.1107858590407567,
      "grad_norm": 0.3870645761489868,
      "learning_rate": 0.001,
      "loss": 2.4214,
      "step": 37496
    },
    {
      "epoch": 2.113206485025895,
      "grad_norm": 0.3607809841632843,
      "learning_rate": 0.001,
      "loss": 2.4171,
      "step": 37539
    },
    {
      "epoch": 2.1156271110110336,
      "grad_norm": 0.4464394152164459,
      "learning_rate": 0.001,
      "loss": 2.4299,
      "step": 37582
    },
    {
      "epoch": 2.118047736996172,
      "grad_norm": 0.34369561076164246,
      "learning_rate": 0.001,
      "loss": 2.4288,
      "step": 37625
    },
    {
      "epoch": 2.1204683629813106,
      "grad_norm": 0.36715954542160034,
      "learning_rate": 0.001,
      "loss": 2.4324,
      "step": 37668
    },
    {
      "epoch": 2.122888988966449,
      "grad_norm": 0.3749127984046936,
      "learning_rate": 0.001,
      "loss": 2.4281,
      "step": 37711
    },
    {
      "epoch": 2.1253096149515875,
      "grad_norm": 0.3637578785419464,
      "learning_rate": 0.001,
      "loss": 2.4125,
      "step": 37754
    },
    {
      "epoch": 2.127730240936726,
      "grad_norm": 0.3196404278278351,
      "learning_rate": 0.001,
      "loss": 2.4275,
      "step": 37797
    },
    {
      "epoch": 2.1301508669218645,
      "grad_norm": 0.4022475481033325,
      "learning_rate": 0.001,
      "loss": 2.4391,
      "step": 37840
    },
    {
      "epoch": 2.1325714929070028,
      "grad_norm": 0.437644362449646,
      "learning_rate": 0.001,
      "loss": 2.4165,
      "step": 37883
    },
    {
      "epoch": 2.1349921188921415,
      "grad_norm": 0.3556492030620575,
      "learning_rate": 0.001,
      "loss": 2.4308,
      "step": 37926
    },
    {
      "epoch": 2.1374127448772797,
      "grad_norm": 0.3737190067768097,
      "learning_rate": 0.001,
      "loss": 2.4258,
      "step": 37969
    },
    {
      "epoch": 2.1398333708624184,
      "grad_norm": 0.3839499354362488,
      "learning_rate": 0.001,
      "loss": 2.4268,
      "step": 38012
    },
    {
      "epoch": 2.1422539968475567,
      "grad_norm": 0.32862794399261475,
      "learning_rate": 0.001,
      "loss": 2.4231,
      "step": 38055
    },
    {
      "epoch": 2.1446746228326954,
      "grad_norm": 0.36925506591796875,
      "learning_rate": 0.001,
      "loss": 2.4277,
      "step": 38098
    },
    {
      "epoch": 2.1470952488178336,
      "grad_norm": 0.38524365425109863,
      "learning_rate": 0.001,
      "loss": 2.4243,
      "step": 38141
    },
    {
      "epoch": 2.1495158748029723,
      "grad_norm": 0.4096418619155884,
      "learning_rate": 0.001,
      "loss": 2.4387,
      "step": 38184
    },
    {
      "epoch": 2.1519365007881106,
      "grad_norm": 0.32776889204978943,
      "learning_rate": 0.001,
      "loss": 2.4454,
      "step": 38227
    },
    {
      "epoch": 2.1543571267732493,
      "grad_norm": 0.29084092378616333,
      "learning_rate": 0.001,
      "loss": 2.4304,
      "step": 38270
    },
    {
      "epoch": 2.1567777527583876,
      "grad_norm": 0.35480961203575134,
      "learning_rate": 0.001,
      "loss": 2.4324,
      "step": 38313
    },
    {
      "epoch": 2.1591983787435263,
      "grad_norm": 0.3448084890842438,
      "learning_rate": 0.001,
      "loss": 2.4271,
      "step": 38356
    },
    {
      "epoch": 2.161337536590858,
      "eval_ag_news_bleu_score": 4.103042154396162,
      "eval_ag_news_bleu_score_sem": 0.16166874867703923,
      "eval_ag_news_emb_cos_sim": 0.9915177822113037,
      "eval_ag_news_emb_cos_sim_sem": 0.00029913079933218913,
      "eval_ag_news_emb_top1_equal": 0.6433333158493042,
      "eval_ag_news_emb_top1_equal_sem": 0.027702163273800266,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.6506762504577637,
      "eval_ag_news_n_ngrams_match_1": 4.952,
      "eval_ag_news_n_ngrams_match_2": 0.9,
      "eval_ag_news_n_ngrams_match_3": 0.248,
      "eval_ag_news_num_pred_words": 22.932,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.27086676714049407,
      "eval_ag_news_runtime": 14.8782,
      "eval_ag_news_samples_per_second": 33.606,
      "eval_ag_news_steps_per_second": 0.134,
      "eval_ag_news_token_set_f1": 0.2323560564533145,
      "eval_ag_news_token_set_f1_sem": 0.005213737808639133,
      "eval_ag_news_token_set_precision": 0.2256792221660771,
      "eval_ag_news_token_set_recall": 0.24671100116751105,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 38394
    },
    {
      "epoch": 2.161337536590858,
      "eval_anthropic_toxic_prompts_bleu_score": 5.33096576541611,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.22426518990930414,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9913881421089172,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00026999263017858915,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6766666769981384,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02705060760513538,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.9572482109069824,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.908,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.528,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.616,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.63,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.284613069237145,
      "eval_anthropic_toxic_prompts_runtime": 6.5948,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.817,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.303,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.31556629220406623,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006596774805622437,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.38849896743580187,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2913563792961351,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 38394
    },
    {
      "epoch": 2.161337536590858,
      "eval_arxiv_bleu_score": 3.8262989058751353,
      "eval_arxiv_bleu_score_sem": 0.12186306668763341,
      "eval_arxiv_emb_cos_sim": 0.9912503361701965,
      "eval_arxiv_emb_cos_sim_sem": 0.0002381731157304169,
      "eval_arxiv_emb_top1_equal": 0.5799999833106995,
      "eval_arxiv_emb_top1_equal_sem": 0.02854322483723857,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.714620590209961,
      "eval_arxiv_n_ngrams_match_1": 4.746,
      "eval_arxiv_n_ngrams_match_2": 0.798,
      "eval_arxiv_n_ngrams_match_3": 0.148,
      "eval_arxiv_num_pred_words": 17.978,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2565300616631933,
      "eval_arxiv_runtime": 6.6072,
      "eval_arxiv_samples_per_second": 75.675,
      "eval_arxiv_steps_per_second": 0.303,
      "eval_arxiv_token_set_f1": 0.2365091441240837,
      "eval_arxiv_token_set_f1_sem": 0.004815680440035803,
      "eval_arxiv_token_set_precision": 0.2046228465683376,
      "eval_arxiv_token_set_recall": 0.2963113850308922,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 38394
    },
    {
      "epoch": 2.161337536590858,
      "eval_python_code_alpaca_bleu_score": 5.77547903871545,
      "eval_python_code_alpaca_bleu_score_sem": 0.21626224051851847,
      "eval_python_code_alpaca_emb_cos_sim": 0.9892141222953796,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00042668784242699966,
      "eval_python_code_alpaca_emb_top1_equal": 0.6133333444595337,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028163139369651306,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.9532041549682617,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.602,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.368,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.48,
      "eval_python_code_alpaca_num_pred_words": 19.826,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3130361703483423,
      "eval_python_code_alpaca_runtime": 7.3017,
      "eval_python_code_alpaca_samples_per_second": 68.477,
      "eval_python_code_alpaca_steps_per_second": 0.274,
      "eval_python_code_alpaca_token_set_f1": 0.3331273902267871,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005861593857973737,
      "eval_python_code_alpaca_token_set_precision": 0.33216927714441197,
      "eval_python_code_alpaca_token_set_recall": 0.3633746628067197,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 38394
    },
    {
      "epoch": 2.161337536590858,
      "eval_wikibio_bleu_score": 3.9999662716410045,
      "eval_wikibio_bleu_score_sem": 0.14362453452243007,
      "eval_wikibio_emb_cos_sim": 0.9909074902534485,
      "eval_wikibio_emb_cos_sim_sem": 0.0003027933652658214,
      "eval_wikibio_emb_top1_equal": 0.6200000047683716,
      "eval_wikibio_emb_top1_equal_sem": 0.028070622391196547,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.98056697845459,
      "eval_wikibio_n_ngrams_match_1": 3.266,
      "eval_wikibio_n_ngrams_match_2": 0.88,
      "eval_wikibio_n_ngrams_match_3": 0.206,
      "eval_wikibio_num_pred_words": 20.384,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2675917401190838,
      "eval_wikibio_runtime": 6.4242,
      "eval_wikibio_samples_per_second": 77.831,
      "eval_wikibio_steps_per_second": 0.311,
      "eval_wikibio_token_set_f1": 0.21104516205478407,
      "eval_wikibio_token_set_f1_sem": 0.005429103237131639,
      "eval_wikibio_token_set_precision": 0.24816984514251844,
      "eval_wikibio_token_set_recall": 0.19043871353770028,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 38394
    },
    {
      "epoch": 2.161337536590858,
      "eval_nq_5round_bleu_score": 10.990280637608052,
      "eval_nq_5round_bleu_score_sem": 0.5191972185943845,
      "eval_nq_5round_emb_cos_sim": 0.9933412671089172,
      "eval_nq_5round_emb_cos_sim_sem": 0.0002334962068784326,
      "eval_nq_5round_emb_top1_equal": 0.6299999952316284,
      "eval_nq_5round_emb_top1_equal_sem": 0.027921293391044915,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 9.926,
      "eval_nq_5round_n_ngrams_match_2": 3.502,
      "eval_nq_5round_n_ngrams_match_3": 1.606,
      "eval_nq_5round_num_pred_words": 24.01,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4005776239441333,
      "eval_nq_5round_token_set_f1": 0.4254086855333136,
      "eval_nq_5round_token_set_f1_sem": 0.0064990914838314266,
      "eval_nq_5round_token_set_precision": 0.4054852847823009,
      "eval_nq_5round_token_set_recall": 0.4557721694995063,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 10.210609311753553,
      "eval_nq_bleu_score_sem": 0.46146853728760995,
      "eval_nq_emb_cos_sim": 0.9931657314300537,
      "eval_nq_emb_cos_sim_sem": 0.00021161969418510978,
      "eval_nq_emb_top1_equal": 0.6266666650772095,
      "eval_nq_emb_top1_equal_sem": 0.027972489250684164,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.1675965785980225,
      "eval_nq_n_ngrams_match_1": 9.652,
      "eval_nq_n_ngrams_match_2": 3.33,
      "eval_nq_n_ngrams_match_3": 1.47,
      "eval_nq_num_pred_words": 23.98,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3867904760071742,
      "eval_nq_runtime": 18.6449,
      "eval_nq_samples_per_second": 26.817,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.4159075096963495,
      "eval_nq_token_set_f1_sem": 0.006228042802210653,
      "eval_nq_token_set_precision": 0.39659163110801016,
      "eval_nq_token_set_recall": 0.4462099540088144,
      "eval_nq_true_num_tokens": 32.0,
      "step": 38394
    },
    {
      "epoch": 2.1616190047286645,
      "grad_norm": 0.3599923253059387,
      "learning_rate": 0.001,
      "loss": 2.4224,
      "step": 38399
    },
    {
      "epoch": 2.1640396307138032,
      "grad_norm": 0.3302769064903259,
      "learning_rate": 0.001,
      "loss": 2.42,
      "step": 38442
    },
    {
      "epoch": 2.1664602566989415,
      "grad_norm": 0.37822437286376953,
      "learning_rate": 0.001,
      "loss": 2.4277,
      "step": 38485
    },
    {
      "epoch": 2.16888088268408,
      "grad_norm": 0.3448626399040222,
      "learning_rate": 0.001,
      "loss": 2.4172,
      "step": 38528
    },
    {
      "epoch": 2.1713015086692184,
      "grad_norm": 0.33718636631965637,
      "learning_rate": 0.001,
      "loss": 2.4297,
      "step": 38571
    },
    {
      "epoch": 2.173722134654357,
      "grad_norm": 0.3774874210357666,
      "learning_rate": 0.001,
      "loss": 2.4338,
      "step": 38614
    },
    {
      "epoch": 2.1761427606394954,
      "grad_norm": 0.4138236343860626,
      "learning_rate": 0.001,
      "loss": 2.4158,
      "step": 38657
    },
    {
      "epoch": 2.178563386624634,
      "grad_norm": 0.37604472041130066,
      "learning_rate": 0.001,
      "loss": 2.4251,
      "step": 38700
    },
    {
      "epoch": 2.1809840126097724,
      "grad_norm": 0.3576754629611969,
      "learning_rate": 0.001,
      "loss": 2.4231,
      "step": 38743
    },
    {
      "epoch": 2.183404638594911,
      "grad_norm": 0.32193464040756226,
      "learning_rate": 0.001,
      "loss": 2.4271,
      "step": 38786
    },
    {
      "epoch": 2.1858252645800493,
      "grad_norm": 0.3594810962677002,
      "learning_rate": 0.001,
      "loss": 2.4312,
      "step": 38829
    },
    {
      "epoch": 2.188245890565188,
      "grad_norm": 0.3455040454864502,
      "learning_rate": 0.001,
      "loss": 2.4212,
      "step": 38872
    },
    {
      "epoch": 2.1906665165503263,
      "grad_norm": 0.3661624789237976,
      "learning_rate": 0.001,
      "loss": 2.4229,
      "step": 38915
    },
    {
      "epoch": 2.193087142535465,
      "grad_norm": 0.48453617095947266,
      "learning_rate": 0.001,
      "loss": 2.4197,
      "step": 38958
    },
    {
      "epoch": 2.1955077685206037,
      "grad_norm": 0.3786635398864746,
      "learning_rate": 0.001,
      "loss": 2.4225,
      "step": 39001
    },
    {
      "epoch": 2.197928394505742,
      "grad_norm": 0.3706575334072113,
      "learning_rate": 0.001,
      "loss": 2.4209,
      "step": 39044
    },
    {
      "epoch": 2.20034902049088,
      "grad_norm": 0.3632626235485077,
      "learning_rate": 0.001,
      "loss": 2.4228,
      "step": 39087
    },
    {
      "epoch": 2.202769646476019,
      "grad_norm": 0.3821455240249634,
      "learning_rate": 0.001,
      "loss": 2.4236,
      "step": 39130
    },
    {
      "epoch": 2.2051902724611576,
      "grad_norm": 0.38064947724342346,
      "learning_rate": 0.001,
      "loss": 2.4202,
      "step": 39173
    },
    {
      "epoch": 2.207610898446296,
      "grad_norm": 0.35686957836151123,
      "learning_rate": 0.001,
      "loss": 2.4231,
      "step": 39216
    },
    {
      "epoch": 2.210031524431434,
      "grad_norm": 0.3592695891857147,
      "learning_rate": 0.001,
      "loss": 2.4209,
      "step": 39259
    },
    {
      "epoch": 2.212452150416573,
      "grad_norm": 0.3529389798641205,
      "learning_rate": 0.001,
      "loss": 2.4218,
      "step": 39302
    },
    {
      "epoch": 2.2148727764017115,
      "grad_norm": 0.3462003469467163,
      "learning_rate": 0.001,
      "loss": 2.4163,
      "step": 39345
    },
    {
      "epoch": 2.21729340238685,
      "grad_norm": 0.38739100098609924,
      "learning_rate": 0.001,
      "loss": 2.4097,
      "step": 39388
    },
    {
      "epoch": 2.219714028371988,
      "grad_norm": 0.39090996980667114,
      "learning_rate": 0.001,
      "loss": 2.426,
      "step": 39431
    },
    {
      "epoch": 2.2221346543571268,
      "grad_norm": 0.3504168689250946,
      "learning_rate": 0.001,
      "loss": 2.4078,
      "step": 39474
    },
    {
      "epoch": 2.2245552803422655,
      "grad_norm": 0.33859142661094666,
      "learning_rate": 0.001,
      "loss": 2.4137,
      "step": 39517
    },
    {
      "epoch": 2.2269759063274037,
      "grad_norm": 0.39702874422073364,
      "learning_rate": 0.001,
      "loss": 2.4093,
      "step": 39560
    },
    {
      "epoch": 2.2293965323125424,
      "grad_norm": 0.4276939034461975,
      "learning_rate": 0.001,
      "loss": 2.4202,
      "step": 39603
    },
    {
      "epoch": 2.2318171582976807,
      "grad_norm": 0.35284125804901123,
      "learning_rate": 0.001,
      "loss": 2.412,
      "step": 39646
    },
    {
      "epoch": 2.2342377842828194,
      "grad_norm": 0.3699091970920563,
      "learning_rate": 0.001,
      "loss": 2.4179,
      "step": 39689
    },
    {
      "epoch": 2.2366584102679576,
      "grad_norm": 0.43949568271636963,
      "learning_rate": 0.001,
      "loss": 2.4125,
      "step": 39732
    },
    {
      "epoch": 2.2390790362530963,
      "grad_norm": 0.39731457829475403,
      "learning_rate": 0.001,
      "loss": 2.4128,
      "step": 39775
    },
    {
      "epoch": 2.2414996622382346,
      "grad_norm": 0.33715033531188965,
      "learning_rate": 0.001,
      "loss": 2.414,
      "step": 39818
    },
    {
      "epoch": 2.2439202882233733,
      "grad_norm": 0.39530423283576965,
      "learning_rate": 0.001,
      "loss": 2.4089,
      "step": 39861
    },
    {
      "epoch": 2.2463409142085116,
      "grad_norm": 0.3994970917701721,
      "learning_rate": 0.001,
      "loss": 2.4214,
      "step": 39904
    },
    {
      "epoch": 2.2487615401936503,
      "grad_norm": 0.3814813196659088,
      "learning_rate": 0.001,
      "loss": 2.4252,
      "step": 39947
    },
    {
      "epoch": 2.2511821661787885,
      "grad_norm": 0.34390074014663696,
      "learning_rate": 0.001,
      "loss": 2.417,
      "step": 39990
    },
    {
      "epoch": 2.2536027921639272,
      "grad_norm": 0.40500909090042114,
      "learning_rate": 0.001,
      "loss": 2.4162,
      "step": 40033
    },
    {
      "epoch": 2.2560234181490655,
      "grad_norm": 0.3856814205646515,
      "learning_rate": 0.001,
      "loss": 2.4102,
      "step": 40076
    },
    {
      "epoch": 2.258444044134204,
      "grad_norm": 0.3713878393173218,
      "learning_rate": 0.001,
      "loss": 2.4173,
      "step": 40119
    },
    {
      "epoch": 2.2608646701193424,
      "grad_norm": 0.3423754572868347,
      "learning_rate": 0.001,
      "loss": 2.4082,
      "step": 40162
    },
    {
      "epoch": 2.263285296104481,
      "grad_norm": 0.40311717987060547,
      "learning_rate": 0.001,
      "loss": 2.4133,
      "step": 40205
    },
    {
      "epoch": 2.2657059220896194,
      "grad_norm": 0.39802461862564087,
      "learning_rate": 0.001,
      "loss": 2.4111,
      "step": 40248
    },
    {
      "epoch": 2.268126548074758,
      "grad_norm": 0.35951799154281616,
      "learning_rate": 0.001,
      "loss": 2.4203,
      "step": 40291
    },
    {
      "epoch": 2.2705471740598964,
      "grad_norm": 0.39954039454460144,
      "learning_rate": 0.001,
      "loss": 2.4088,
      "step": 40334
    },
    {
      "epoch": 2.272967800045035,
      "grad_norm": 0.3854176700115204,
      "learning_rate": 0.001,
      "loss": 2.4073,
      "step": 40377
    },
    {
      "epoch": 2.2753884260301733,
      "grad_norm": 0.4559727609157562,
      "learning_rate": 0.001,
      "loss": 2.4161,
      "step": 40420
    },
    {
      "epoch": 2.277809052015312,
      "grad_norm": 0.4036829173564911,
      "learning_rate": 0.001,
      "loss": 2.4079,
      "step": 40463
    },
    {
      "epoch": 2.2802296780004503,
      "grad_norm": 0.37271565198898315,
      "learning_rate": 0.001,
      "loss": 2.4068,
      "step": 40506
    },
    {
      "epoch": 2.281411844179239,
      "eval_ag_news_bleu_score": 4.106287070282323,
      "eval_ag_news_bleu_score_sem": 0.1587335699204289,
      "eval_ag_news_emb_cos_sim": 0.9915873408317566,
      "eval_ag_news_emb_cos_sim_sem": 0.0002840499728518303,
      "eval_ag_news_emb_top1_equal": 0.6333333253860474,
      "eval_ag_news_emb_top1_equal_sem": 0.02786867456387452,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.642498016357422,
      "eval_ag_news_n_ngrams_match_1": 4.81,
      "eval_ag_news_n_ngrams_match_2": 0.934,
      "eval_ag_news_n_ngrams_match_3": 0.256,
      "eval_ag_news_num_pred_words": 22.616,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2649386057204274,
      "eval_ag_news_runtime": 20.4155,
      "eval_ag_news_samples_per_second": 24.491,
      "eval_ag_news_steps_per_second": 0.098,
      "eval_ag_news_token_set_f1": 0.22400420780866853,
      "eval_ag_news_token_set_f1_sem": 0.005147222947163847,
      "eval_ag_news_token_set_precision": 0.21668286051959054,
      "eval_ag_news_token_set_recall": 0.2383078421141386,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 40527
    },
    {
      "epoch": 2.281411844179239,
      "eval_anthropic_toxic_prompts_bleu_score": 5.3605530158751336,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.22406278676625654,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9914950728416443,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002740741445523253,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6033333539962769,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028291497235256893,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.948965549468994,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.894,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.498,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.59,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.298,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.29173925692866287,
      "eval_anthropic_toxic_prompts_runtime": 6.5566,
      "eval_anthropic_toxic_prompts_samples_per_second": 76.259,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.305,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3208631579965067,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006148248551279521,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3879869729808572,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.2992045219044877,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 40527
    },
    {
      "epoch": 2.281411844179239,
      "eval_arxiv_bleu_score": 3.684597522901239,
      "eval_arxiv_bleu_score_sem": 0.12081810312433174,
      "eval_arxiv_emb_cos_sim": 0.9911803603172302,
      "eval_arxiv_emb_cos_sim_sem": 0.00026654326203293246,
      "eval_arxiv_emb_top1_equal": 0.5933333039283752,
      "eval_arxiv_emb_top1_equal_sem": 0.028407504362121784,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.6896963119506836,
      "eval_arxiv_n_ngrams_match_1": 4.782,
      "eval_arxiv_n_ngrams_match_2": 0.744,
      "eval_arxiv_n_ngrams_match_3": 0.148,
      "eval_arxiv_num_pred_words": 18.082,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.25446375870006366,
      "eval_arxiv_runtime": 6.6615,
      "eval_arxiv_samples_per_second": 75.058,
      "eval_arxiv_steps_per_second": 0.3,
      "eval_arxiv_token_set_f1": 0.23630591377670895,
      "eval_arxiv_token_set_f1_sem": 0.005140382703847919,
      "eval_arxiv_token_set_precision": 0.20529669215774524,
      "eval_arxiv_token_set_recall": 0.29668678353548483,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 40527
    },
    {
      "epoch": 2.281411844179239,
      "eval_python_code_alpaca_bleu_score": 5.66787956119342,
      "eval_python_code_alpaca_bleu_score_sem": 0.20514195276456418,
      "eval_python_code_alpaca_emb_cos_sim": 0.9888928532600403,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00046976032864504174,
      "eval_python_code_alpaca_emb_top1_equal": 0.6333333253860474,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02786867456387452,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.944753885269165,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.678,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.37,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.45,
      "eval_python_code_alpaca_num_pred_words": 19.92,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3161475622717477,
      "eval_python_code_alpaca_runtime": 6.6211,
      "eval_python_code_alpaca_samples_per_second": 75.517,
      "eval_python_code_alpaca_steps_per_second": 0.302,
      "eval_python_code_alpaca_token_set_f1": 0.3380238642324391,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005967555252661301,
      "eval_python_code_alpaca_token_set_precision": 0.337471835501138,
      "eval_python_code_alpaca_token_set_recall": 0.36814366669398013,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 40527
    },
    {
      "epoch": 2.281411844179239,
      "eval_wikibio_bleu_score": 3.9025565436074086,
      "eval_wikibio_bleu_score_sem": 0.14526776252694143,
      "eval_wikibio_emb_cos_sim": 0.9912044405937195,
      "eval_wikibio_emb_cos_sim_sem": 0.0003061310263737683,
      "eval_wikibio_emb_top1_equal": 0.6299999952316284,
      "eval_wikibio_emb_top1_equal_sem": 0.027921293391044915,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.991060972213745,
      "eval_wikibio_n_ngrams_match_1": 3.194,
      "eval_wikibio_n_ngrams_match_2": 0.872,
      "eval_wikibio_n_ngrams_match_3": 0.19,
      "eval_wikibio_num_pred_words": 20.07,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2713261452954442,
      "eval_wikibio_runtime": 6.4844,
      "eval_wikibio_samples_per_second": 77.108,
      "eval_wikibio_steps_per_second": 0.308,
      "eval_wikibio_token_set_f1": 0.2074170902759044,
      "eval_wikibio_token_set_f1_sem": 0.005496964630134211,
      "eval_wikibio_token_set_precision": 0.24249639393473216,
      "eval_wikibio_token_set_recall": 0.1885522497151079,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 40527
    },
    {
      "epoch": 2.281411844179239,
      "eval_nq_5round_bleu_score": 11.67947883656196,
      "eval_nq_5round_bleu_score_sem": 0.5679376209976416,
      "eval_nq_5round_emb_cos_sim": 0.9935846328735352,
      "eval_nq_5round_emb_cos_sim_sem": 0.00023427471486663052,
      "eval_nq_5round_emb_top1_equal": 0.6499999761581421,
      "eval_nq_5round_emb_top1_equal_sem": 0.027583864257272155,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 9.99,
      "eval_nq_5round_n_ngrams_match_2": 3.724,
      "eval_nq_5round_n_ngrams_match_3": 1.77,
      "eval_nq_5round_num_pred_words": 23.652,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4066027959854136,
      "eval_nq_5round_token_set_f1": 0.4289500740919191,
      "eval_nq_5round_token_set_f1_sem": 0.006752404797608124,
      "eval_nq_5round_token_set_precision": 0.4087633270211916,
      "eval_nq_5round_token_set_recall": 0.4592936045068488,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 10.470128529348008,
      "eval_nq_bleu_score_sem": 0.45036954767361354,
      "eval_nq_emb_cos_sim": 0.9935137033462524,
      "eval_nq_emb_cos_sim_sem": 0.0002026435564969648,
      "eval_nq_emb_top1_equal": 0.6399999856948853,
      "eval_nq_emb_top1_equal_sem": 0.02775911810844162,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.1512904167175293,
      "eval_nq_n_ngrams_match_1": 9.824,
      "eval_nq_n_ngrams_match_2": 3.426,
      "eval_nq_n_ngrams_match_3": 1.548,
      "eval_nq_num_pred_words": 23.888,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.3942125788200874,
      "eval_nq_runtime": 18.9454,
      "eval_nq_samples_per_second": 26.392,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.4227395597381446,
      "eval_nq_token_set_f1_sem": 0.006168591890490001,
      "eval_nq_token_set_precision": 0.40139390171282885,
      "eval_nq_token_set_recall": 0.45548165809422786,
      "eval_nq_true_num_tokens": 32.0,
      "step": 40527
    },
    {
      "epoch": 2.282650303985589,
      "grad_norm": 0.3350335955619812,
      "learning_rate": 0.001,
      "loss": 2.4042,
      "step": 40549
    },
    {
      "epoch": 2.2850709299707272,
      "grad_norm": 0.3534840941429138,
      "learning_rate": 0.001,
      "loss": 2.424,
      "step": 40592
    },
    {
      "epoch": 2.287491555955866,
      "grad_norm": 0.34957125782966614,
      "learning_rate": 0.001,
      "loss": 2.4039,
      "step": 40635
    },
    {
      "epoch": 2.289912181941004,
      "grad_norm": 0.3435848355293274,
      "learning_rate": 0.001,
      "loss": 2.4078,
      "step": 40678
    },
    {
      "epoch": 2.292332807926143,
      "grad_norm": 0.3563764989376068,
      "learning_rate": 0.001,
      "loss": 2.4182,
      "step": 40721
    },
    {
      "epoch": 2.294753433911281,
      "grad_norm": 0.3831741511821747,
      "learning_rate": 0.001,
      "loss": 2.4099,
      "step": 40764
    },
    {
      "epoch": 2.29717405989642,
      "grad_norm": 0.3922659456729889,
      "learning_rate": 0.001,
      "loss": 2.4141,
      "step": 40807
    },
    {
      "epoch": 2.299594685881558,
      "grad_norm": 0.33624616265296936,
      "learning_rate": 0.001,
      "loss": 2.4096,
      "step": 40850
    },
    {
      "epoch": 2.302015311866697,
      "grad_norm": 0.3586716949939728,
      "learning_rate": 0.001,
      "loss": 2.4226,
      "step": 40893
    },
    {
      "epoch": 2.304435937851835,
      "grad_norm": 0.3742067217826843,
      "learning_rate": 0.001,
      "loss": 2.4222,
      "step": 40936
    },
    {
      "epoch": 2.306856563836974,
      "grad_norm": 0.3860240876674652,
      "learning_rate": 0.001,
      "loss": 2.4204,
      "step": 40979
    },
    {
      "epoch": 2.309277189822112,
      "grad_norm": 0.36864137649536133,
      "learning_rate": 0.001,
      "loss": 2.3967,
      "step": 41022
    },
    {
      "epoch": 2.3116978158072508,
      "grad_norm": 0.3558599054813385,
      "learning_rate": 0.001,
      "loss": 2.41,
      "step": 41065
    },
    {
      "epoch": 2.314118441792389,
      "grad_norm": 0.42291557788848877,
      "learning_rate": 0.001,
      "loss": 2.406,
      "step": 41108
    },
    {
      "epoch": 2.3165390677775277,
      "grad_norm": 0.41387447714805603,
      "learning_rate": 0.001,
      "loss": 2.4103,
      "step": 41151
    },
    {
      "epoch": 2.318959693762666,
      "grad_norm": 0.3869604170322418,
      "learning_rate": 0.001,
      "loss": 2.3912,
      "step": 41194
    },
    {
      "epoch": 2.3213803197478047,
      "grad_norm": 0.364279180765152,
      "learning_rate": 0.001,
      "loss": 2.4146,
      "step": 41237
    },
    {
      "epoch": 2.323800945732943,
      "grad_norm": 0.42195117473602295,
      "learning_rate": 0.001,
      "loss": 2.4117,
      "step": 41280
    },
    {
      "epoch": 2.3262215717180816,
      "grad_norm": 0.369698703289032,
      "learning_rate": 0.001,
      "loss": 2.3961,
      "step": 41323
    },
    {
      "epoch": 2.32864219770322,
      "grad_norm": 0.34724557399749756,
      "learning_rate": 0.001,
      "loss": 2.4086,
      "step": 41366
    },
    {
      "epoch": 2.3310628236883586,
      "grad_norm": 0.33751964569091797,
      "learning_rate": 0.001,
      "loss": 2.4098,
      "step": 41409
    },
    {
      "epoch": 2.333483449673497,
      "grad_norm": 0.35075289011001587,
      "learning_rate": 0.001,
      "loss": 2.3918,
      "step": 41452
    },
    {
      "epoch": 2.3359040756586356,
      "grad_norm": 0.3818661570549011,
      "learning_rate": 0.001,
      "loss": 2.4043,
      "step": 41495
    },
    {
      "epoch": 2.338324701643774,
      "grad_norm": 0.4003363847732544,
      "learning_rate": 0.001,
      "loss": 2.4014,
      "step": 41538
    },
    {
      "epoch": 2.3407453276289125,
      "grad_norm": 0.368821382522583,
      "learning_rate": 0.001,
      "loss": 2.4045,
      "step": 41581
    },
    {
      "epoch": 2.343165953614051,
      "grad_norm": 0.3649406135082245,
      "learning_rate": 0.001,
      "loss": 2.4045,
      "step": 41624
    },
    {
      "epoch": 2.3455865795991895,
      "grad_norm": 0.3854084312915802,
      "learning_rate": 0.001,
      "loss": 2.4098,
      "step": 41667
    },
    {
      "epoch": 2.3480072055843277,
      "grad_norm": 0.3940819501876831,
      "learning_rate": 0.001,
      "loss": 2.4003,
      "step": 41710
    },
    {
      "epoch": 2.3504278315694664,
      "grad_norm": 0.3675088584423065,
      "learning_rate": 0.001,
      "loss": 2.409,
      "step": 41753
    },
    {
      "epoch": 2.3528484575546047,
      "grad_norm": 0.36895254254341125,
      "learning_rate": 0.001,
      "loss": 2.4111,
      "step": 41796
    },
    {
      "epoch": 2.3552690835397434,
      "grad_norm": 0.36916065216064453,
      "learning_rate": 0.001,
      "loss": 2.4077,
      "step": 41839
    },
    {
      "epoch": 2.3576897095248817,
      "grad_norm": 0.41352593898773193,
      "learning_rate": 0.001,
      "loss": 2.3976,
      "step": 41882
    },
    {
      "epoch": 2.3601103355100204,
      "grad_norm": 0.3848950266838074,
      "learning_rate": 0.001,
      "loss": 2.4149,
      "step": 41925
    },
    {
      "epoch": 2.3625309614951586,
      "grad_norm": 0.3679824769496918,
      "learning_rate": 0.001,
      "loss": 2.3899,
      "step": 41968
    },
    {
      "epoch": 2.3649515874802973,
      "grad_norm": 0.34399837255477905,
      "learning_rate": 0.001,
      "loss": 2.4004,
      "step": 42011
    },
    {
      "epoch": 2.3673722134654356,
      "grad_norm": 0.3594101667404175,
      "learning_rate": 0.001,
      "loss": 2.3948,
      "step": 42054
    },
    {
      "epoch": 2.3697928394505743,
      "grad_norm": 0.36709678173065186,
      "learning_rate": 0.001,
      "loss": 2.3959,
      "step": 42097
    },
    {
      "epoch": 2.3722134654357125,
      "grad_norm": 0.47512298822402954,
      "learning_rate": 0.001,
      "loss": 2.3942,
      "step": 42140
    },
    {
      "epoch": 2.3746340914208512,
      "grad_norm": 0.37823736667633057,
      "learning_rate": 0.001,
      "loss": 2.3942,
      "step": 42183
    },
    {
      "epoch": 2.3770547174059895,
      "grad_norm": 0.32226452231407166,
      "learning_rate": 0.001,
      "loss": 2.3884,
      "step": 42226
    },
    {
      "epoch": 2.379475343391128,
      "grad_norm": 0.38377368450164795,
      "learning_rate": 0.001,
      "loss": 2.409,
      "step": 42269
    },
    {
      "epoch": 2.3818959693762665,
      "grad_norm": 0.35552746057510376,
      "learning_rate": 0.001,
      "loss": 2.4049,
      "step": 42312
    },
    {
      "epoch": 2.384316595361405,
      "grad_norm": 0.3489093780517578,
      "learning_rate": 0.001,
      "loss": 2.4072,
      "step": 42355
    },
    {
      "epoch": 2.3867372213465434,
      "grad_norm": 0.3539813160896301,
      "learning_rate": 0.001,
      "loss": 2.3985,
      "step": 42398
    },
    {
      "epoch": 2.389157847331682,
      "grad_norm": 0.41008514165878296,
      "learning_rate": 0.001,
      "loss": 2.403,
      "step": 42441
    },
    {
      "epoch": 2.3915784733168204,
      "grad_norm": 0.367999404668808,
      "learning_rate": 0.001,
      "loss": 2.4074,
      "step": 42484
    },
    {
      "epoch": 2.393999099301959,
      "grad_norm": 0.38965097069740295,
      "learning_rate": 0.001,
      "loss": 2.3976,
      "step": 42527
    },
    {
      "epoch": 2.3964197252870973,
      "grad_norm": 0.39069244265556335,
      "learning_rate": 0.001,
      "loss": 2.3986,
      "step": 42570
    },
    {
      "epoch": 2.398840351272236,
      "grad_norm": 0.34019941091537476,
      "learning_rate": 0.001,
      "loss": 2.4006,
      "step": 42613
    },
    {
      "epoch": 2.4012609772573743,
      "grad_norm": 0.3735220432281494,
      "learning_rate": 0.001,
      "loss": 2.3966,
      "step": 42656
    },
    {
      "epoch": 2.40148615176762,
      "eval_ag_news_bleu_score": 4.07002752107067,
      "eval_ag_news_bleu_score_sem": 0.14660254453728924,
      "eval_ag_news_emb_cos_sim": 0.991386353969574,
      "eval_ag_news_emb_cos_sim_sem": 0.0003047045906818867,
      "eval_ag_news_emb_top1_equal": 0.6433333158493042,
      "eval_ag_news_emb_top1_equal_sem": 0.02770216499443815,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.6151061058044434,
      "eval_ag_news_n_ngrams_match_1": 4.966,
      "eval_ag_news_n_ngrams_match_2": 0.92,
      "eval_ag_news_n_ngrams_match_3": 0.244,
      "eval_ag_news_num_pred_words": 22.746,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2688646690399838,
      "eval_ag_news_runtime": 15.3499,
      "eval_ag_news_samples_per_second": 32.574,
      "eval_ag_news_steps_per_second": 0.13,
      "eval_ag_news_token_set_f1": 0.23179312179679065,
      "eval_ag_news_token_set_f1_sem": 0.00529379663774587,
      "eval_ag_news_token_set_precision": 0.2256537407063932,
      "eval_ag_news_token_set_recall": 0.24613772545009327,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 42660
    },
    {
      "epoch": 2.40148615176762,
      "eval_anthropic_toxic_prompts_bleu_score": 5.458553654383775,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.23717582508402538,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9915730953216553,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002708735161209625,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6133333444595337,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028163139369651306,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.9150447845458984,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.902,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.496,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.616,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.45,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2871299776795447,
      "eval_anthropic_toxic_prompts_runtime": 6.8304,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.202,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.293,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.31759957823920587,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006348523347267382,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.3886459311806911,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.29292699658861066,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 42660
    },
    {
      "epoch": 2.40148615176762,
      "eval_arxiv_bleu_score": 3.9305225983428698,
      "eval_arxiv_bleu_score_sem": 0.12126305614536304,
      "eval_arxiv_emb_cos_sim": 0.990936279296875,
      "eval_arxiv_emb_cos_sim_sem": 0.00027007401097357055,
      "eval_arxiv_emb_top1_equal": 0.5666666626930237,
      "eval_arxiv_emb_top1_equal_sem": 0.028657564665992605,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.692587375640869,
      "eval_arxiv_n_ngrams_match_1": 4.924,
      "eval_arxiv_n_ngrams_match_2": 0.828,
      "eval_arxiv_n_ngrams_match_3": 0.176,
      "eval_arxiv_num_pred_words": 18.774,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.26105833491375924,
      "eval_arxiv_runtime": 6.9367,
      "eval_arxiv_samples_per_second": 72.081,
      "eval_arxiv_steps_per_second": 0.288,
      "eval_arxiv_token_set_f1": 0.24001167536210724,
      "eval_arxiv_token_set_f1_sem": 0.004698335939760652,
      "eval_arxiv_token_set_precision": 0.2111319712308602,
      "eval_arxiv_token_set_recall": 0.2923227268920194,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 42660
    },
    {
      "epoch": 2.40148615176762,
      "eval_python_code_alpaca_bleu_score": 6.064004159256193,
      "eval_python_code_alpaca_bleu_score_sem": 0.21940232530594245,
      "eval_python_code_alpaca_emb_cos_sim": 0.9901056289672852,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003574315760743537,
      "eval_python_code_alpaca_emb_top1_equal": 0.6366666555404663,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027814619004069896,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.927064895629883,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.028,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.518,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.554,
      "eval_python_code_alpaca_num_pred_words": 21.248,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.32585397271575656,
      "eval_python_code_alpaca_runtime": 6.8401,
      "eval_python_code_alpaca_samples_per_second": 73.098,
      "eval_python_code_alpaca_steps_per_second": 0.292,
      "eval_python_code_alpaca_token_set_f1": 0.35467147744804917,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005868471722637684,
      "eval_python_code_alpaca_token_set_precision": 0.36297136082318615,
      "eval_python_code_alpaca_token_set_recall": 0.3768767982301873,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 42660
    },
    {
      "epoch": 2.40148615176762,
      "eval_wikibio_bleu_score": 3.82273580910339,
      "eval_wikibio_bleu_score_sem": 0.13843454088503446,
      "eval_wikibio_emb_cos_sim": 0.9910023212432861,
      "eval_wikibio_emb_cos_sim_sem": 0.00033251676638015946,
      "eval_wikibio_emb_top1_equal": 0.6166666746139526,
      "eval_wikibio_emb_top1_equal_sem": 0.028117578599086417,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9164671897888184,
      "eval_wikibio_n_ngrams_match_1": 3.198,
      "eval_wikibio_n_ngrams_match_2": 0.84,
      "eval_wikibio_n_ngrams_match_3": 0.18,
      "eval_wikibio_num_pred_words": 20.35,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.27227699118421866,
      "eval_wikibio_runtime": 6.7317,
      "eval_wikibio_samples_per_second": 74.276,
      "eval_wikibio_steps_per_second": 0.297,
      "eval_wikibio_token_set_f1": 0.20604733889541263,
      "eval_wikibio_token_set_f1_sem": 0.005539197831628905,
      "eval_wikibio_token_set_precision": 0.24166729538450382,
      "eval_wikibio_token_set_recall": 0.18566739231939625,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 42660
    },
    {
      "epoch": 2.40148615176762,
      "eval_nq_5round_bleu_score": 11.997194270804478,
      "eval_nq_5round_bleu_score_sem": 0.5714087586453318,
      "eval_nq_5round_emb_cos_sim": 0.9935112595558167,
      "eval_nq_5round_emb_cos_sim_sem": 0.00024132282403436158,
      "eval_nq_5round_emb_top1_equal": 0.6566666960716248,
      "eval_nq_5round_emb_top1_equal_sem": 0.027459642805142834,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.128,
      "eval_nq_5round_n_ngrams_match_2": 3.774,
      "eval_nq_5round_n_ngrams_match_3": 1.86,
      "eval_nq_5round_num_pred_words": 23.924,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.409997747100111,
      "eval_nq_5round_token_set_f1": 0.435823175438589,
      "eval_nq_5round_token_set_f1_sem": 0.006911327476742612,
      "eval_nq_5round_token_set_precision": 0.41556700400807967,
      "eval_nq_5round_token_set_recall": 0.4684333228956178,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 10.925161597870863,
      "eval_nq_bleu_score_sem": 0.49169583916002335,
      "eval_nq_emb_cos_sim": 0.9932632446289062,
      "eval_nq_emb_cos_sim_sem": 0.00023012192844548912,
      "eval_nq_emb_top1_equal": 0.6600000262260437,
      "eval_nq_emb_top1_equal_sem": 0.027395285786318915,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.1340603828430176,
      "eval_nq_n_ngrams_match_1": 9.916,
      "eval_nq_n_ngrams_match_2": 3.516,
      "eval_nq_n_ngrams_match_3": 1.62,
      "eval_nq_num_pred_words": 24.222,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.397518515751206,
      "eval_nq_runtime": 19.2062,
      "eval_nq_samples_per_second": 26.033,
      "eval_nq_steps_per_second": 0.104,
      "eval_nq_token_set_f1": 0.42713450270759823,
      "eval_nq_token_set_f1_sem": 0.006346912814575102,
      "eval_nq_token_set_precision": 0.40590097393371677,
      "eval_nq_token_set_recall": 0.4624077233269119,
      "eval_nq_true_num_tokens": 32.0,
      "step": 42660
    },
    {
      "epoch": 2.403681603242513,
      "grad_norm": 0.37970879673957825,
      "learning_rate": 0.001,
      "loss": 2.3893,
      "step": 42699
    },
    {
      "epoch": 2.4061022292276513,
      "grad_norm": 0.3726083040237427,
      "learning_rate": 0.001,
      "loss": 2.3965,
      "step": 42742
    },
    {
      "epoch": 2.40852285521279,
      "grad_norm": 0.3693784177303314,
      "learning_rate": 0.001,
      "loss": 2.3945,
      "step": 42785
    },
    {
      "epoch": 2.4109434811979282,
      "grad_norm": 0.37905552983283997,
      "learning_rate": 0.001,
      "loss": 2.3923,
      "step": 42828
    },
    {
      "epoch": 2.413364107183067,
      "grad_norm": 0.42317068576812744,
      "learning_rate": 0.001,
      "loss": 2.3838,
      "step": 42871
    },
    {
      "epoch": 2.415784733168205,
      "grad_norm": 0.3850644528865814,
      "learning_rate": 0.001,
      "loss": 2.3971,
      "step": 42914
    },
    {
      "epoch": 2.418205359153344,
      "grad_norm": 0.39900922775268555,
      "learning_rate": 0.001,
      "loss": 2.397,
      "step": 42957
    },
    {
      "epoch": 2.420625985138482,
      "grad_norm": 0.434496134519577,
      "learning_rate": 0.001,
      "loss": 2.3928,
      "step": 43000
    },
    {
      "epoch": 2.423046611123621,
      "grad_norm": 0.35049858689308167,
      "learning_rate": 0.001,
      "loss": 2.3923,
      "step": 43043
    },
    {
      "epoch": 2.425467237108759,
      "grad_norm": 0.39952364563941956,
      "learning_rate": 0.001,
      "loss": 2.4006,
      "step": 43086
    },
    {
      "epoch": 2.427887863093898,
      "grad_norm": 0.40778404474258423,
      "learning_rate": 0.001,
      "loss": 2.3942,
      "step": 43129
    },
    {
      "epoch": 2.430308489079036,
      "grad_norm": 0.3824518918991089,
      "learning_rate": 0.001,
      "loss": 2.3847,
      "step": 43172
    },
    {
      "epoch": 2.4327291150641748,
      "grad_norm": 0.3447646200656891,
      "learning_rate": 0.001,
      "loss": 2.3951,
      "step": 43215
    },
    {
      "epoch": 2.435149741049313,
      "grad_norm": 0.39206811785697937,
      "learning_rate": 0.001,
      "loss": 2.3955,
      "step": 43258
    },
    {
      "epoch": 2.4375703670344517,
      "grad_norm": 0.44096142053604126,
      "learning_rate": 0.001,
      "loss": 2.3993,
      "step": 43301
    },
    {
      "epoch": 2.43999099301959,
      "grad_norm": 0.3559003472328186,
      "learning_rate": 0.001,
      "loss": 2.397,
      "step": 43344
    },
    {
      "epoch": 2.4424116190047287,
      "grad_norm": 0.3267379701137543,
      "learning_rate": 0.001,
      "loss": 2.3811,
      "step": 43387
    },
    {
      "epoch": 2.444832244989867,
      "grad_norm": 0.3728923499584198,
      "learning_rate": 0.001,
      "loss": 2.3894,
      "step": 43430
    },
    {
      "epoch": 2.4472528709750057,
      "grad_norm": 0.3699650466442108,
      "learning_rate": 0.001,
      "loss": 2.3961,
      "step": 43473
    },
    {
      "epoch": 2.449673496960144,
      "grad_norm": 0.39126962423324585,
      "learning_rate": 0.001,
      "loss": 2.391,
      "step": 43516
    },
    {
      "epoch": 2.4520941229452826,
      "grad_norm": 0.37535351514816284,
      "learning_rate": 0.001,
      "loss": 2.4016,
      "step": 43559
    },
    {
      "epoch": 2.4545147489304213,
      "grad_norm": 0.37305498123168945,
      "learning_rate": 0.001,
      "loss": 2.3863,
      "step": 43602
    },
    {
      "epoch": 2.4569353749155596,
      "grad_norm": 0.3785671889781952,
      "learning_rate": 0.001,
      "loss": 2.3863,
      "step": 43645
    },
    {
      "epoch": 2.459356000900698,
      "grad_norm": 0.4318099021911621,
      "learning_rate": 0.001,
      "loss": 2.3954,
      "step": 43688
    },
    {
      "epoch": 2.4617766268858365,
      "grad_norm": 0.32604125142097473,
      "learning_rate": 0.001,
      "loss": 2.3906,
      "step": 43731
    },
    {
      "epoch": 2.4641972528709752,
      "grad_norm": 0.339754581451416,
      "learning_rate": 0.001,
      "loss": 2.3901,
      "step": 43774
    },
    {
      "epoch": 2.4666178788561135,
      "grad_norm": 0.39620405435562134,
      "learning_rate": 0.001,
      "loss": 2.3891,
      "step": 43817
    },
    {
      "epoch": 2.4690385048412518,
      "grad_norm": 0.3856825828552246,
      "learning_rate": 0.001,
      "loss": 2.3875,
      "step": 43860
    },
    {
      "epoch": 2.4714591308263905,
      "grad_norm": 0.34739282727241516,
      "learning_rate": 0.001,
      "loss": 2.3944,
      "step": 43903
    },
    {
      "epoch": 2.473879756811529,
      "grad_norm": 0.3379807770252228,
      "learning_rate": 0.001,
      "loss": 2.3977,
      "step": 43946
    },
    {
      "epoch": 2.4763003827966674,
      "grad_norm": 0.33213287591934204,
      "learning_rate": 0.001,
      "loss": 2.3869,
      "step": 43989
    },
    {
      "epoch": 2.4787210087818057,
      "grad_norm": 0.3742017447948456,
      "learning_rate": 0.001,
      "loss": 2.3931,
      "step": 44032
    },
    {
      "epoch": 2.4811416347669444,
      "grad_norm": 0.3591045141220093,
      "learning_rate": 0.001,
      "loss": 2.3932,
      "step": 44075
    },
    {
      "epoch": 2.483562260752083,
      "grad_norm": 0.3915739953517914,
      "learning_rate": 0.001,
      "loss": 2.3862,
      "step": 44118
    },
    {
      "epoch": 2.4859828867372213,
      "grad_norm": 0.385274201631546,
      "learning_rate": 0.001,
      "loss": 2.3812,
      "step": 44161
    },
    {
      "epoch": 2.4884035127223596,
      "grad_norm": 0.4219292998313904,
      "learning_rate": 0.001,
      "loss": 2.393,
      "step": 44204
    },
    {
      "epoch": 2.4908241387074983,
      "grad_norm": 0.33788347244262695,
      "learning_rate": 0.001,
      "loss": 2.387,
      "step": 44247
    },
    {
      "epoch": 2.493244764692637,
      "grad_norm": 0.37386953830718994,
      "learning_rate": 0.001,
      "loss": 2.3721,
      "step": 44290
    },
    {
      "epoch": 2.4956653906777753,
      "grad_norm": 0.42849719524383545,
      "learning_rate": 0.001,
      "loss": 2.4006,
      "step": 44333
    },
    {
      "epoch": 2.4980860166629135,
      "grad_norm": 0.3691047728061676,
      "learning_rate": 0.001,
      "loss": 2.3887,
      "step": 44376
    },
    {
      "epoch": 2.5005066426480522,
      "grad_norm": 0.3592837154865265,
      "learning_rate": 0.001,
      "loss": 2.3937,
      "step": 44419
    },
    {
      "epoch": 2.502927268633191,
      "grad_norm": 0.34035637974739075,
      "learning_rate": 0.001,
      "loss": 2.3885,
      "step": 44462
    },
    {
      "epoch": 2.505347894618329,
      "grad_norm": 0.3920776844024658,
      "learning_rate": 0.001,
      "loss": 2.3835,
      "step": 44505
    },
    {
      "epoch": 2.5077685206034674,
      "grad_norm": 0.41436436772346497,
      "learning_rate": 0.001,
      "loss": 2.3856,
      "step": 44548
    },
    {
      "epoch": 2.510189146588606,
      "grad_norm": 0.3250133693218231,
      "learning_rate": 0.001,
      "loss": 2.3942,
      "step": 44591
    },
    {
      "epoch": 2.512609772573745,
      "grad_norm": 0.40845218300819397,
      "learning_rate": 0.001,
      "loss": 2.383,
      "step": 44634
    },
    {
      "epoch": 2.515030398558883,
      "grad_norm": 0.44528019428253174,
      "learning_rate": 0.001,
      "loss": 2.3899,
      "step": 44677
    },
    {
      "epoch": 2.5174510245440214,
      "grad_norm": 0.3726034164428711,
      "learning_rate": 0.001,
      "loss": 2.391,
      "step": 44720
    },
    {
      "epoch": 2.51987165052916,
      "grad_norm": 0.33024823665618896,
      "learning_rate": 0.001,
      "loss": 2.3877,
      "step": 44763
    },
    {
      "epoch": 2.521560459356001,
      "eval_ag_news_bleu_score": 4.121296346088122,
      "eval_ag_news_bleu_score_sem": 0.15677807105832658,
      "eval_ag_news_emb_cos_sim": 0.991741955280304,
      "eval_ag_news_emb_cos_sim_sem": 0.000278876418005234,
      "eval_ag_news_emb_top1_equal": 0.5899999737739563,
      "eval_ag_news_emb_top1_equal_sem": 0.028443455370097265,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.6560404300689697,
      "eval_ag_news_n_ngrams_match_1": 4.878,
      "eval_ag_news_n_ngrams_match_2": 0.906,
      "eval_ag_news_n_ngrams_match_3": 0.236,
      "eval_ag_news_num_pred_words": 22.512,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.27144123141356896,
      "eval_ag_news_runtime": 15.193,
      "eval_ag_news_samples_per_second": 32.91,
      "eval_ag_news_steps_per_second": 0.132,
      "eval_ag_news_token_set_f1": 0.23003297947064139,
      "eval_ag_news_token_set_f1_sem": 0.005176629106296154,
      "eval_ag_news_token_set_precision": 0.2223230901026999,
      "eval_ag_news_token_set_recall": 0.245361088398074,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 44793
    },
    {
      "epoch": 2.521560459356001,
      "eval_anthropic_toxic_prompts_bleu_score": 5.489880546467526,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.23382521434878303,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9916748404502869,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00027944264229433287,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6566666960716248,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027459642805142834,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.91304874420166,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 4.99,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.556,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.65,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.16,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.29538682556400797,
      "eval_anthropic_toxic_prompts_runtime": 6.7045,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.577,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.298,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.32782055006248906,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006405459981811513,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.39992694041619403,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3049421144221239,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 44793
    },
    {
      "epoch": 2.521560459356001,
      "eval_arxiv_bleu_score": 3.6796225596702468,
      "eval_arxiv_bleu_score_sem": 0.11596687104162863,
      "eval_arxiv_emb_cos_sim": 0.9912837743759155,
      "eval_arxiv_emb_cos_sim_sem": 0.0002443915548178293,
      "eval_arxiv_emb_top1_equal": 0.6100000143051147,
      "eval_arxiv_emb_top1_equal_sem": 0.028207306423529106,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.6923975944519043,
      "eval_arxiv_n_ngrams_match_1": 4.748,
      "eval_arxiv_n_ngrams_match_2": 0.704,
      "eval_arxiv_n_ngrams_match_3": 0.136,
      "eval_arxiv_num_pred_words": 17.682,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2592397228558524,
      "eval_arxiv_runtime": 6.6539,
      "eval_arxiv_samples_per_second": 75.144,
      "eval_arxiv_steps_per_second": 0.301,
      "eval_arxiv_token_set_f1": 0.23809695227747008,
      "eval_arxiv_token_set_f1_sem": 0.004674420714300239,
      "eval_arxiv_token_set_precision": 0.20565650325900914,
      "eval_arxiv_token_set_recall": 0.2998440561106535,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 44793
    },
    {
      "epoch": 2.521560459356001,
      "eval_python_code_alpaca_bleu_score": 5.768912738621095,
      "eval_python_code_alpaca_bleu_score_sem": 0.20675252596033422,
      "eval_python_code_alpaca_emb_cos_sim": 0.9898436665534973,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003328216687904282,
      "eval_python_code_alpaca_emb_top1_equal": 0.6299999952316284,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027921293391044915,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.944685220718384,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.514,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.326,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.45,
      "eval_python_code_alpaca_num_pred_words": 19.396,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3173194893999313,
      "eval_python_code_alpaca_runtime": 6.5125,
      "eval_python_code_alpaca_samples_per_second": 76.776,
      "eval_python_code_alpaca_steps_per_second": 0.307,
      "eval_python_code_alpaca_token_set_f1": 0.33349313894298194,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005868766431952118,
      "eval_python_code_alpaca_token_set_precision": 0.32857645613147596,
      "eval_python_code_alpaca_token_set_recall": 0.3701822368307127,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 44793
    },
    {
      "epoch": 2.521560459356001,
      "eval_wikibio_bleu_score": 3.856627088093524,
      "eval_wikibio_bleu_score_sem": 0.1305461353661074,
      "eval_wikibio_emb_cos_sim": 0.9910803437232971,
      "eval_wikibio_emb_cos_sim_sem": 0.0002578153801296761,
      "eval_wikibio_emb_top1_equal": 0.6333333253860474,
      "eval_wikibio_emb_top1_equal_sem": 0.02786867456387452,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9636528491973877,
      "eval_wikibio_n_ngrams_match_1": 3.238,
      "eval_wikibio_n_ngrams_match_2": 0.864,
      "eval_wikibio_n_ngrams_match_3": 0.172,
      "eval_wikibio_num_pred_words": 20.268,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2732318132009849,
      "eval_wikibio_runtime": 6.429,
      "eval_wikibio_samples_per_second": 77.772,
      "eval_wikibio_steps_per_second": 0.311,
      "eval_wikibio_token_set_f1": 0.20811394785717136,
      "eval_wikibio_token_set_f1_sem": 0.005450500435590213,
      "eval_wikibio_token_set_precision": 0.2442823264031254,
      "eval_wikibio_token_set_recall": 0.18821275076555974,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 44793
    },
    {
      "epoch": 2.521560459356001,
      "eval_nq_5round_bleu_score": 11.922144650969727,
      "eval_nq_5round_bleu_score_sem": 0.5470092576081583,
      "eval_nq_5round_emb_cos_sim": 0.9937997460365295,
      "eval_nq_5round_emb_cos_sim_sem": 0.00023367595976751265,
      "eval_nq_5round_emb_top1_equal": 0.6399999856948853,
      "eval_nq_5round_emb_top1_equal_sem": 0.02775911810844162,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.206,
      "eval_nq_5round_n_ngrams_match_2": 3.852,
      "eval_nq_5round_n_ngrams_match_3": 1.85,
      "eval_nq_5round_num_pred_words": 23.68,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.41462694841018105,
      "eval_nq_5round_token_set_f1": 0.4407076933793558,
      "eval_nq_5round_token_set_f1_sem": 0.006650065888266262,
      "eval_nq_5round_token_set_precision": 0.4192541797226905,
      "eval_nq_5round_token_set_recall": 0.47529776096285675,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 10.86719157051471,
      "eval_nq_bleu_score_sem": 0.49527901470718705,
      "eval_nq_emb_cos_sim": 0.9937382340431213,
      "eval_nq_emb_cos_sim_sem": 0.00020761107839927976,
      "eval_nq_emb_top1_equal": 0.7200000286102295,
      "eval_nq_emb_top1_equal_sem": 0.025966275374921653,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.1084303855895996,
      "eval_nq_n_ngrams_match_1": 10.046,
      "eval_nq_n_ngrams_match_2": 3.584,
      "eval_nq_n_ngrams_match_3": 1.606,
      "eval_nq_num_pred_words": 23.85,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4034820323160959,
      "eval_nq_runtime": 19.2306,
      "eval_nq_samples_per_second": 26.0,
      "eval_nq_steps_per_second": 0.104,
      "eval_nq_token_set_f1": 0.4343228430246876,
      "eval_nq_token_set_f1_sem": 0.006096889949596072,
      "eval_nq_token_set_precision": 0.41275834288975,
      "eval_nq_token_set_recall": 0.46883261013590954,
      "eval_nq_true_num_tokens": 32.0,
      "step": 44793
    },
    {
      "epoch": 2.5222922765142988,
      "grad_norm": 0.3470325767993927,
      "learning_rate": 0.001,
      "loss": 2.3842,
      "step": 44806
    },
    {
      "epoch": 2.524712902499437,
      "grad_norm": 0.3789697289466858,
      "learning_rate": 0.001,
      "loss": 2.3786,
      "step": 44849
    },
    {
      "epoch": 2.5271335284845753,
      "grad_norm": 0.3629646301269531,
      "learning_rate": 0.001,
      "loss": 2.3822,
      "step": 44892
    },
    {
      "epoch": 2.529554154469714,
      "grad_norm": 0.4172182083129883,
      "learning_rate": 0.001,
      "loss": 2.3804,
      "step": 44935
    },
    {
      "epoch": 2.5319747804548527,
      "grad_norm": 0.3804088532924652,
      "learning_rate": 0.001,
      "loss": 2.367,
      "step": 44978
    },
    {
      "epoch": 2.534395406439991,
      "grad_norm": 0.3911835551261902,
      "learning_rate": 0.001,
      "loss": 2.3804,
      "step": 45021
    },
    {
      "epoch": 2.536816032425129,
      "grad_norm": 0.4093911945819855,
      "learning_rate": 0.001,
      "loss": 2.3803,
      "step": 45064
    },
    {
      "epoch": 2.539236658410268,
      "grad_norm": 0.35848113894462585,
      "learning_rate": 0.001,
      "loss": 2.3867,
      "step": 45107
    },
    {
      "epoch": 2.5416572843954066,
      "grad_norm": 0.38880595564842224,
      "learning_rate": 0.001,
      "loss": 2.3959,
      "step": 45150
    },
    {
      "epoch": 2.544077910380545,
      "grad_norm": 0.383248507976532,
      "learning_rate": 0.001,
      "loss": 2.3822,
      "step": 45193
    },
    {
      "epoch": 2.546498536365683,
      "grad_norm": 0.4003937244415283,
      "learning_rate": 0.001,
      "loss": 2.3758,
      "step": 45236
    },
    {
      "epoch": 2.548919162350822,
      "grad_norm": 0.35542163252830505,
      "learning_rate": 0.001,
      "loss": 2.3858,
      "step": 45279
    },
    {
      "epoch": 2.5513397883359605,
      "grad_norm": 0.37692761421203613,
      "learning_rate": 0.001,
      "loss": 2.3747,
      "step": 45322
    },
    {
      "epoch": 2.553760414321099,
      "grad_norm": 0.3633362948894501,
      "learning_rate": 0.001,
      "loss": 2.3844,
      "step": 45365
    },
    {
      "epoch": 2.5561810403062375,
      "grad_norm": 0.35952213406562805,
      "learning_rate": 0.001,
      "loss": 2.3819,
      "step": 45408
    },
    {
      "epoch": 2.5586016662913758,
      "grad_norm": 0.3664102256298065,
      "learning_rate": 0.001,
      "loss": 2.395,
      "step": 45451
    },
    {
      "epoch": 2.5610222922765145,
      "grad_norm": 0.3441017270088196,
      "learning_rate": 0.001,
      "loss": 2.3798,
      "step": 45494
    },
    {
      "epoch": 2.5634429182616527,
      "grad_norm": 0.39890629053115845,
      "learning_rate": 0.001,
      "loss": 2.38,
      "step": 45537
    },
    {
      "epoch": 2.5658635442467914,
      "grad_norm": 0.33530572056770325,
      "learning_rate": 0.001,
      "loss": 2.3766,
      "step": 45580
    },
    {
      "epoch": 2.5682841702319297,
      "grad_norm": 0.3830295503139496,
      "learning_rate": 0.001,
      "loss": 2.3744,
      "step": 45623
    },
    {
      "epoch": 2.5707047962170684,
      "grad_norm": 0.3401520848274231,
      "learning_rate": 0.001,
      "loss": 2.381,
      "step": 45666
    },
    {
      "epoch": 2.5731254222022066,
      "grad_norm": 0.440909743309021,
      "learning_rate": 0.001,
      "loss": 2.3723,
      "step": 45709
    },
    {
      "epoch": 2.5755460481873453,
      "grad_norm": 0.3871539831161499,
      "learning_rate": 0.001,
      "loss": 2.3696,
      "step": 45752
    },
    {
      "epoch": 2.5779666741724836,
      "grad_norm": 0.3366640508174896,
      "learning_rate": 0.001,
      "loss": 2.3925,
      "step": 45795
    },
    {
      "epoch": 2.5803873001576223,
      "grad_norm": 0.3579271733760834,
      "learning_rate": 0.001,
      "loss": 2.3877,
      "step": 45838
    },
    {
      "epoch": 2.5828079261427606,
      "grad_norm": 0.3817802667617798,
      "learning_rate": 0.001,
      "loss": 2.3754,
      "step": 45881
    },
    {
      "epoch": 2.5852285521278993,
      "grad_norm": 0.4094647467136383,
      "learning_rate": 0.001,
      "loss": 2.3769,
      "step": 45924
    },
    {
      "epoch": 2.5876491781130375,
      "grad_norm": 0.3934791088104248,
      "learning_rate": 0.001,
      "loss": 2.3719,
      "step": 45967
    },
    {
      "epoch": 2.590069804098176,
      "grad_norm": 0.397417813539505,
      "learning_rate": 0.001,
      "loss": 2.3705,
      "step": 46010
    },
    {
      "epoch": 2.5924904300833145,
      "grad_norm": 0.44781187176704407,
      "learning_rate": 0.001,
      "loss": 2.3642,
      "step": 46053
    },
    {
      "epoch": 2.594911056068453,
      "grad_norm": 0.3281353712081909,
      "learning_rate": 0.001,
      "loss": 2.3777,
      "step": 46096
    },
    {
      "epoch": 2.5973316820535914,
      "grad_norm": 0.4140641689300537,
      "learning_rate": 0.001,
      "loss": 2.368,
      "step": 46139
    },
    {
      "epoch": 2.59975230803873,
      "grad_norm": 0.37922021746635437,
      "learning_rate": 0.001,
      "loss": 2.3718,
      "step": 46182
    },
    {
      "epoch": 2.6021729340238684,
      "grad_norm": 0.41801437735557556,
      "learning_rate": 0.001,
      "loss": 2.3774,
      "step": 46225
    },
    {
      "epoch": 2.604593560009007,
      "grad_norm": 0.3387153446674347,
      "learning_rate": 0.001,
      "loss": 2.3663,
      "step": 46268
    },
    {
      "epoch": 2.6070141859941454,
      "grad_norm": 0.3605867028236389,
      "learning_rate": 0.001,
      "loss": 2.3787,
      "step": 46311
    },
    {
      "epoch": 2.609434811979284,
      "grad_norm": 0.3328022360801697,
      "learning_rate": 0.001,
      "loss": 2.3817,
      "step": 46354
    },
    {
      "epoch": 2.6118554379644223,
      "grad_norm": 0.3726743161678314,
      "learning_rate": 0.001,
      "loss": 2.368,
      "step": 46397
    },
    {
      "epoch": 2.614276063949561,
      "grad_norm": 0.43381696939468384,
      "learning_rate": 0.001,
      "loss": 2.3698,
      "step": 46440
    },
    {
      "epoch": 2.6166966899346993,
      "grad_norm": 0.4110385775566101,
      "learning_rate": 0.001,
      "loss": 2.367,
      "step": 46483
    },
    {
      "epoch": 2.619117315919838,
      "grad_norm": 0.3883928954601288,
      "learning_rate": 0.001,
      "loss": 2.3709,
      "step": 46526
    },
    {
      "epoch": 2.6215379419049762,
      "grad_norm": 0.3288014829158783,
      "learning_rate": 0.001,
      "loss": 2.377,
      "step": 46569
    },
    {
      "epoch": 2.623958567890115,
      "grad_norm": 0.3974944055080414,
      "learning_rate": 0.001,
      "loss": 2.3659,
      "step": 46612
    },
    {
      "epoch": 2.626379193875253,
      "grad_norm": 0.34351909160614014,
      "learning_rate": 0.001,
      "loss": 2.3586,
      "step": 46655
    },
    {
      "epoch": 2.628799819860392,
      "grad_norm": 0.40419018268585205,
      "learning_rate": 0.001,
      "loss": 2.3795,
      "step": 46698
    },
    {
      "epoch": 2.63122044584553,
      "grad_norm": 0.40513312816619873,
      "learning_rate": 0.001,
      "loss": 2.3761,
      "step": 46741
    },
    {
      "epoch": 2.633641071830669,
      "grad_norm": 0.401887983083725,
      "learning_rate": 0.001,
      "loss": 2.3622,
      "step": 46784
    },
    {
      "epoch": 2.636061697815807,
      "grad_norm": 0.3317054808139801,
      "learning_rate": 0.001,
      "loss": 2.3773,
      "step": 46827
    },
    {
      "epoch": 2.638482323800946,
      "grad_norm": 0.33688563108444214,
      "learning_rate": 0.001,
      "loss": 2.3706,
      "step": 46870
    },
    {
      "epoch": 2.640902949786084,
      "grad_norm": 0.3498205244541168,
      "learning_rate": 0.001,
      "loss": 2.363,
      "step": 46913
    },
    {
      "epoch": 2.641634766944382,
      "eval_ag_news_bleu_score": 4.197539527620954,
      "eval_ag_news_bleu_score_sem": 0.15755702714569508,
      "eval_ag_news_emb_cos_sim": 0.991930365562439,
      "eval_ag_news_emb_cos_sim_sem": 0.00027673865985758205,
      "eval_ag_news_emb_top1_equal": 0.6233333349227905,
      "eval_ag_news_emb_top1_equal_sem": 0.028022260422154388,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.633270025253296,
      "eval_ag_news_n_ngrams_match_1": 5.096,
      "eval_ag_news_n_ngrams_match_2": 0.986,
      "eval_ag_news_n_ngrams_match_3": 0.252,
      "eval_ag_news_num_pred_words": 22.782,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.27806747017537176,
      "eval_ag_news_runtime": 15.2589,
      "eval_ag_news_samples_per_second": 32.768,
      "eval_ag_news_steps_per_second": 0.131,
      "eval_ag_news_token_set_f1": 0.23958026136053687,
      "eval_ag_news_token_set_f1_sem": 0.005209987749144385,
      "eval_ag_news_token_set_precision": 0.2322861460255841,
      "eval_ag_news_token_set_recall": 0.25826358881975925,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 46926
    },
    {
      "epoch": 2.641634766944382,
      "eval_anthropic_toxic_prompts_bleu_score": 5.7086186950330955,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.23814806363618427,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9915887713432312,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00027647846714736417,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6666666865348816,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027262027544015993,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.910087823867798,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.104,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.67,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.696,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.542,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2970849964576204,
      "eval_anthropic_toxic_prompts_runtime": 6.7483,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.093,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.296,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.32869006828337544,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006712143967206406,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4014480082977993,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.302104023500268,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 46926
    },
    {
      "epoch": 2.641634766944382,
      "eval_arxiv_bleu_score": 4.044957409081823,
      "eval_arxiv_bleu_score_sem": 0.13038503652085268,
      "eval_arxiv_emb_cos_sim": 0.9911471605300903,
      "eval_arxiv_emb_cos_sim_sem": 0.00025937271872567545,
      "eval_arxiv_emb_top1_equal": 0.5866666436195374,
      "eval_arxiv_emb_top1_equal_sem": 0.02847805739797067,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.639821767807007,
      "eval_arxiv_n_ngrams_match_1": 5.096,
      "eval_arxiv_n_ngrams_match_2": 0.864,
      "eval_arxiv_n_ngrams_match_3": 0.178,
      "eval_arxiv_num_pred_words": 18.74,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2690578427850656,
      "eval_arxiv_runtime": 6.5913,
      "eval_arxiv_samples_per_second": 75.857,
      "eval_arxiv_steps_per_second": 0.303,
      "eval_arxiv_token_set_f1": 0.2499108685095226,
      "eval_arxiv_token_set_f1_sem": 0.005053671195492807,
      "eval_arxiv_token_set_precision": 0.21667327336134884,
      "eval_arxiv_token_set_recall": 0.31178200502889253,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 46926
    },
    {
      "epoch": 2.641634766944382,
      "eval_python_code_alpaca_bleu_score": 6.206490684013209,
      "eval_python_code_alpaca_bleu_score_sem": 0.21521376674153078,
      "eval_python_code_alpaca_emb_cos_sim": 0.9893166422843933,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.000481432490819718,
      "eval_python_code_alpaca_emb_top1_equal": 0.6399999856948853,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02775911810844162,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.883559465408325,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.168,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.57,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.572,
      "eval_python_code_alpaca_num_pred_words": 21.06,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.33400185194269605,
      "eval_python_code_alpaca_runtime": 6.5506,
      "eval_python_code_alpaca_samples_per_second": 76.329,
      "eval_python_code_alpaca_steps_per_second": 0.305,
      "eval_python_code_alpaca_token_set_f1": 0.3601612193971409,
      "eval_python_code_alpaca_token_set_f1_sem": 0.0056793742213093945,
      "eval_python_code_alpaca_token_set_precision": 0.3670554051428105,
      "eval_python_code_alpaca_token_set_recall": 0.37849754241122446,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 46926
    },
    {
      "epoch": 2.641634766944382,
      "eval_wikibio_bleu_score": 3.865118990133376,
      "eval_wikibio_bleu_score_sem": 0.1290988941912754,
      "eval_wikibio_emb_cos_sim": 0.9909161329269409,
      "eval_wikibio_emb_cos_sim_sem": 0.0003187588146989607,
      "eval_wikibio_emb_top1_equal": 0.6200000047683716,
      "eval_wikibio_emb_top1_equal_sem": 0.028070624111834433,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.8843705654144287,
      "eval_wikibio_n_ngrams_match_1": 3.258,
      "eval_wikibio_n_ngrams_match_2": 0.88,
      "eval_wikibio_n_ngrams_match_3": 0.182,
      "eval_wikibio_num_pred_words": 20.24,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2759749505468428,
      "eval_wikibio_runtime": 6.4517,
      "eval_wikibio_samples_per_second": 77.499,
      "eval_wikibio_steps_per_second": 0.31,
      "eval_wikibio_token_set_f1": 0.2097735656415341,
      "eval_wikibio_token_set_f1_sem": 0.005670409637184574,
      "eval_wikibio_token_set_precision": 0.24583358744342668,
      "eval_wikibio_token_set_recall": 0.18956040653899,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 46926
    },
    {
      "epoch": 2.641634766944382,
      "eval_nq_5round_bleu_score": 12.336756515731574,
      "eval_nq_5round_bleu_score_sem": 0.5496249899592666,
      "eval_nq_5round_emb_cos_sim": 0.9938399195671082,
      "eval_nq_5round_emb_cos_sim_sem": 0.00024261854501689427,
      "eval_nq_5round_emb_top1_equal": 0.6433333158493042,
      "eval_nq_5round_emb_top1_equal_sem": 0.027702163273800266,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.578,
      "eval_nq_5round_n_ngrams_match_2": 3.992,
      "eval_nq_5round_n_ngrams_match_3": 1.926,
      "eval_nq_5round_num_pred_words": 24.05,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4284080767400543,
      "eval_nq_5round_token_set_f1": 0.45101943026014746,
      "eval_nq_5round_token_set_f1_sem": 0.006592267673347352,
      "eval_nq_5round_token_set_precision": 0.43214348653603024,
      "eval_nq_5round_token_set_recall": 0.4807334789480522,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.329908214493221,
      "eval_nq_bleu_score_sem": 0.5075717375045526,
      "eval_nq_emb_cos_sim": 0.9931704998016357,
      "eval_nq_emb_cos_sim_sem": 0.0002666628463659609,
      "eval_nq_emb_top1_equal": 0.6433333158493042,
      "eval_nq_emb_top1_equal_sem": 0.02770216499443815,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.099414825439453,
      "eval_nq_n_ngrams_match_1": 10.026,
      "eval_nq_n_ngrams_match_2": 3.672,
      "eval_nq_n_ngrams_match_3": 1.71,
      "eval_nq_num_pred_words": 23.912,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.40605602741057334,
      "eval_nq_runtime": 18.6497,
      "eval_nq_samples_per_second": 26.81,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.43077395584774547,
      "eval_nq_token_set_f1_sem": 0.00625383563801083,
      "eval_nq_token_set_precision": 0.40959021701177,
      "eval_nq_token_set_recall": 0.46526217452250607,
      "eval_nq_true_num_tokens": 32.0,
      "step": 46926
    },
    {
      "epoch": 2.643323575771223,
      "grad_norm": 0.3680292069911957,
      "learning_rate": 0.001,
      "loss": 2.3631,
      "step": 46956
    },
    {
      "epoch": 2.645744201756361,
      "grad_norm": 0.46452194452285767,
      "learning_rate": 0.001,
      "loss": 2.3634,
      "step": 46999
    },
    {
      "epoch": 2.6481648277414997,
      "grad_norm": 0.33882033824920654,
      "learning_rate": 0.001,
      "loss": 2.3725,
      "step": 47042
    },
    {
      "epoch": 2.650585453726638,
      "grad_norm": 0.4213804304599762,
      "learning_rate": 0.001,
      "loss": 2.3602,
      "step": 47085
    },
    {
      "epoch": 2.6530060797117767,
      "grad_norm": 0.3946714997291565,
      "learning_rate": 0.001,
      "loss": 2.3705,
      "step": 47128
    },
    {
      "epoch": 2.655426705696915,
      "grad_norm": 0.34017056226730347,
      "learning_rate": 0.001,
      "loss": 2.3674,
      "step": 47171
    },
    {
      "epoch": 2.6578473316820537,
      "grad_norm": 0.42140525579452515,
      "learning_rate": 0.001,
      "loss": 2.3689,
      "step": 47214
    },
    {
      "epoch": 2.660267957667192,
      "grad_norm": 0.377823144197464,
      "learning_rate": 0.001,
      "loss": 2.3599,
      "step": 47257
    },
    {
      "epoch": 2.6626885836523306,
      "grad_norm": 0.3680105209350586,
      "learning_rate": 0.001,
      "loss": 2.3692,
      "step": 47300
    },
    {
      "epoch": 2.665109209637469,
      "grad_norm": 0.3778042495250702,
      "learning_rate": 0.001,
      "loss": 2.3755,
      "step": 47343
    },
    {
      "epoch": 2.6675298356226076,
      "grad_norm": 0.4193517565727234,
      "learning_rate": 0.001,
      "loss": 2.3661,
      "step": 47386
    },
    {
      "epoch": 2.669950461607746,
      "grad_norm": 0.3847842216491699,
      "learning_rate": 0.001,
      "loss": 2.3656,
      "step": 47429
    },
    {
      "epoch": 2.6723710875928846,
      "grad_norm": 0.39740270376205444,
      "learning_rate": 0.001,
      "loss": 2.3703,
      "step": 47472
    },
    {
      "epoch": 2.674791713578023,
      "grad_norm": 0.41266173124313354,
      "learning_rate": 0.001,
      "loss": 2.3656,
      "step": 47515
    },
    {
      "epoch": 2.6772123395631615,
      "grad_norm": 0.3682873249053955,
      "learning_rate": 0.001,
      "loss": 2.36,
      "step": 47558
    },
    {
      "epoch": 2.6796329655482998,
      "grad_norm": 0.40641966462135315,
      "learning_rate": 0.001,
      "loss": 2.3569,
      "step": 47601
    },
    {
      "epoch": 2.6820535915334385,
      "grad_norm": 0.3979203999042511,
      "learning_rate": 0.001,
      "loss": 2.3713,
      "step": 47644
    },
    {
      "epoch": 2.6844742175185767,
      "grad_norm": 0.3398992717266083,
      "learning_rate": 0.001,
      "loss": 2.3647,
      "step": 47687
    },
    {
      "epoch": 2.6868948435037154,
      "grad_norm": 0.3842313587665558,
      "learning_rate": 0.001,
      "loss": 2.3629,
      "step": 47730
    },
    {
      "epoch": 2.6893154694888537,
      "grad_norm": 0.3603452146053314,
      "learning_rate": 0.001,
      "loss": 2.3741,
      "step": 47773
    },
    {
      "epoch": 2.6917360954739924,
      "grad_norm": 0.407206267118454,
      "learning_rate": 0.001,
      "loss": 2.3652,
      "step": 47816
    },
    {
      "epoch": 2.694156721459131,
      "grad_norm": 0.3871298134326935,
      "learning_rate": 0.001,
      "loss": 2.3698,
      "step": 47859
    },
    {
      "epoch": 2.6965773474442694,
      "grad_norm": 0.41647645831108093,
      "learning_rate": 0.001,
      "loss": 2.3572,
      "step": 47902
    },
    {
      "epoch": 2.6989979734294076,
      "grad_norm": 0.34122151136398315,
      "learning_rate": 0.001,
      "loss": 2.3616,
      "step": 47945
    },
    {
      "epoch": 2.7014185994145463,
      "grad_norm": 0.45216619968414307,
      "learning_rate": 0.001,
      "loss": 2.357,
      "step": 47988
    },
    {
      "epoch": 2.703839225399685,
      "grad_norm": 0.37675797939300537,
      "learning_rate": 0.001,
      "loss": 2.3712,
      "step": 48031
    },
    {
      "epoch": 2.7062598513848233,
      "grad_norm": 0.41175007820129395,
      "learning_rate": 0.001,
      "loss": 2.3637,
      "step": 48074
    },
    {
      "epoch": 2.7086804773699615,
      "grad_norm": 0.3629567325115204,
      "learning_rate": 0.001,
      "loss": 2.3646,
      "step": 48117
    },
    {
      "epoch": 2.7111011033551002,
      "grad_norm": 0.39334070682525635,
      "learning_rate": 0.001,
      "loss": 2.3504,
      "step": 48160
    },
    {
      "epoch": 2.713521729340239,
      "grad_norm": 0.4162062704563141,
      "learning_rate": 0.001,
      "loss": 2.3681,
      "step": 48203
    },
    {
      "epoch": 2.715942355325377,
      "grad_norm": 0.4380643367767334,
      "learning_rate": 0.001,
      "loss": 2.3553,
      "step": 48246
    },
    {
      "epoch": 2.7183629813105155,
      "grad_norm": 0.3430883288383484,
      "learning_rate": 0.001,
      "loss": 2.3591,
      "step": 48289
    },
    {
      "epoch": 2.720783607295654,
      "grad_norm": 0.4688119888305664,
      "learning_rate": 0.001,
      "loss": 2.3487,
      "step": 48332
    },
    {
      "epoch": 2.723204233280793,
      "grad_norm": 0.35573816299438477,
      "learning_rate": 0.001,
      "loss": 2.3577,
      "step": 48375
    },
    {
      "epoch": 2.725624859265931,
      "grad_norm": 0.38904181122779846,
      "learning_rate": 0.001,
      "loss": 2.3598,
      "step": 48418
    },
    {
      "epoch": 2.7280454852510694,
      "grad_norm": 0.40671297907829285,
      "learning_rate": 0.001,
      "loss": 2.3664,
      "step": 48461
    },
    {
      "epoch": 2.730466111236208,
      "grad_norm": 0.3987201452255249,
      "learning_rate": 0.001,
      "loss": 2.3615,
      "step": 48504
    },
    {
      "epoch": 2.732886737221347,
      "grad_norm": 0.3713367283344269,
      "learning_rate": 0.001,
      "loss": 2.3464,
      "step": 48547
    },
    {
      "epoch": 2.735307363206485,
      "grad_norm": 0.42813804745674133,
      "learning_rate": 0.001,
      "loss": 2.3658,
      "step": 48590
    },
    {
      "epoch": 2.7377279891916233,
      "grad_norm": 0.33698004484176636,
      "learning_rate": 0.001,
      "loss": 2.3513,
      "step": 48633
    },
    {
      "epoch": 2.740148615176762,
      "grad_norm": 0.3946531116962433,
      "learning_rate": 0.001,
      "loss": 2.3555,
      "step": 48676
    },
    {
      "epoch": 2.7425692411619007,
      "grad_norm": 0.38504648208618164,
      "learning_rate": 0.001,
      "loss": 2.3617,
      "step": 48719
    },
    {
      "epoch": 2.744989867147039,
      "grad_norm": 0.49033796787261963,
      "learning_rate": 0.001,
      "loss": 2.37,
      "step": 48762
    },
    {
      "epoch": 2.7474104931321772,
      "grad_norm": 0.38549768924713135,
      "learning_rate": 0.001,
      "loss": 2.3605,
      "step": 48805
    },
    {
      "epoch": 2.749831119117316,
      "grad_norm": 0.38301563262939453,
      "learning_rate": 0.001,
      "loss": 2.358,
      "step": 48848
    },
    {
      "epoch": 2.7522517451024546,
      "grad_norm": 0.39941298961639404,
      "learning_rate": 0.001,
      "loss": 2.3684,
      "step": 48891
    },
    {
      "epoch": 2.754672371087593,
      "grad_norm": 0.3987928330898285,
      "learning_rate": 0.001,
      "loss": 2.3579,
      "step": 48934
    },
    {
      "epoch": 2.757092997072731,
      "grad_norm": 0.42401111125946045,
      "learning_rate": 0.001,
      "loss": 2.3582,
      "step": 48977
    },
    {
      "epoch": 2.75951362305787,
      "grad_norm": 0.4021932780742645,
      "learning_rate": 0.001,
      "loss": 2.3709,
      "step": 49020
    },
    {
      "epoch": 2.761709074532763,
      "eval_ag_news_bleu_score": 4.209617997678192,
      "eval_ag_news_bleu_score_sem": 0.16236880182698857,
      "eval_ag_news_emb_cos_sim": 0.9915399551391602,
      "eval_ag_news_emb_cos_sim_sem": 0.00032593892905445474,
      "eval_ag_news_emb_top1_equal": 0.6333333253860474,
      "eval_ag_news_emb_top1_equal_sem": 0.02786867456387452,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.6178274154663086,
      "eval_ag_news_n_ngrams_match_1": 5.056,
      "eval_ag_news_n_ngrams_match_2": 0.952,
      "eval_ag_news_n_ngrams_match_3": 0.252,
      "eval_ag_news_num_pred_words": 22.534,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.27822136007416365,
      "eval_ag_news_runtime": 14.9479,
      "eval_ag_news_samples_per_second": 33.45,
      "eval_ag_news_steps_per_second": 0.134,
      "eval_ag_news_token_set_f1": 0.23707253430464859,
      "eval_ag_news_token_set_f1_sem": 0.005189259370466797,
      "eval_ag_news_token_set_precision": 0.2288426787146094,
      "eval_ag_news_token_set_recall": 0.2548070741761044,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 49059
    },
    {
      "epoch": 2.761709074532763,
      "eval_anthropic_toxic_prompts_bleu_score": 5.705855396951003,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.24502006207266783,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9914274215698242,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002785776184824647,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6366666555404663,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027814619004069896,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.847844123840332,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.016,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.612,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.658,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.114,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.2990737015307958,
      "eval_anthropic_toxic_prompts_runtime": 6.4923,
      "eval_anthropic_toxic_prompts_samples_per_second": 77.014,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.308,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.32587220192722444,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006494023848475875,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.39568426183342775,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.300768862765015,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 49059
    },
    {
      "epoch": 2.761709074532763,
      "eval_arxiv_bleu_score": 3.85325153000644,
      "eval_arxiv_bleu_score_sem": 0.12291412288619158,
      "eval_arxiv_emb_cos_sim": 0.9914005994796753,
      "eval_arxiv_emb_cos_sim_sem": 0.0002425423261355688,
      "eval_arxiv_emb_top1_equal": 0.6166666746139526,
      "eval_arxiv_emb_top1_equal_sem": 0.028117578599086417,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.62481689453125,
      "eval_arxiv_n_ngrams_match_1": 4.912,
      "eval_arxiv_n_ngrams_match_2": 0.764,
      "eval_arxiv_n_ngrams_match_3": 0.16,
      "eval_arxiv_num_pred_words": 18.1,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2628520214651507,
      "eval_arxiv_runtime": 6.5122,
      "eval_arxiv_samples_per_second": 76.779,
      "eval_arxiv_steps_per_second": 0.307,
      "eval_arxiv_token_set_f1": 0.24108495774169386,
      "eval_arxiv_token_set_f1_sem": 0.004893534376002534,
      "eval_arxiv_token_set_precision": 0.21018824716503368,
      "eval_arxiv_token_set_recall": 0.29777778602785476,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 49059
    },
    {
      "epoch": 2.761709074532763,
      "eval_python_code_alpaca_bleu_score": 6.177940427961755,
      "eval_python_code_alpaca_bleu_score_sem": 0.24611235872130235,
      "eval_python_code_alpaca_emb_cos_sim": 0.9900827407836914,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00033408641829102545,
      "eval_python_code_alpaca_emb_top1_equal": 0.6566666960716248,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027459642805142834,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.9076993465423584,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.714,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.442,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.526,
      "eval_python_code_alpaca_num_pred_words": 19.49,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3215222005035755,
      "eval_python_code_alpaca_runtime": 6.5249,
      "eval_python_code_alpaca_samples_per_second": 76.63,
      "eval_python_code_alpaca_steps_per_second": 0.307,
      "eval_python_code_alpaca_token_set_f1": 0.34759251965145266,
      "eval_python_code_alpaca_token_set_f1_sem": 0.0060660140847854215,
      "eval_python_code_alpaca_token_set_precision": 0.3408450197771403,
      "eval_python_code_alpaca_token_set_recall": 0.38914633097829016,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 49059
    },
    {
      "epoch": 2.761709074532763,
      "eval_wikibio_bleu_score": 3.845614351341233,
      "eval_wikibio_bleu_score_sem": 0.13932669600041298,
      "eval_wikibio_emb_cos_sim": 0.9913191199302673,
      "eval_wikibio_emb_cos_sim_sem": 0.0002656335822909473,
      "eval_wikibio_emb_top1_equal": 0.6299999952316284,
      "eval_wikibio_emb_top1_equal_sem": 0.027921293391044915,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9510557651519775,
      "eval_wikibio_n_ngrams_match_1": 3.156,
      "eval_wikibio_n_ngrams_match_2": 0.83,
      "eval_wikibio_n_ngrams_match_3": 0.174,
      "eval_wikibio_num_pred_words": 20.174,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.27762473636925333,
      "eval_wikibio_runtime": 6.467,
      "eval_wikibio_samples_per_second": 77.316,
      "eval_wikibio_steps_per_second": 0.309,
      "eval_wikibio_token_set_f1": 0.20587755222908027,
      "eval_wikibio_token_set_f1_sem": 0.0054603135772809215,
      "eval_wikibio_token_set_precision": 0.2393965798049336,
      "eval_wikibio_token_set_recall": 0.1865822973588684,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 49059
    },
    {
      "epoch": 2.761709074532763,
      "eval_nq_5round_bleu_score": 13.123773103520918,
      "eval_nq_5round_bleu_score_sem": 0.6204496389421832,
      "eval_nq_5round_emb_cos_sim": 0.9942554235458374,
      "eval_nq_5round_emb_cos_sim_sem": 0.00021684560750485647,
      "eval_nq_5round_emb_top1_equal": 0.6233333349227905,
      "eval_nq_5round_emb_top1_equal_sem": 0.028022260422154388,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.726,
      "eval_nq_5round_n_ngrams_match_2": 4.136,
      "eval_nq_5round_n_ngrams_match_3": 2.102,
      "eval_nq_5round_num_pred_words": 23.81,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4408116863095022,
      "eval_nq_5round_token_set_f1": 0.4633042011741436,
      "eval_nq_5round_token_set_f1_sem": 0.006851748477809549,
      "eval_nq_5round_token_set_precision": 0.44408381699975674,
      "eval_nq_5round_token_set_recall": 0.4930341281087957,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.5514584986066,
      "eval_nq_bleu_score_sem": 0.5223548469174542,
      "eval_nq_emb_cos_sim": 0.9937440156936646,
      "eval_nq_emb_cos_sim_sem": 0.00021687699570377972,
      "eval_nq_emb_top1_equal": 0.6600000262260437,
      "eval_nq_emb_top1_equal_sem": 0.0273952875069568,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.0753002166748047,
      "eval_nq_n_ngrams_match_1": 10.214,
      "eval_nq_n_ngrams_match_2": 3.718,
      "eval_nq_n_ngrams_match_3": 1.726,
      "eval_nq_num_pred_words": 23.88,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.41388469660896,
      "eval_nq_runtime": 18.6846,
      "eval_nq_samples_per_second": 26.76,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.4400551730058439,
      "eval_nq_token_set_f1_sem": 0.00644530721738114,
      "eval_nq_token_set_precision": 0.42060327566164357,
      "eval_nq_token_set_recall": 0.4712320135673396,
      "eval_nq_true_num_tokens": 32.0,
      "step": 49059
    },
    {
      "epoch": 2.7619342490430085,
      "grad_norm": 0.45435261726379395,
      "learning_rate": 0.001,
      "loss": 2.3542,
      "step": 49063
    },
    {
      "epoch": 2.764354875028147,
      "grad_norm": 0.3972301483154297,
      "learning_rate": 0.001,
      "loss": 2.3679,
      "step": 49106
    },
    {
      "epoch": 2.766775501013285,
      "grad_norm": 0.41639795899391174,
      "learning_rate": 0.001,
      "loss": 2.3693,
      "step": 49149
    },
    {
      "epoch": 2.7691961269984238,
      "grad_norm": 0.376834511756897,
      "learning_rate": 0.001,
      "loss": 2.3649,
      "step": 49192
    },
    {
      "epoch": 2.7716167529835625,
      "grad_norm": 0.4844721853733063,
      "learning_rate": 0.001,
      "loss": 2.3531,
      "step": 49235
    },
    {
      "epoch": 2.7740373789687007,
      "grad_norm": 0.3745051622390747,
      "learning_rate": 0.001,
      "loss": 2.3627,
      "step": 49278
    },
    {
      "epoch": 2.776458004953839,
      "grad_norm": 0.42490705847740173,
      "learning_rate": 0.001,
      "loss": 2.3534,
      "step": 49321
    },
    {
      "epoch": 2.7788786309389777,
      "grad_norm": 0.40002599358558655,
      "learning_rate": 0.001,
      "loss": 2.3701,
      "step": 49364
    },
    {
      "epoch": 2.7812992569241164,
      "grad_norm": 0.41728246212005615,
      "learning_rate": 0.001,
      "loss": 2.35,
      "step": 49407
    },
    {
      "epoch": 2.7837198829092547,
      "grad_norm": 0.43016719818115234,
      "learning_rate": 0.001,
      "loss": 2.344,
      "step": 49450
    },
    {
      "epoch": 2.786140508894393,
      "grad_norm": 0.417157918214798,
      "learning_rate": 0.001,
      "loss": 2.343,
      "step": 49493
    },
    {
      "epoch": 2.7885611348795316,
      "grad_norm": 0.3577554523944855,
      "learning_rate": 0.001,
      "loss": 2.3415,
      "step": 49536
    },
    {
      "epoch": 2.7909817608646703,
      "grad_norm": 0.3956291377544403,
      "learning_rate": 0.001,
      "loss": 2.3584,
      "step": 49579
    },
    {
      "epoch": 2.7934023868498086,
      "grad_norm": 0.36730337142944336,
      "learning_rate": 0.001,
      "loss": 2.3401,
      "step": 49622
    },
    {
      "epoch": 2.795823012834947,
      "grad_norm": 0.4165145456790924,
      "learning_rate": 0.001,
      "loss": 2.3655,
      "step": 49665
    },
    {
      "epoch": 2.7982436388200855,
      "grad_norm": 0.37215283513069153,
      "learning_rate": 0.001,
      "loss": 2.3457,
      "step": 49708
    },
    {
      "epoch": 2.8006642648052242,
      "grad_norm": 0.42801520228385925,
      "learning_rate": 0.001,
      "loss": 2.3527,
      "step": 49751
    },
    {
      "epoch": 2.8030848907903625,
      "grad_norm": 0.34249308705329895,
      "learning_rate": 0.001,
      "loss": 2.357,
      "step": 49794
    },
    {
      "epoch": 2.8055055167755008,
      "grad_norm": 0.32792118191719055,
      "learning_rate": 0.001,
      "loss": 2.3589,
      "step": 49837
    },
    {
      "epoch": 2.8079261427606395,
      "grad_norm": 0.4032750129699707,
      "learning_rate": 0.001,
      "loss": 2.3572,
      "step": 49880
    },
    {
      "epoch": 2.810346768745778,
      "grad_norm": 0.39931589365005493,
      "learning_rate": 0.001,
      "loss": 2.3584,
      "step": 49923
    },
    {
      "epoch": 2.8127673947309164,
      "grad_norm": 0.39806920289993286,
      "learning_rate": 0.001,
      "loss": 2.3552,
      "step": 49966
    },
    {
      "epoch": 2.8151880207160547,
      "grad_norm": 0.4629383385181427,
      "learning_rate": 0.001,
      "loss": 2.3544,
      "step": 50009
    },
    {
      "epoch": 2.8176086467011934,
      "grad_norm": 0.4589201807975769,
      "learning_rate": 0.001,
      "loss": 2.3423,
      "step": 50052
    },
    {
      "epoch": 2.820029272686332,
      "grad_norm": 0.42653754353523254,
      "learning_rate": 0.001,
      "loss": 2.3492,
      "step": 50095
    },
    {
      "epoch": 2.8224498986714703,
      "grad_norm": 0.346945196390152,
      "learning_rate": 0.001,
      "loss": 2.3502,
      "step": 50138
    },
    {
      "epoch": 2.8248705246566086,
      "grad_norm": 0.4170111119747162,
      "learning_rate": 0.001,
      "loss": 2.3525,
      "step": 50181
    },
    {
      "epoch": 2.8272911506417473,
      "grad_norm": 0.37393733859062195,
      "learning_rate": 0.001,
      "loss": 2.3501,
      "step": 50224
    },
    {
      "epoch": 2.829711776626886,
      "grad_norm": 0.3640117347240448,
      "learning_rate": 0.001,
      "loss": 2.3593,
      "step": 50267
    },
    {
      "epoch": 2.8321324026120243,
      "grad_norm": 0.3725832998752594,
      "learning_rate": 0.001,
      "loss": 2.3542,
      "step": 50310
    },
    {
      "epoch": 2.834553028597163,
      "grad_norm": 0.35561805963516235,
      "learning_rate": 0.001,
      "loss": 2.3521,
      "step": 50353
    },
    {
      "epoch": 2.836973654582301,
      "grad_norm": 0.4639611542224884,
      "learning_rate": 0.001,
      "loss": 2.3447,
      "step": 50396
    },
    {
      "epoch": 2.83939428056744,
      "grad_norm": 0.39092347025871277,
      "learning_rate": 0.001,
      "loss": 2.3511,
      "step": 50439
    },
    {
      "epoch": 2.841814906552578,
      "grad_norm": 0.40049174427986145,
      "learning_rate": 0.001,
      "loss": 2.3501,
      "step": 50482
    },
    {
      "epoch": 2.844235532537717,
      "grad_norm": 0.3411664068698883,
      "learning_rate": 0.001,
      "loss": 2.3582,
      "step": 50525
    },
    {
      "epoch": 2.846656158522855,
      "grad_norm": 0.45036011934280396,
      "learning_rate": 0.001,
      "loss": 2.349,
      "step": 50568
    },
    {
      "epoch": 2.849076784507994,
      "grad_norm": 0.3835832476615906,
      "learning_rate": 0.001,
      "loss": 2.3425,
      "step": 50611
    },
    {
      "epoch": 2.851497410493132,
      "grad_norm": 0.38475748896598816,
      "learning_rate": 0.001,
      "loss": 2.3564,
      "step": 50654
    },
    {
      "epoch": 2.853918036478271,
      "grad_norm": 0.4234263002872467,
      "learning_rate": 0.001,
      "loss": 2.3473,
      "step": 50697
    },
    {
      "epoch": 2.856338662463409,
      "grad_norm": 0.3861537277698517,
      "learning_rate": 0.001,
      "loss": 2.3425,
      "step": 50740
    },
    {
      "epoch": 2.8587592884485478,
      "grad_norm": 0.3899056613445282,
      "learning_rate": 0.001,
      "loss": 2.346,
      "step": 50783
    },
    {
      "epoch": 2.861179914433686,
      "grad_norm": 0.40977463126182556,
      "learning_rate": 0.001,
      "loss": 2.3442,
      "step": 50826
    },
    {
      "epoch": 2.8636005404188247,
      "grad_norm": 0.424645334482193,
      "learning_rate": 0.001,
      "loss": 2.3463,
      "step": 50869
    },
    {
      "epoch": 2.866021166403963,
      "grad_norm": 0.3778836727142334,
      "learning_rate": 0.001,
      "loss": 2.3412,
      "step": 50912
    },
    {
      "epoch": 2.8684417923891017,
      "grad_norm": 0.41537660360336304,
      "learning_rate": 0.001,
      "loss": 2.3393,
      "step": 50955
    },
    {
      "epoch": 2.87086241837424,
      "grad_norm": 0.3698303699493408,
      "learning_rate": 0.001,
      "loss": 2.3407,
      "step": 50998
    },
    {
      "epoch": 2.8732830443593786,
      "grad_norm": 0.43733757734298706,
      "learning_rate": 0.001,
      "loss": 2.3308,
      "step": 51041
    },
    {
      "epoch": 2.875703670344517,
      "grad_norm": 0.39628684520721436,
      "learning_rate": 0.001,
      "loss": 2.3562,
      "step": 51084
    },
    {
      "epoch": 2.8781242963296556,
      "grad_norm": 0.41509613394737244,
      "learning_rate": 0.001,
      "loss": 2.3394,
      "step": 51127
    },
    {
      "epoch": 2.880544922314794,
      "grad_norm": 0.4138786792755127,
      "learning_rate": 0.001,
      "loss": 2.343,
      "step": 51170
    },
    {
      "epoch": 2.881783382121144,
      "eval_ag_news_bleu_score": 4.355313784337991,
      "eval_ag_news_bleu_score_sem": 0.16831787247443653,
      "eval_ag_news_emb_cos_sim": 0.9918799996376038,
      "eval_ag_news_emb_cos_sim_sem": 0.00028122296480628047,
      "eval_ag_news_emb_top1_equal": 0.6133333444595337,
      "eval_ag_news_emb_top1_equal_sem": 0.028163139369651306,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.549003839492798,
      "eval_ag_news_n_ngrams_match_1": 5.134,
      "eval_ag_news_n_ngrams_match_2": 0.988,
      "eval_ag_news_n_ngrams_match_3": 0.24,
      "eval_ag_news_num_pred_words": 22.364,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2856069762511186,
      "eval_ag_news_runtime": 15.699,
      "eval_ag_news_samples_per_second": 31.849,
      "eval_ag_news_steps_per_second": 0.127,
      "eval_ag_news_token_set_f1": 0.24280685584104478,
      "eval_ag_news_token_set_f1_sem": 0.005247282044332534,
      "eval_ag_news_token_set_precision": 0.23366105543346385,
      "eval_ag_news_token_set_recall": 0.26342818852505484,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 51192
    },
    {
      "epoch": 2.881783382121144,
      "eval_anthropic_toxic_prompts_bleu_score": 5.911337515291984,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2394824070185179,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9918864369392395,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002262566498609948,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6166666746139526,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.028117578599086417,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8602330684661865,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.192,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.702,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.718,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.448,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.30554579369804596,
      "eval_anthropic_toxic_prompts_runtime": 6.7267,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.33,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.297,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.33874167470817945,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006526177313091385,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4085567107471676,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.31440329647812953,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 51192
    },
    {
      "epoch": 2.881783382121144,
      "eval_arxiv_bleu_score": 3.7132608879848914,
      "eval_arxiv_bleu_score_sem": 0.11326061824249234,
      "eval_arxiv_emb_cos_sim": 0.9915195107460022,
      "eval_arxiv_emb_cos_sim_sem": 0.00023785065743672467,
      "eval_arxiv_emb_top1_equal": 0.5866666436195374,
      "eval_arxiv_emb_top1_equal_sem": 0.02847805567733279,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.6259570121765137,
      "eval_arxiv_n_ngrams_match_1": 4.838,
      "eval_arxiv_n_ngrams_match_2": 0.732,
      "eval_arxiv_n_ngrams_match_3": 0.122,
      "eval_arxiv_num_pred_words": 18.054,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.26358620286576034,
      "eval_arxiv_runtime": 6.6878,
      "eval_arxiv_samples_per_second": 74.763,
      "eval_arxiv_steps_per_second": 0.299,
      "eval_arxiv_token_set_f1": 0.23935979531429558,
      "eval_arxiv_token_set_f1_sem": 0.004939826609392321,
      "eval_arxiv_token_set_precision": 0.21001787858666723,
      "eval_arxiv_token_set_recall": 0.2906627312847339,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 51192
    },
    {
      "epoch": 2.881783382121144,
      "eval_python_code_alpaca_bleu_score": 6.111734981783877,
      "eval_python_code_alpaca_bleu_score_sem": 0.21577121119859927,
      "eval_python_code_alpaca_emb_cos_sim": 0.990050196647644,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00033763776111635316,
      "eval_python_code_alpaca_emb_top1_equal": 0.6866666674613953,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.026825057787940674,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.9396185874938965,
      "eval_python_code_alpaca_n_ngrams_match_1": 5.906,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.45,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.498,
      "eval_python_code_alpaca_num_pred_words": 19.682,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3334304591646431,
      "eval_python_code_alpaca_runtime": 6.6256,
      "eval_python_code_alpaca_samples_per_second": 75.465,
      "eval_python_code_alpaca_steps_per_second": 0.302,
      "eval_python_code_alpaca_token_set_f1": 0.3525821230184145,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005746071158251165,
      "eval_python_code_alpaca_token_set_precision": 0.35485608595955315,
      "eval_python_code_alpaca_token_set_recall": 0.37827644124358023,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 51192
    },
    {
      "epoch": 2.881783382121144,
      "eval_wikibio_bleu_score": 3.9384491089215934,
      "eval_wikibio_bleu_score_sem": 0.12468548329456719,
      "eval_wikibio_emb_cos_sim": 0.9912716150283813,
      "eval_wikibio_emb_cos_sim_sem": 0.0002687261600401066,
      "eval_wikibio_emb_top1_equal": 0.6499999761581421,
      "eval_wikibio_emb_top1_equal_sem": 0.027583864257272155,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9341254234313965,
      "eval_wikibio_n_ngrams_match_1": 3.292,
      "eval_wikibio_n_ngrams_match_2": 0.874,
      "eval_wikibio_n_ngrams_match_3": 0.16,
      "eval_wikibio_num_pred_words": 20.028,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2905036109895679,
      "eval_wikibio_runtime": 6.5805,
      "eval_wikibio_samples_per_second": 75.982,
      "eval_wikibio_steps_per_second": 0.304,
      "eval_wikibio_token_set_f1": 0.2142728209173035,
      "eval_wikibio_token_set_f1_sem": 0.00530576901463422,
      "eval_wikibio_token_set_precision": 0.2516486485935955,
      "eval_wikibio_token_set_recall": 0.19313792027217214,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 51192
    },
    {
      "epoch": 2.881783382121144,
      "eval_nq_5round_bleu_score": 13.188020510197013,
      "eval_nq_5round_bleu_score_sem": 0.6065398845993543,
      "eval_nq_5round_emb_cos_sim": 0.9941616058349609,
      "eval_nq_5round_emb_cos_sim_sem": 0.00023877939862028294,
      "eval_nq_5round_emb_top1_equal": 0.6033333539962769,
      "eval_nq_5round_emb_top1_equal_sem": 0.028291497235256893,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.766,
      "eval_nq_5round_n_ngrams_match_2": 4.24,
      "eval_nq_5round_n_ngrams_match_3": 2.082,
      "eval_nq_5round_num_pred_words": 23.868,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4387910866870198,
      "eval_nq_5round_token_set_f1": 0.4639251811260319,
      "eval_nq_5round_token_set_f1_sem": 0.006933309876319887,
      "eval_nq_5round_token_set_precision": 0.44167831624743664,
      "eval_nq_5round_token_set_recall": 0.49798213696482874,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.63217583931798,
      "eval_nq_bleu_score_sem": 0.5321417700927487,
      "eval_nq_emb_cos_sim": 0.9938931465148926,
      "eval_nq_emb_cos_sim_sem": 0.00020598584181922845,
      "eval_nq_emb_top1_equal": 0.6899999976158142,
      "eval_nq_emb_top1_equal_sem": 0.026746674129075776,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.0595483779907227,
      "eval_nq_n_ngrams_match_1": 10.208,
      "eval_nq_n_ngrams_match_2": 3.704,
      "eval_nq_n_ngrams_match_3": 1.762,
      "eval_nq_num_pred_words": 23.77,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4132307626764155,
      "eval_nq_runtime": 22.5714,
      "eval_nq_samples_per_second": 22.152,
      "eval_nq_steps_per_second": 0.089,
      "eval_nq_token_set_f1": 0.43902160791538497,
      "eval_nq_token_set_f1_sem": 0.006494087110493651,
      "eval_nq_token_set_precision": 0.4161191972385361,
      "eval_nq_token_set_recall": 0.4750053592320848,
      "eval_nq_true_num_tokens": 32.0,
      "step": 51192
    },
    {
      "epoch": 2.8829655482999326,
      "grad_norm": 0.33793455362319946,
      "learning_rate": 0.001,
      "loss": 2.3381,
      "step": 51213
    },
    {
      "epoch": 2.885386174285071,
      "grad_norm": 0.3836718797683716,
      "learning_rate": 0.001,
      "loss": 2.3421,
      "step": 51256
    },
    {
      "epoch": 2.8878068002702095,
      "grad_norm": 0.38045796751976013,
      "learning_rate": 0.001,
      "loss": 2.3385,
      "step": 51299
    },
    {
      "epoch": 2.890227426255348,
      "grad_norm": 0.4255308508872986,
      "learning_rate": 0.001,
      "loss": 2.345,
      "step": 51342
    },
    {
      "epoch": 2.8926480522404865,
      "grad_norm": 0.3759765923023224,
      "learning_rate": 0.001,
      "loss": 2.3407,
      "step": 51385
    },
    {
      "epoch": 2.8950686782256247,
      "grad_norm": 0.36301809549331665,
      "learning_rate": 0.001,
      "loss": 2.3387,
      "step": 51428
    },
    {
      "epoch": 2.8974893042107634,
      "grad_norm": 0.3596934974193573,
      "learning_rate": 0.001,
      "loss": 2.3408,
      "step": 51471
    },
    {
      "epoch": 2.8999099301959017,
      "grad_norm": 0.3649439215660095,
      "learning_rate": 0.001,
      "loss": 2.3382,
      "step": 51514
    },
    {
      "epoch": 2.9023305561810404,
      "grad_norm": 0.39899420738220215,
      "learning_rate": 0.001,
      "loss": 2.345,
      "step": 51557
    },
    {
      "epoch": 2.9047511821661787,
      "grad_norm": 0.3704439103603363,
      "learning_rate": 0.001,
      "loss": 2.3433,
      "step": 51600
    },
    {
      "epoch": 2.9071718081513174,
      "grad_norm": 0.4402637481689453,
      "learning_rate": 0.001,
      "loss": 2.3438,
      "step": 51643
    },
    {
      "epoch": 2.9095924341364556,
      "grad_norm": 0.3675265610218048,
      "learning_rate": 0.001,
      "loss": 2.3351,
      "step": 51686
    },
    {
      "epoch": 2.9120130601215943,
      "grad_norm": 0.48207202553749084,
      "learning_rate": 0.001,
      "loss": 2.34,
      "step": 51729
    },
    {
      "epoch": 2.9144336861067326,
      "grad_norm": 0.4268812835216522,
      "learning_rate": 0.001,
      "loss": 2.3287,
      "step": 51772
    },
    {
      "epoch": 2.9168543120918713,
      "grad_norm": 0.4033609628677368,
      "learning_rate": 0.001,
      "loss": 2.3385,
      "step": 51815
    },
    {
      "epoch": 2.9192749380770096,
      "grad_norm": 0.3876606225967407,
      "learning_rate": 0.001,
      "loss": 2.3417,
      "step": 51858
    },
    {
      "epoch": 2.9216955640621483,
      "grad_norm": 0.40050381422042847,
      "learning_rate": 0.001,
      "loss": 2.3411,
      "step": 51901
    },
    {
      "epoch": 2.9241161900472865,
      "grad_norm": 0.4080885052680969,
      "learning_rate": 0.001,
      "loss": 2.3454,
      "step": 51944
    },
    {
      "epoch": 2.926536816032425,
      "grad_norm": 0.37711140513420105,
      "learning_rate": 0.001,
      "loss": 2.3573,
      "step": 51987
    },
    {
      "epoch": 2.9289574420175635,
      "grad_norm": 0.3947446942329407,
      "learning_rate": 0.001,
      "loss": 2.3406,
      "step": 52030
    },
    {
      "epoch": 2.931378068002702,
      "grad_norm": 0.4006566107273102,
      "learning_rate": 0.001,
      "loss": 2.3364,
      "step": 52073
    },
    {
      "epoch": 2.9337986939878404,
      "grad_norm": 0.35701245069503784,
      "learning_rate": 0.001,
      "loss": 2.3372,
      "step": 52116
    },
    {
      "epoch": 2.936219319972979,
      "grad_norm": 0.3934337794780731,
      "learning_rate": 0.001,
      "loss": 2.3412,
      "step": 52159
    },
    {
      "epoch": 2.9386399459581174,
      "grad_norm": 0.4216381311416626,
      "learning_rate": 0.001,
      "loss": 2.3428,
      "step": 52202
    },
    {
      "epoch": 2.941060571943256,
      "grad_norm": 0.3740784227848053,
      "learning_rate": 0.001,
      "loss": 2.3367,
      "step": 52245
    },
    {
      "epoch": 2.9434811979283944,
      "grad_norm": 0.39236384630203247,
      "learning_rate": 0.001,
      "loss": 2.3352,
      "step": 52288
    },
    {
      "epoch": 2.945901823913533,
      "grad_norm": 0.4496864676475525,
      "learning_rate": 0.001,
      "loss": 2.3508,
      "step": 52331
    },
    {
      "epoch": 2.9483224498986713,
      "grad_norm": 0.43271955847740173,
      "learning_rate": 0.001,
      "loss": 2.3401,
      "step": 52374
    },
    {
      "epoch": 2.95074307588381,
      "grad_norm": 0.44769519567489624,
      "learning_rate": 0.001,
      "loss": 2.3407,
      "step": 52417
    },
    {
      "epoch": 2.9531637018689483,
      "grad_norm": 0.4225563406944275,
      "learning_rate": 0.001,
      "loss": 2.3286,
      "step": 52460
    },
    {
      "epoch": 2.955584327854087,
      "grad_norm": 0.3623674809932709,
      "learning_rate": 0.001,
      "loss": 2.3328,
      "step": 52503
    },
    {
      "epoch": 2.9580049538392252,
      "grad_norm": 0.38770195841789246,
      "learning_rate": 0.001,
      "loss": 2.3414,
      "step": 52546
    },
    {
      "epoch": 2.960425579824364,
      "grad_norm": 0.4080352485179901,
      "learning_rate": 0.001,
      "loss": 2.3395,
      "step": 52589
    },
    {
      "epoch": 2.962846205809502,
      "grad_norm": 0.4001011550426483,
      "learning_rate": 0.001,
      "loss": 2.33,
      "step": 52632
    },
    {
      "epoch": 2.965266831794641,
      "grad_norm": 0.38674217462539673,
      "learning_rate": 0.001,
      "loss": 2.33,
      "step": 52675
    },
    {
      "epoch": 2.967687457779779,
      "grad_norm": 0.435863196849823,
      "learning_rate": 0.001,
      "loss": 2.337,
      "step": 52718
    },
    {
      "epoch": 2.970108083764918,
      "grad_norm": 0.35077035427093506,
      "learning_rate": 0.001,
      "loss": 2.3274,
      "step": 52761
    },
    {
      "epoch": 2.9725287097500566,
      "grad_norm": 0.3825863301753998,
      "learning_rate": 0.001,
      "loss": 2.3434,
      "step": 52804
    },
    {
      "epoch": 2.974949335735195,
      "grad_norm": 0.41368573904037476,
      "learning_rate": 0.001,
      "loss": 2.3345,
      "step": 52847
    },
    {
      "epoch": 2.977369961720333,
      "grad_norm": 0.4171009957790375,
      "learning_rate": 0.001,
      "loss": 2.3302,
      "step": 52890
    },
    {
      "epoch": 2.979790587705472,
      "grad_norm": 0.37666305899620056,
      "learning_rate": 0.001,
      "loss": 2.3284,
      "step": 52933
    },
    {
      "epoch": 2.9822112136906105,
      "grad_norm": 0.4124838709831238,
      "learning_rate": 0.001,
      "loss": 2.3393,
      "step": 52976
    },
    {
      "epoch": 2.9846318396757487,
      "grad_norm": 0.43958693742752075,
      "learning_rate": 0.001,
      "loss": 2.3383,
      "step": 53019
    },
    {
      "epoch": 2.987052465660887,
      "grad_norm": 0.4249688982963562,
      "learning_rate": 0.001,
      "loss": 2.3373,
      "step": 53062
    },
    {
      "epoch": 2.9894730916460257,
      "grad_norm": 0.43102797865867615,
      "learning_rate": 0.001,
      "loss": 2.3354,
      "step": 53105
    },
    {
      "epoch": 2.9918937176311644,
      "grad_norm": 0.3637291491031647,
      "learning_rate": 0.001,
      "loss": 2.338,
      "step": 53148
    },
    {
      "epoch": 2.9943143436163027,
      "grad_norm": 0.378751277923584,
      "learning_rate": 0.001,
      "loss": 2.3411,
      "step": 53191
    },
    {
      "epoch": 2.996734969601441,
      "grad_norm": 0.42051395773887634,
      "learning_rate": 0.001,
      "loss": 2.3353,
      "step": 53234
    },
    {
      "epoch": 2.9991555955865796,
      "grad_norm": 0.37460005283355713,
      "learning_rate": 0.001,
      "loss": 2.3363,
      "step": 53277
    },
    {
      "epoch": 3.001576221571718,
      "grad_norm": 0.41763439774513245,
      "learning_rate": 0.001,
      "loss": 2.3108,
      "step": 53320
    },
    {
      "epoch": 3.001857689709525,
      "eval_ag_news_bleu_score": 4.321925460280258,
      "eval_ag_news_bleu_score_sem": 0.16347550075891104,
      "eval_ag_news_emb_cos_sim": 0.9918847680091858,
      "eval_ag_news_emb_cos_sim_sem": 0.0003003161575253599,
      "eval_ag_news_emb_top1_equal": 0.6133333444595337,
      "eval_ag_news_emb_top1_equal_sem": 0.028163139369651306,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.585192918777466,
      "eval_ag_news_n_ngrams_match_1": 5.076,
      "eval_ag_news_n_ngrams_match_2": 1.022,
      "eval_ag_news_n_ngrams_match_3": 0.266,
      "eval_ag_news_num_pred_words": 22.48,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2833054505291288,
      "eval_ag_news_runtime": 16.8605,
      "eval_ag_news_samples_per_second": 29.655,
      "eval_ag_news_steps_per_second": 0.119,
      "eval_ag_news_token_set_f1": 0.23746043305819875,
      "eval_ag_news_token_set_f1_sem": 0.005356528412427952,
      "eval_ag_news_token_set_precision": 0.2305667070558599,
      "eval_ag_news_token_set_recall": 0.25411230974919913,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 53325
    },
    {
      "epoch": 3.001857689709525,
      "eval_anthropic_toxic_prompts_bleu_score": 6.124580433138715,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2465201508222283,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9917296767234802,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00027358758042032253,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6466666460037231,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027643749338232177,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8684639930725098,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.24,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.8,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.776,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.564,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3083768186510827,
      "eval_anthropic_toxic_prompts_runtime": 7.0121,
      "eval_anthropic_toxic_prompts_samples_per_second": 71.306,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.285,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3472858652870321,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006565355682789178,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.41554567610622795,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3258067250307018,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 53325
    },
    {
      "epoch": 3.001857689709525,
      "eval_arxiv_bleu_score": 3.9859813178036982,
      "eval_arxiv_bleu_score_sem": 0.11450936816798937,
      "eval_arxiv_emb_cos_sim": 0.9912902116775513,
      "eval_arxiv_emb_cos_sim_sem": 0.00026920648685977523,
      "eval_arxiv_emb_top1_equal": 0.5633333325386047,
      "eval_arxiv_emb_top1_equal_sem": 0.028682839115889792,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5954744815826416,
      "eval_arxiv_n_ngrams_match_1": 5.126,
      "eval_arxiv_n_ngrams_match_2": 0.844,
      "eval_arxiv_n_ngrams_match_3": 0.172,
      "eval_arxiv_num_pred_words": 18.576,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.274478448600258,
      "eval_arxiv_runtime": 6.7516,
      "eval_arxiv_samples_per_second": 74.056,
      "eval_arxiv_steps_per_second": 0.296,
      "eval_arxiv_token_set_f1": 0.2517372889131928,
      "eval_arxiv_token_set_f1_sem": 0.0047407105646377285,
      "eval_arxiv_token_set_precision": 0.2220706537269786,
      "eval_arxiv_token_set_recall": 0.30492449703113955,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 53325
    },
    {
      "epoch": 3.001857689709525,
      "eval_python_code_alpaca_bleu_score": 6.446821169289777,
      "eval_python_code_alpaca_bleu_score_sem": 0.23522429643438494,
      "eval_python_code_alpaca_emb_cos_sim": 0.9907196164131165,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003009924757541511,
      "eval_python_code_alpaca_emb_top1_equal": 0.6433333158493042,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027702163273800266,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.8474833965301514,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.022,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.568,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.568,
      "eval_python_code_alpaca_num_pred_words": 20.012,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.33990541463017887,
      "eval_python_code_alpaca_runtime": 6.7024,
      "eval_python_code_alpaca_samples_per_second": 74.6,
      "eval_python_code_alpaca_steps_per_second": 0.298,
      "eval_python_code_alpaca_token_set_f1": 0.3620572080590661,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006198664555653903,
      "eval_python_code_alpaca_token_set_precision": 0.3619296699114842,
      "eval_python_code_alpaca_token_set_recall": 0.39312352072308154,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 53325
    },
    {
      "epoch": 3.001857689709525,
      "eval_wikibio_bleu_score": 4.065823605155038,
      "eval_wikibio_bleu_score_sem": 0.14974010884760136,
      "eval_wikibio_emb_cos_sim": 0.9910346865653992,
      "eval_wikibio_emb_cos_sim_sem": 0.00027603276816513724,
      "eval_wikibio_emb_top1_equal": 0.653333306312561,
      "eval_wikibio_emb_top1_equal_sem": 0.027522495986455002,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9525411128997803,
      "eval_wikibio_n_ngrams_match_1": 3.24,
      "eval_wikibio_n_ngrams_match_2": 0.888,
      "eval_wikibio_n_ngrams_match_3": 0.192,
      "eval_wikibio_num_pred_words": 19.636,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.29057654385452325,
      "eval_wikibio_runtime": 6.6399,
      "eval_wikibio_samples_per_second": 75.302,
      "eval_wikibio_steps_per_second": 0.301,
      "eval_wikibio_token_set_f1": 0.21313973508197226,
      "eval_wikibio_token_set_f1_sem": 0.005475531997764196,
      "eval_wikibio_token_set_precision": 0.24643742136320487,
      "eval_wikibio_token_set_recall": 0.19487848568827007,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 53325
    },
    {
      "epoch": 3.001857689709525,
      "eval_nq_5round_bleu_score": 13.823382608042488,
      "eval_nq_5round_bleu_score_sem": 0.6296948458765295,
      "eval_nq_5round_emb_cos_sim": 0.9944865107536316,
      "eval_nq_5round_emb_cos_sim_sem": 0.00021303239165887597,
      "eval_nq_5round_emb_top1_equal": 0.6866666674613953,
      "eval_nq_5round_emb_top1_equal_sem": 0.02682505950857856,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 10.892,
      "eval_nq_5round_n_ngrams_match_2": 4.422,
      "eval_nq_5round_n_ngrams_match_3": 2.24,
      "eval_nq_5round_num_pred_words": 23.784,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4442161035884881,
      "eval_nq_5round_token_set_f1": 0.4677255211987438,
      "eval_nq_5round_token_set_f1_sem": 0.007091679045385308,
      "eval_nq_5round_token_set_precision": 0.4480835173815382,
      "eval_nq_5round_token_set_recall": 0.49708393512113014,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.937543166196988,
      "eval_nq_bleu_score_sem": 0.5188770965173711,
      "eval_nq_emb_cos_sim": 0.9937622547149658,
      "eval_nq_emb_cos_sim_sem": 0.0002485512237905119,
      "eval_nq_emb_top1_equal": 0.6499999761581421,
      "eval_nq_emb_top1_equal_sem": 0.027583864257272155,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.047739028930664,
      "eval_nq_n_ngrams_match_1": 10.164,
      "eval_nq_n_ngrams_match_2": 3.812,
      "eval_nq_n_ngrams_match_3": 1.826,
      "eval_nq_num_pred_words": 23.752,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4143502085031048,
      "eval_nq_runtime": 18.9772,
      "eval_nq_samples_per_second": 26.347,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.44131718935109193,
      "eval_nq_token_set_f1_sem": 0.0064274854046051585,
      "eval_nq_token_set_precision": 0.4197022148073054,
      "eval_nq_token_set_recall": 0.4745744343911579,
      "eval_nq_true_num_tokens": 32.0,
      "step": 53325
    },
    {
      "epoch": 3.0039968475568566,
      "grad_norm": 0.3745954930782318,
      "learning_rate": 0.001,
      "loss": 2.2782,
      "step": 53363
    },
    {
      "epoch": 3.006417473541995,
      "grad_norm": 0.4108395576477051,
      "learning_rate": 0.001,
      "loss": 2.2812,
      "step": 53406
    },
    {
      "epoch": 3.0088380995271335,
      "grad_norm": 0.36995700001716614,
      "learning_rate": 0.001,
      "loss": 2.271,
      "step": 53449
    },
    {
      "epoch": 3.011258725512272,
      "grad_norm": 0.4308566451072693,
      "learning_rate": 0.001,
      "loss": 2.2737,
      "step": 53492
    },
    {
      "epoch": 3.0136793514974105,
      "grad_norm": 0.3811626434326172,
      "learning_rate": 0.001,
      "loss": 2.282,
      "step": 53535
    },
    {
      "epoch": 3.0160999774825488,
      "grad_norm": 0.42669597268104553,
      "learning_rate": 0.001,
      "loss": 2.2882,
      "step": 53578
    },
    {
      "epoch": 3.0185206034676875,
      "grad_norm": 0.4143531322479248,
      "learning_rate": 0.001,
      "loss": 2.2843,
      "step": 53621
    },
    {
      "epoch": 3.0209412294528257,
      "grad_norm": 0.40538978576660156,
      "learning_rate": 0.001,
      "loss": 2.2855,
      "step": 53664
    },
    {
      "epoch": 3.0233618554379644,
      "grad_norm": 0.41121789813041687,
      "learning_rate": 0.001,
      "loss": 2.2705,
      "step": 53707
    },
    {
      "epoch": 3.0257824814231027,
      "grad_norm": 0.4574834108352661,
      "learning_rate": 0.001,
      "loss": 2.2935,
      "step": 53750
    },
    {
      "epoch": 3.0282031074082414,
      "grad_norm": 0.38775020837783813,
      "learning_rate": 0.001,
      "loss": 2.2779,
      "step": 53793
    },
    {
      "epoch": 3.03062373339338,
      "grad_norm": 0.45131412148475647,
      "learning_rate": 0.001,
      "loss": 2.2872,
      "step": 53836
    },
    {
      "epoch": 3.0330443593785184,
      "grad_norm": 0.42348894476890564,
      "learning_rate": 0.001,
      "loss": 2.2844,
      "step": 53879
    },
    {
      "epoch": 3.0354649853636566,
      "grad_norm": 0.43472227454185486,
      "learning_rate": 0.001,
      "loss": 2.2678,
      "step": 53922
    },
    {
      "epoch": 3.0378856113487953,
      "grad_norm": 0.4022725522518158,
      "learning_rate": 0.001,
      "loss": 2.2884,
      "step": 53965
    },
    {
      "epoch": 3.040306237333934,
      "grad_norm": 0.4347687065601349,
      "learning_rate": 0.001,
      "loss": 2.284,
      "step": 54008
    },
    {
      "epoch": 3.0427268633190723,
      "grad_norm": 0.45618724822998047,
      "learning_rate": 0.001,
      "loss": 2.276,
      "step": 54051
    },
    {
      "epoch": 3.045147489304211,
      "grad_norm": 0.4210381805896759,
      "learning_rate": 0.001,
      "loss": 2.2734,
      "step": 54094
    },
    {
      "epoch": 3.0475681152893492,
      "grad_norm": 0.36106744408607483,
      "learning_rate": 0.001,
      "loss": 2.2789,
      "step": 54137
    },
    {
      "epoch": 3.049988741274488,
      "grad_norm": 0.37616223096847534,
      "learning_rate": 0.001,
      "loss": 2.2837,
      "step": 54180
    },
    {
      "epoch": 3.052409367259626,
      "grad_norm": 0.4477735161781311,
      "learning_rate": 0.001,
      "loss": 2.2859,
      "step": 54223
    },
    {
      "epoch": 3.054829993244765,
      "grad_norm": 0.4151265621185303,
      "learning_rate": 0.001,
      "loss": 2.284,
      "step": 54266
    },
    {
      "epoch": 3.057250619229903,
      "grad_norm": 0.4059627056121826,
      "learning_rate": 0.001,
      "loss": 2.2791,
      "step": 54309
    },
    {
      "epoch": 3.059671245215042,
      "grad_norm": 0.42024728655815125,
      "learning_rate": 0.001,
      "loss": 2.2788,
      "step": 54352
    },
    {
      "epoch": 3.06209187120018,
      "grad_norm": 0.35471323132514954,
      "learning_rate": 0.001,
      "loss": 2.28,
      "step": 54395
    },
    {
      "epoch": 3.064512497185319,
      "grad_norm": 0.3479269742965698,
      "learning_rate": 0.001,
      "loss": 2.2766,
      "step": 54438
    },
    {
      "epoch": 3.066933123170457,
      "grad_norm": 0.42797625064849854,
      "learning_rate": 0.001,
      "loss": 2.2791,
      "step": 54481
    },
    {
      "epoch": 3.069353749155596,
      "grad_norm": 0.4203141927719116,
      "learning_rate": 0.001,
      "loss": 2.2801,
      "step": 54524
    },
    {
      "epoch": 3.071774375140734,
      "grad_norm": 0.39509502053260803,
      "learning_rate": 0.001,
      "loss": 2.2816,
      "step": 54567
    },
    {
      "epoch": 3.0741950011258727,
      "grad_norm": 0.40240004658699036,
      "learning_rate": 0.001,
      "loss": 2.2864,
      "step": 54610
    },
    {
      "epoch": 3.076615627111011,
      "grad_norm": 0.4156615436077118,
      "learning_rate": 0.001,
      "loss": 2.2837,
      "step": 54653
    },
    {
      "epoch": 3.0790362530961497,
      "grad_norm": 0.3927633762359619,
      "learning_rate": 0.001,
      "loss": 2.2709,
      "step": 54696
    },
    {
      "epoch": 3.081456879081288,
      "grad_norm": 0.4687662720680237,
      "learning_rate": 0.001,
      "loss": 2.2918,
      "step": 54739
    },
    {
      "epoch": 3.0838775050664267,
      "grad_norm": 0.4581640362739563,
      "learning_rate": 0.001,
      "loss": 2.2752,
      "step": 54782
    },
    {
      "epoch": 3.086298131051565,
      "grad_norm": 0.3969971239566803,
      "learning_rate": 0.001,
      "loss": 2.2895,
      "step": 54825
    },
    {
      "epoch": 3.0887187570367036,
      "grad_norm": 0.40480953454971313,
      "learning_rate": 0.001,
      "loss": 2.2877,
      "step": 54868
    },
    {
      "epoch": 3.091139383021842,
      "grad_norm": 0.4049796164035797,
      "learning_rate": 0.001,
      "loss": 2.2722,
      "step": 54911
    },
    {
      "epoch": 3.0935600090069806,
      "grad_norm": 0.4506540894508362,
      "learning_rate": 0.001,
      "loss": 2.2939,
      "step": 54954
    },
    {
      "epoch": 3.095980634992119,
      "grad_norm": 0.43072324991226196,
      "learning_rate": 0.001,
      "loss": 2.2762,
      "step": 54997
    },
    {
      "epoch": 3.0984012609772575,
      "grad_norm": 0.44462889432907104,
      "learning_rate": 0.001,
      "loss": 2.2706,
      "step": 55040
    },
    {
      "epoch": 3.100821886962396,
      "grad_norm": 0.4317513406276703,
      "learning_rate": 0.001,
      "loss": 2.2807,
      "step": 55083
    },
    {
      "epoch": 3.1032425129475345,
      "grad_norm": 0.3678582012653351,
      "learning_rate": 0.001,
      "loss": 2.283,
      "step": 55126
    },
    {
      "epoch": 3.1056631389326728,
      "grad_norm": 0.36899274587631226,
      "learning_rate": 0.001,
      "loss": 2.2827,
      "step": 55169
    },
    {
      "epoch": 3.1080837649178115,
      "grad_norm": 0.4250926077365875,
      "learning_rate": 0.001,
      "loss": 2.2758,
      "step": 55212
    },
    {
      "epoch": 3.1105043909029497,
      "grad_norm": 0.3872144818305969,
      "learning_rate": 0.001,
      "loss": 2.2868,
      "step": 55255
    },
    {
      "epoch": 3.1129250168880884,
      "grad_norm": 0.49390658736228943,
      "learning_rate": 0.001,
      "loss": 2.2824,
      "step": 55298
    },
    {
      "epoch": 3.1153456428732267,
      "grad_norm": 0.43651673197746277,
      "learning_rate": 0.001,
      "loss": 2.292,
      "step": 55341
    },
    {
      "epoch": 3.1177662688583654,
      "grad_norm": 0.40253040194511414,
      "learning_rate": 0.001,
      "loss": 2.2852,
      "step": 55384
    },
    {
      "epoch": 3.1201868948435036,
      "grad_norm": 0.3711717426776886,
      "learning_rate": 0.001,
      "loss": 2.2775,
      "step": 55427
    },
    {
      "epoch": 3.121931997297906,
      "eval_ag_news_bleu_score": 4.463811250927712,
      "eval_ag_news_bleu_score_sem": 0.168033112994722,
      "eval_ag_news_emb_cos_sim": 0.9923207759857178,
      "eval_ag_news_emb_cos_sim_sem": 0.0002887287367665347,
      "eval_ag_news_emb_top1_equal": 0.6633333563804626,
      "eval_ag_news_emb_top1_equal_sem": 0.027329419768069584,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.5841269493103027,
      "eval_ag_news_n_ngrams_match_1": 5.2,
      "eval_ag_news_n_ngrams_match_2": 1.016,
      "eval_ag_news_n_ngrams_match_3": 0.272,
      "eval_ag_news_num_pred_words": 22.474,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2864033569422888,
      "eval_ag_news_runtime": 15.1276,
      "eval_ag_news_samples_per_second": 33.052,
      "eval_ag_news_steps_per_second": 0.132,
      "eval_ag_news_token_set_f1": 0.24360237673777146,
      "eval_ag_news_token_set_f1_sem": 0.00530865543177802,
      "eval_ag_news_token_set_precision": 0.23700207606658827,
      "eval_ag_news_token_set_recall": 0.2581792004289516,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 55458
    },
    {
      "epoch": 3.121931997297906,
      "eval_anthropic_toxic_prompts_bleu_score": 6.11500641819681,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.25726801410762007,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9918687343597412,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00025998854577881153,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6299999952316284,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027921293391044915,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8438472747802734,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.358,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.792,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.75,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.454,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.31614334343154743,
      "eval_anthropic_toxic_prompts_runtime": 6.7991,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.539,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.294,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3439254800088849,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006482462862408153,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4212571234669664,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.31343676724388886,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 55458
    },
    {
      "epoch": 3.121931997297906,
      "eval_arxiv_bleu_score": 4.038560504565108,
      "eval_arxiv_bleu_score_sem": 0.11807190409724984,
      "eval_arxiv_emb_cos_sim": 0.9917646646499634,
      "eval_arxiv_emb_cos_sim_sem": 0.00026831293809796475,
      "eval_arxiv_emb_top1_equal": 0.5366666913032532,
      "eval_arxiv_emb_top1_equal_sem": 0.02883789095764794,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.602611780166626,
      "eval_arxiv_n_ngrams_match_1": 5.302,
      "eval_arxiv_n_ngrams_match_2": 0.898,
      "eval_arxiv_n_ngrams_match_3": 0.16,
      "eval_arxiv_num_pred_words": 18.556,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.27793683003829783,
      "eval_arxiv_runtime": 7.3642,
      "eval_arxiv_samples_per_second": 67.896,
      "eval_arxiv_steps_per_second": 0.272,
      "eval_arxiv_token_set_f1": 0.25786911237776866,
      "eval_arxiv_token_set_f1_sem": 0.004716953022173594,
      "eval_arxiv_token_set_precision": 0.22809248485644412,
      "eval_arxiv_token_set_recall": 0.3126720695600152,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 55458
    },
    {
      "epoch": 3.121931997297906,
      "eval_python_code_alpaca_bleu_score": 6.762691376028305,
      "eval_python_code_alpaca_bleu_score_sem": 0.2742917808589299,
      "eval_python_code_alpaca_emb_cos_sim": 0.9900730848312378,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00038484725227993236,
      "eval_python_code_alpaca_emb_top1_equal": 0.6299999952316284,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027921293391044915,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.880096912384033,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.188,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.688,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.646,
      "eval_python_code_alpaca_num_pred_words": 20.536,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3423730379813872,
      "eval_python_code_alpaca_runtime": 6.6818,
      "eval_python_code_alpaca_samples_per_second": 74.83,
      "eval_python_code_alpaca_steps_per_second": 0.299,
      "eval_python_code_alpaca_token_set_f1": 0.3689783007817277,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006049050680449783,
      "eval_python_code_alpaca_token_set_precision": 0.3727247855628469,
      "eval_python_code_alpaca_token_set_recall": 0.3944960385780151,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 55458
    },
    {
      "epoch": 3.121931997297906,
      "eval_wikibio_bleu_score": 3.955702630936329,
      "eval_wikibio_bleu_score_sem": 0.14485842543525626,
      "eval_wikibio_emb_cos_sim": 0.9917250871658325,
      "eval_wikibio_emb_cos_sim_sem": 0.0002456820870017391,
      "eval_wikibio_emb_top1_equal": 0.6800000071525574,
      "eval_wikibio_emb_top1_equal_sem": 0.026977012481505278,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9534800052642822,
      "eval_wikibio_n_ngrams_match_1": 3.226,
      "eval_wikibio_n_ngrams_match_2": 0.884,
      "eval_wikibio_n_ngrams_match_3": 0.198,
      "eval_wikibio_num_pred_words": 19.838,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2834138757993222,
      "eval_wikibio_runtime": 6.6014,
      "eval_wikibio_samples_per_second": 75.741,
      "eval_wikibio_steps_per_second": 0.303,
      "eval_wikibio_token_set_f1": 0.20924024513403125,
      "eval_wikibio_token_set_f1_sem": 0.005607908048323222,
      "eval_wikibio_token_set_precision": 0.24412521880655932,
      "eval_wikibio_token_set_recall": 0.18917037017274482,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 55458
    },
    {
      "epoch": 3.121931997297906,
      "eval_nq_5round_bleu_score": 13.481314441836865,
      "eval_nq_5round_bleu_score_sem": 0.6191907086149322,
      "eval_nq_5round_emb_cos_sim": 0.9941333532333374,
      "eval_nq_5round_emb_cos_sim_sem": 0.00022140857786687347,
      "eval_nq_5round_emb_top1_equal": 0.6833333373069763,
      "eval_nq_5round_emb_top1_equal_sem": 0.02690183265038281,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.838,
      "eval_nq_5round_n_ngrams_match_2": 4.272,
      "eval_nq_5round_n_ngrams_match_3": 2.166,
      "eval_nq_5round_num_pred_words": 23.836,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.44318213573256293,
      "eval_nq_5round_token_set_f1": 0.46598355402788233,
      "eval_nq_5round_token_set_f1_sem": 0.0068726733647394705,
      "eval_nq_5round_token_set_precision": 0.44410299043665974,
      "eval_nq_5round_token_set_recall": 0.4984868372665212,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.819549491217579,
      "eval_nq_bleu_score_sem": 0.5238885942153766,
      "eval_nq_emb_cos_sim": 0.9938676357269287,
      "eval_nq_emb_cos_sim_sem": 0.00022610055574283762,
      "eval_nq_emb_top1_equal": 0.6800000071525574,
      "eval_nq_emb_top1_equal_sem": 0.026977012481505278,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.0298683643341064,
      "eval_nq_n_ngrams_match_1": 10.232,
      "eval_nq_n_ngrams_match_2": 3.81,
      "eval_nq_n_ngrams_match_3": 1.808,
      "eval_nq_num_pred_words": 23.712,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4144600733753244,
      "eval_nq_runtime": 18.8616,
      "eval_nq_samples_per_second": 26.509,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.44097597224060336,
      "eval_nq_token_set_f1_sem": 0.00644385489793395,
      "eval_nq_token_set_precision": 0.4185488839343076,
      "eval_nq_token_set_recall": 0.4762883729391119,
      "eval_nq_true_num_tokens": 32.0,
      "step": 55458
    },
    {
      "epoch": 3.1226075208286423,
      "grad_norm": 0.4427945613861084,
      "learning_rate": 0.001,
      "loss": 2.2809,
      "step": 55470
    },
    {
      "epoch": 3.1250281468137806,
      "grad_norm": 0.4289681315422058,
      "learning_rate": 0.001,
      "loss": 2.2792,
      "step": 55513
    },
    {
      "epoch": 3.1274487727989193,
      "grad_norm": 0.36305564641952515,
      "learning_rate": 0.001,
      "loss": 2.2867,
      "step": 55556
    },
    {
      "epoch": 3.1298693987840576,
      "grad_norm": 0.4429032802581787,
      "learning_rate": 0.001,
      "loss": 2.2817,
      "step": 55599
    },
    {
      "epoch": 3.1322900247691963,
      "grad_norm": 0.3783322274684906,
      "learning_rate": 0.001,
      "loss": 2.2755,
      "step": 55642
    },
    {
      "epoch": 3.1347106507543345,
      "grad_norm": 0.40916576981544495,
      "learning_rate": 0.001,
      "loss": 2.291,
      "step": 55685
    },
    {
      "epoch": 3.1371312767394732,
      "grad_norm": 0.42195039987564087,
      "learning_rate": 0.001,
      "loss": 2.2773,
      "step": 55728
    },
    {
      "epoch": 3.1395519027246115,
      "grad_norm": 0.42976686358451843,
      "learning_rate": 0.001,
      "loss": 2.2847,
      "step": 55771
    },
    {
      "epoch": 3.14197252870975,
      "grad_norm": 0.41050148010253906,
      "learning_rate": 0.001,
      "loss": 2.2742,
      "step": 55814
    },
    {
      "epoch": 3.1443931546948884,
      "grad_norm": 0.513163149356842,
      "learning_rate": 0.001,
      "loss": 2.2779,
      "step": 55857
    },
    {
      "epoch": 3.146813780680027,
      "grad_norm": 0.4128406345844269,
      "learning_rate": 0.001,
      "loss": 2.2906,
      "step": 55900
    },
    {
      "epoch": 3.1492344066651654,
      "grad_norm": 0.36547616124153137,
      "learning_rate": 0.001,
      "loss": 2.2722,
      "step": 55943
    },
    {
      "epoch": 3.151655032650304,
      "grad_norm": 0.4125881493091583,
      "learning_rate": 0.001,
      "loss": 2.29,
      "step": 55986
    },
    {
      "epoch": 3.1540756586354424,
      "grad_norm": 0.40098071098327637,
      "learning_rate": 0.001,
      "loss": 2.267,
      "step": 56029
    },
    {
      "epoch": 3.156496284620581,
      "grad_norm": 0.38548344373703003,
      "learning_rate": 0.001,
      "loss": 2.2833,
      "step": 56072
    },
    {
      "epoch": 3.1589169106057193,
      "grad_norm": 0.36711210012435913,
      "learning_rate": 0.001,
      "loss": 2.2765,
      "step": 56115
    },
    {
      "epoch": 3.161337536590858,
      "grad_norm": 0.37986627221107483,
      "learning_rate": 0.001,
      "loss": 2.2794,
      "step": 56158
    },
    {
      "epoch": 3.1637581625759963,
      "grad_norm": 0.39683371782302856,
      "learning_rate": 0.001,
      "loss": 2.2767,
      "step": 56201
    },
    {
      "epoch": 3.166178788561135,
      "grad_norm": 0.3973598778247833,
      "learning_rate": 0.001,
      "loss": 2.2736,
      "step": 56244
    },
    {
      "epoch": 3.1685994145462733,
      "grad_norm": 0.39236050844192505,
      "learning_rate": 0.001,
      "loss": 2.2725,
      "step": 56287
    },
    {
      "epoch": 3.171020040531412,
      "grad_norm": 0.4222198724746704,
      "learning_rate": 0.001,
      "loss": 2.2814,
      "step": 56330
    },
    {
      "epoch": 3.17344066651655,
      "grad_norm": 0.4188011884689331,
      "learning_rate": 0.001,
      "loss": 2.2763,
      "step": 56373
    },
    {
      "epoch": 3.175861292501689,
      "grad_norm": 0.44655972719192505,
      "learning_rate": 0.001,
      "loss": 2.2841,
      "step": 56416
    },
    {
      "epoch": 3.178281918486827,
      "grad_norm": 0.39385268092155457,
      "learning_rate": 0.001,
      "loss": 2.2858,
      "step": 56459
    },
    {
      "epoch": 3.180702544471966,
      "grad_norm": 0.38987618684768677,
      "learning_rate": 0.001,
      "loss": 2.2786,
      "step": 56502
    },
    {
      "epoch": 3.183123170457104,
      "grad_norm": 0.3684488832950592,
      "learning_rate": 0.001,
      "loss": 2.2815,
      "step": 56545
    },
    {
      "epoch": 3.185543796442243,
      "grad_norm": 0.4242568016052246,
      "learning_rate": 0.001,
      "loss": 2.2766,
      "step": 56588
    },
    {
      "epoch": 3.187964422427381,
      "grad_norm": 0.40748852491378784,
      "learning_rate": 0.001,
      "loss": 2.2776,
      "step": 56631
    },
    {
      "epoch": 3.19038504841252,
      "grad_norm": 0.39608103036880493,
      "learning_rate": 0.001,
      "loss": 2.276,
      "step": 56674
    },
    {
      "epoch": 3.192805674397658,
      "grad_norm": 0.35585010051727295,
      "learning_rate": 0.001,
      "loss": 2.2721,
      "step": 56717
    },
    {
      "epoch": 3.1952263003827968,
      "grad_norm": 0.3934483528137207,
      "learning_rate": 0.001,
      "loss": 2.2836,
      "step": 56760
    },
    {
      "epoch": 3.197646926367935,
      "grad_norm": 0.3660542368888855,
      "learning_rate": 0.001,
      "loss": 2.2783,
      "step": 56803
    },
    {
      "epoch": 3.2000675523530737,
      "grad_norm": 0.34695035219192505,
      "learning_rate": 0.001,
      "loss": 2.2702,
      "step": 56846
    },
    {
      "epoch": 3.202488178338212,
      "grad_norm": 0.4875432550907135,
      "learning_rate": 0.001,
      "loss": 2.2627,
      "step": 56889
    },
    {
      "epoch": 3.2049088043233507,
      "grad_norm": 0.46332278847694397,
      "learning_rate": 0.001,
      "loss": 2.2627,
      "step": 56932
    },
    {
      "epoch": 3.207329430308489,
      "grad_norm": 0.35458657145500183,
      "learning_rate": 0.001,
      "loss": 2.2738,
      "step": 56975
    },
    {
      "epoch": 3.2097500562936276,
      "grad_norm": 0.44589802622795105,
      "learning_rate": 0.001,
      "loss": 2.2694,
      "step": 57018
    },
    {
      "epoch": 3.212170682278766,
      "grad_norm": 0.398385614156723,
      "learning_rate": 0.001,
      "loss": 2.2736,
      "step": 57061
    },
    {
      "epoch": 3.2145913082639046,
      "grad_norm": 0.4530303478240967,
      "learning_rate": 0.001,
      "loss": 2.2758,
      "step": 57104
    },
    {
      "epoch": 3.217011934249043,
      "grad_norm": 0.4278252124786377,
      "learning_rate": 0.001,
      "loss": 2.284,
      "step": 57147
    },
    {
      "epoch": 3.2194325602341816,
      "grad_norm": 0.4007197618484497,
      "learning_rate": 0.001,
      "loss": 2.2729,
      "step": 57190
    },
    {
      "epoch": 3.22185318621932,
      "grad_norm": 0.3839760720729828,
      "learning_rate": 0.001,
      "loss": 2.2865,
      "step": 57233
    },
    {
      "epoch": 3.2242738122044585,
      "grad_norm": 0.38461920619010925,
      "learning_rate": 0.001,
      "loss": 2.2628,
      "step": 57276
    },
    {
      "epoch": 3.226694438189597,
      "grad_norm": 0.3688443899154663,
      "learning_rate": 0.001,
      "loss": 2.2721,
      "step": 57319
    },
    {
      "epoch": 3.2291150641747355,
      "grad_norm": 0.3745008707046509,
      "learning_rate": 0.001,
      "loss": 2.2743,
      "step": 57362
    },
    {
      "epoch": 3.2315356901598737,
      "grad_norm": 0.443968266248703,
      "learning_rate": 0.001,
      "loss": 2.2782,
      "step": 57405
    },
    {
      "epoch": 3.2339563161450124,
      "grad_norm": 0.3911362290382385,
      "learning_rate": 0.001,
      "loss": 2.2778,
      "step": 57448
    },
    {
      "epoch": 3.2363769421301507,
      "grad_norm": 0.4062647223472595,
      "learning_rate": 0.001,
      "loss": 2.2797,
      "step": 57491
    },
    {
      "epoch": 3.2387975681152894,
      "grad_norm": 0.40339791774749756,
      "learning_rate": 0.001,
      "loss": 2.2759,
      "step": 57534
    },
    {
      "epoch": 3.2412181941004277,
      "grad_norm": 0.464725524187088,
      "learning_rate": 0.001,
      "loss": 2.2853,
      "step": 57577
    },
    {
      "epoch": 3.242006304886287,
      "eval_ag_news_bleu_score": 4.566737027382573,
      "eval_ag_news_bleu_score_sem": 0.16913201267455408,
      "eval_ag_news_emb_cos_sim": 0.9920859932899475,
      "eval_ag_news_emb_cos_sim_sem": 0.00026927990970453705,
      "eval_ag_news_emb_top1_equal": 0.6466666460037231,
      "eval_ag_news_emb_top1_equal_sem": 0.027643749338232177,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.574152946472168,
      "eval_ag_news_n_ngrams_match_1": 5.332,
      "eval_ag_news_n_ngrams_match_2": 1.03,
      "eval_ag_news_n_ngrams_match_3": 0.256,
      "eval_ag_news_num_pred_words": 22.928,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2943063017750889,
      "eval_ag_news_runtime": 15.2843,
      "eval_ag_news_samples_per_second": 32.713,
      "eval_ag_news_steps_per_second": 0.131,
      "eval_ag_news_token_set_f1": 0.24665893208575226,
      "eval_ag_news_token_set_f1_sem": 0.005297173676743402,
      "eval_ag_news_token_set_precision": 0.24150548775066993,
      "eval_ag_news_token_set_recall": 0.2600347208712031,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 57591
    },
    {
      "epoch": 3.242006304886287,
      "eval_anthropic_toxic_prompts_bleu_score": 6.063154002837746,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2565394218141557,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9919100403785706,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002792840747592156,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6333333253860474,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02786867456387452,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8372559547424316,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.346,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.738,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.758,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.54,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.31436769927694713,
      "eval_anthropic_toxic_prompts_runtime": 6.8566,
      "eval_anthropic_toxic_prompts_samples_per_second": 72.923,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.292,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.34100692053452497,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.00666712419194127,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4222485604226453,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.308553718400697,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 57591
    },
    {
      "epoch": 3.242006304886287,
      "eval_arxiv_bleu_score": 3.9990690815928107,
      "eval_arxiv_bleu_score_sem": 0.11877393943603655,
      "eval_arxiv_emb_cos_sim": 0.9918750524520874,
      "eval_arxiv_emb_cos_sim_sem": 0.00023103908220828857,
      "eval_arxiv_emb_top1_equal": 0.6166666746139526,
      "eval_arxiv_emb_top1_equal_sem": 0.028117578599086417,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5912842750549316,
      "eval_arxiv_n_ngrams_match_1": 5.166,
      "eval_arxiv_n_ngrams_match_2": 0.858,
      "eval_arxiv_n_ngrams_match_3": 0.176,
      "eval_arxiv_num_pred_words": 18.478,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2749982435477911,
      "eval_arxiv_runtime": 6.7177,
      "eval_arxiv_samples_per_second": 74.43,
      "eval_arxiv_steps_per_second": 0.298,
      "eval_arxiv_token_set_f1": 0.25302553913236653,
      "eval_arxiv_token_set_f1_sem": 0.004913316047659322,
      "eval_arxiv_token_set_precision": 0.22365539214169716,
      "eval_arxiv_token_set_recall": 0.3054098025795269,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 57591
    },
    {
      "epoch": 3.242006304886287,
      "eval_python_code_alpaca_bleu_score": 6.885603728032104,
      "eval_python_code_alpaca_bleu_score_sem": 0.2586077731511652,
      "eval_python_code_alpaca_emb_cos_sim": 0.9910852313041687,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003090623749760815,
      "eval_python_code_alpaca_emb_top1_equal": 0.6066666841506958,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028250091805185003,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.8275249004364014,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.304,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.718,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.65,
      "eval_python_code_alpaca_num_pred_words": 20.376,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35028381823536126,
      "eval_python_code_alpaca_runtime": 14.0855,
      "eval_python_code_alpaca_samples_per_second": 35.497,
      "eval_python_code_alpaca_steps_per_second": 0.142,
      "eval_python_code_alpaca_token_set_f1": 0.3734175425771506,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006035483239816862,
      "eval_python_code_alpaca_token_set_precision": 0.37377252062776967,
      "eval_python_code_alpaca_token_set_recall": 0.4026315657002695,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 57591
    },
    {
      "epoch": 3.242006304886287,
      "eval_wikibio_bleu_score": 4.084694729058138,
      "eval_wikibio_bleu_score_sem": 0.14509510320288374,
      "eval_wikibio_emb_cos_sim": 0.9914524555206299,
      "eval_wikibio_emb_cos_sim_sem": 0.0002748704235036824,
      "eval_wikibio_emb_top1_equal": 0.6466666460037231,
      "eval_wikibio_emb_top1_equal_sem": 0.027643749338232177,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.912811517715454,
      "eval_wikibio_n_ngrams_match_1": 3.294,
      "eval_wikibio_n_ngrams_match_2": 0.932,
      "eval_wikibio_n_ngrams_match_3": 0.204,
      "eval_wikibio_num_pred_words": 20.05,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2963043891752096,
      "eval_wikibio_runtime": 7.3089,
      "eval_wikibio_samples_per_second": 68.41,
      "eval_wikibio_steps_per_second": 0.274,
      "eval_wikibio_token_set_f1": 0.21506181636800076,
      "eval_wikibio_token_set_f1_sem": 0.005426807484213645,
      "eval_wikibio_token_set_precision": 0.25234948730977674,
      "eval_wikibio_token_set_recall": 0.19460426972640021,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 57591
    },
    {
      "epoch": 3.242006304886287,
      "eval_nq_5round_bleu_score": 13.199800722874825,
      "eval_nq_5round_bleu_score_sem": 0.6195706448674273,
      "eval_nq_5round_emb_cos_sim": 0.9940650463104248,
      "eval_nq_5round_emb_cos_sim_sem": 0.000322100051507513,
      "eval_nq_5round_emb_top1_equal": 0.6499999761581421,
      "eval_nq_5round_emb_top1_equal_sem": 0.027583864257272155,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 10.876,
      "eval_nq_5round_n_ngrams_match_2": 4.226,
      "eval_nq_5round_n_ngrams_match_3": 2.098,
      "eval_nq_5round_num_pred_words": 23.778,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.44614857670705893,
      "eval_nq_5round_token_set_f1": 0.46639318779404615,
      "eval_nq_5round_token_set_f1_sem": 0.007061794445808627,
      "eval_nq_5round_token_set_precision": 0.44926423291112005,
      "eval_nq_5round_token_set_recall": 0.49470449630590585,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.735491631542843,
      "eval_nq_bleu_score_sem": 0.542163237791629,
      "eval_nq_emb_cos_sim": 0.99374920129776,
      "eval_nq_emb_cos_sim_sem": 0.00031211185612320744,
      "eval_nq_emb_top1_equal": 0.6499999761581421,
      "eval_nq_emb_top1_equal_sem": 0.027583864257272155,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.0188112258911133,
      "eval_nq_n_ngrams_match_1": 10.302,
      "eval_nq_n_ngrams_match_2": 3.784,
      "eval_nq_n_ngrams_match_3": 1.776,
      "eval_nq_num_pred_words": 23.96,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4180194383689945,
      "eval_nq_runtime": 19.1266,
      "eval_nq_samples_per_second": 26.142,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.44080516454032836,
      "eval_nq_token_set_f1_sem": 0.006526246600520513,
      "eval_nq_token_set_precision": 0.42133342024888853,
      "eval_nq_token_set_recall": 0.4721240780768545,
      "eval_nq_true_num_tokens": 32.0,
      "step": 57591
    },
    {
      "epoch": 3.2436388200855664,
      "grad_norm": 0.42209723591804504,
      "learning_rate": 0.001,
      "loss": 2.2641,
      "step": 57620
    },
    {
      "epoch": 3.2460594460707046,
      "grad_norm": 0.3870490491390228,
      "learning_rate": 0.001,
      "loss": 2.2848,
      "step": 57663
    },
    {
      "epoch": 3.2484800720558433,
      "grad_norm": 0.4374302923679352,
      "learning_rate": 0.001,
      "loss": 2.2682,
      "step": 57706
    },
    {
      "epoch": 3.2509006980409816,
      "grad_norm": 0.35285845398902893,
      "learning_rate": 0.001,
      "loss": 2.2749,
      "step": 57749
    },
    {
      "epoch": 3.2533213240261203,
      "grad_norm": 0.44102704524993896,
      "learning_rate": 0.001,
      "loss": 2.279,
      "step": 57792
    },
    {
      "epoch": 3.2557419500112585,
      "grad_norm": 0.43795058131217957,
      "learning_rate": 0.001,
      "loss": 2.2756,
      "step": 57835
    },
    {
      "epoch": 3.2581625759963972,
      "grad_norm": 0.4928187429904938,
      "learning_rate": 0.001,
      "loss": 2.2773,
      "step": 57878
    },
    {
      "epoch": 3.2605832019815355,
      "grad_norm": 0.4266267716884613,
      "learning_rate": 0.001,
      "loss": 2.2823,
      "step": 57921
    },
    {
      "epoch": 3.263003827966674,
      "grad_norm": 0.4568571150302887,
      "learning_rate": 0.001,
      "loss": 2.2672,
      "step": 57964
    },
    {
      "epoch": 3.2654244539518125,
      "grad_norm": 0.3819501996040344,
      "learning_rate": 0.001,
      "loss": 2.2681,
      "step": 58007
    },
    {
      "epoch": 3.267845079936951,
      "grad_norm": 0.38962218165397644,
      "learning_rate": 0.001,
      "loss": 2.2658,
      "step": 58050
    },
    {
      "epoch": 3.2702657059220894,
      "grad_norm": 0.38046994805336,
      "learning_rate": 0.001,
      "loss": 2.2733,
      "step": 58093
    },
    {
      "epoch": 3.272686331907228,
      "grad_norm": 0.40478062629699707,
      "learning_rate": 0.001,
      "loss": 2.2768,
      "step": 58136
    },
    {
      "epoch": 3.2751069578923664,
      "grad_norm": 0.3647924065589905,
      "learning_rate": 0.001,
      "loss": 2.2752,
      "step": 58179
    },
    {
      "epoch": 3.277527583877505,
      "grad_norm": 0.45673948526382446,
      "learning_rate": 0.001,
      "loss": 2.2813,
      "step": 58222
    },
    {
      "epoch": 3.279948209862644,
      "grad_norm": 0.509404182434082,
      "learning_rate": 0.001,
      "loss": 2.2885,
      "step": 58265
    },
    {
      "epoch": 3.282368835847782,
      "grad_norm": 0.40781915187835693,
      "learning_rate": 0.001,
      "loss": 2.2779,
      "step": 58308
    },
    {
      "epoch": 3.2847894618329203,
      "grad_norm": 0.40707191824913025,
      "learning_rate": 0.001,
      "loss": 2.2645,
      "step": 58351
    },
    {
      "epoch": 3.287210087818059,
      "grad_norm": 0.409408301115036,
      "learning_rate": 0.001,
      "loss": 2.2773,
      "step": 58394
    },
    {
      "epoch": 3.2896307138031977,
      "grad_norm": 0.40930864214897156,
      "learning_rate": 0.001,
      "loss": 2.28,
      "step": 58437
    },
    {
      "epoch": 3.292051339788336,
      "grad_norm": 0.41524896025657654,
      "learning_rate": 0.001,
      "loss": 2.2806,
      "step": 58480
    },
    {
      "epoch": 3.2944719657734742,
      "grad_norm": 0.4157450199127197,
      "learning_rate": 0.001,
      "loss": 2.2767,
      "step": 58523
    },
    {
      "epoch": 3.296892591758613,
      "grad_norm": 0.4288618862628937,
      "learning_rate": 0.001,
      "loss": 2.2804,
      "step": 58566
    },
    {
      "epoch": 3.2993132177437516,
      "grad_norm": 0.38753339648246765,
      "learning_rate": 0.001,
      "loss": 2.2783,
      "step": 58609
    },
    {
      "epoch": 3.30173384372889,
      "grad_norm": 0.4036675691604614,
      "learning_rate": 0.001,
      "loss": 2.2801,
      "step": 58652
    },
    {
      "epoch": 3.304154469714028,
      "grad_norm": 0.3815600872039795,
      "learning_rate": 0.001,
      "loss": 2.2688,
      "step": 58695
    },
    {
      "epoch": 3.306575095699167,
      "grad_norm": 0.3629373610019684,
      "learning_rate": 0.001,
      "loss": 2.269,
      "step": 58738
    },
    {
      "epoch": 3.3089957216843056,
      "grad_norm": 0.3697056174278259,
      "learning_rate": 0.001,
      "loss": 2.2729,
      "step": 58781
    },
    {
      "epoch": 3.311416347669444,
      "grad_norm": 0.40573838353157043,
      "learning_rate": 0.001,
      "loss": 2.2807,
      "step": 58824
    },
    {
      "epoch": 3.313836973654582,
      "grad_norm": 0.4398876130580902,
      "learning_rate": 0.001,
      "loss": 2.2772,
      "step": 58867
    },
    {
      "epoch": 3.3162575996397208,
      "grad_norm": 0.44173771142959595,
      "learning_rate": 0.001,
      "loss": 2.2618,
      "step": 58910
    },
    {
      "epoch": 3.3186782256248595,
      "grad_norm": 0.36918413639068604,
      "learning_rate": 0.001,
      "loss": 2.2727,
      "step": 58953
    },
    {
      "epoch": 3.3210988516099977,
      "grad_norm": 0.36295849084854126,
      "learning_rate": 0.001,
      "loss": 2.2693,
      "step": 58996
    },
    {
      "epoch": 3.323519477595136,
      "grad_norm": 0.4358343780040741,
      "learning_rate": 0.001,
      "loss": 2.2801,
      "step": 59039
    },
    {
      "epoch": 3.3259401035802747,
      "grad_norm": 0.4165166914463043,
      "learning_rate": 0.001,
      "loss": 2.2728,
      "step": 59082
    },
    {
      "epoch": 3.3283607295654134,
      "grad_norm": 0.3880733251571655,
      "learning_rate": 0.001,
      "loss": 2.2717,
      "step": 59125
    },
    {
      "epoch": 3.3307813555505517,
      "grad_norm": 0.38660842180252075,
      "learning_rate": 0.001,
      "loss": 2.2696,
      "step": 59168
    },
    {
      "epoch": 3.33320198153569,
      "grad_norm": 0.4538370966911316,
      "learning_rate": 0.001,
      "loss": 2.2771,
      "step": 59211
    },
    {
      "epoch": 3.3356226075208286,
      "grad_norm": 0.4286336302757263,
      "learning_rate": 0.001,
      "loss": 2.2674,
      "step": 59254
    },
    {
      "epoch": 3.3380432335059673,
      "grad_norm": 0.3763386309146881,
      "learning_rate": 0.001,
      "loss": 2.2697,
      "step": 59297
    },
    {
      "epoch": 3.3404638594911056,
      "grad_norm": 0.4499020576477051,
      "learning_rate": 0.001,
      "loss": 2.2621,
      "step": 59340
    },
    {
      "epoch": 3.3428844854762443,
      "grad_norm": 0.3854259252548218,
      "learning_rate": 0.001,
      "loss": 2.2731,
      "step": 59383
    },
    {
      "epoch": 3.3453051114613825,
      "grad_norm": 0.5042581558227539,
      "learning_rate": 0.001,
      "loss": 2.2543,
      "step": 59426
    },
    {
      "epoch": 3.3477257374465212,
      "grad_norm": 0.4334642291069031,
      "learning_rate": 0.001,
      "loss": 2.2724,
      "step": 59469
    },
    {
      "epoch": 3.3501463634316595,
      "grad_norm": 0.39893239736557007,
      "learning_rate": 0.001,
      "loss": 2.2697,
      "step": 59512
    },
    {
      "epoch": 3.352566989416798,
      "grad_norm": 0.4107424318790436,
      "learning_rate": 0.001,
      "loss": 2.2648,
      "step": 59555
    },
    {
      "epoch": 3.3549876154019365,
      "grad_norm": 0.47167596220970154,
      "learning_rate": 0.001,
      "loss": 2.273,
      "step": 59598
    },
    {
      "epoch": 3.357408241387075,
      "grad_norm": 0.4233260452747345,
      "learning_rate": 0.001,
      "loss": 2.2716,
      "step": 59641
    },
    {
      "epoch": 3.3598288673722134,
      "grad_norm": 0.40942907333374023,
      "learning_rate": 0.001,
      "loss": 2.2638,
      "step": 59684
    },
    {
      "epoch": 3.362080612474668,
      "eval_ag_news_bleu_score": 4.253594480996815,
      "eval_ag_news_bleu_score_sem": 0.15977589580488427,
      "eval_ag_news_emb_cos_sim": 0.9920467138290405,
      "eval_ag_news_emb_cos_sim_sem": 0.0003234927196808956,
      "eval_ag_news_emb_top1_equal": 0.6399999856948853,
      "eval_ag_news_emb_top1_equal_sem": 0.02775911810844162,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.546962022781372,
      "eval_ag_news_n_ngrams_match_1": 5.146,
      "eval_ag_news_n_ngrams_match_2": 0.966,
      "eval_ag_news_n_ngrams_match_3": 0.236,
      "eval_ag_news_num_pred_words": 22.482,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2865841073686378,
      "eval_ag_news_runtime": 15.3574,
      "eval_ag_news_samples_per_second": 32.558,
      "eval_ag_news_steps_per_second": 0.13,
      "eval_ag_news_token_set_f1": 0.24105603915540647,
      "eval_ag_news_token_set_f1_sem": 0.005151707496124049,
      "eval_ag_news_token_set_precision": 0.2340295537298839,
      "eval_ag_news_token_set_recall": 0.25812447297008384,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 59724
    },
    {
      "epoch": 3.362080612474668,
      "eval_anthropic_toxic_prompts_bleu_score": 6.039057020128317,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2445555226199016,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9920989274978638,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002642967541939361,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6633333563804626,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027329419768069584,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8090829849243164,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.408,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.782,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.722,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.922,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.31349336700956565,
      "eval_anthropic_toxic_prompts_runtime": 6.7012,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.614,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.298,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.34731750344166296,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0063611453457250895,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.42656072583262855,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3169032568402272,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 59724
    },
    {
      "epoch": 3.362080612474668,
      "eval_arxiv_bleu_score": 4.1357598348779705,
      "eval_arxiv_bleu_score_sem": 0.14023464649022782,
      "eval_arxiv_emb_cos_sim": 0.9917938709259033,
      "eval_arxiv_emb_cos_sim_sem": 0.0002639856413563001,
      "eval_arxiv_emb_top1_equal": 0.5699999928474426,
      "eval_arxiv_emb_top1_equal_sem": 0.02863097048683739,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5679094791412354,
      "eval_arxiv_n_ngrams_match_1": 5.264,
      "eval_arxiv_n_ngrams_match_2": 0.888,
      "eval_arxiv_n_ngrams_match_3": 0.192,
      "eval_arxiv_num_pred_words": 18.484,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.280830903329363,
      "eval_arxiv_runtime": 6.7706,
      "eval_arxiv_samples_per_second": 73.849,
      "eval_arxiv_steps_per_second": 0.295,
      "eval_arxiv_token_set_f1": 0.25680824557202125,
      "eval_arxiv_token_set_f1_sem": 0.00494503657369403,
      "eval_arxiv_token_set_precision": 0.22580955439085906,
      "eval_arxiv_token_set_recall": 0.31055214788912866,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 59724
    },
    {
      "epoch": 3.362080612474668,
      "eval_python_code_alpaca_bleu_score": 6.227616809196396,
      "eval_python_code_alpaca_bleu_score_sem": 0.22861774670072202,
      "eval_python_code_alpaca_emb_cos_sim": 0.9906556606292725,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003940686884066006,
      "eval_python_code_alpaca_emb_top1_equal": 0.6566666960716248,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027459639363867062,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.813305616378784,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.066,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.564,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.554,
      "eval_python_code_alpaca_num_pred_words": 20.574,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.33608046481066334,
      "eval_python_code_alpaca_runtime": 6.7574,
      "eval_python_code_alpaca_samples_per_second": 73.993,
      "eval_python_code_alpaca_steps_per_second": 0.296,
      "eval_python_code_alpaca_token_set_f1": 0.36229599968268017,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006105341725915066,
      "eval_python_code_alpaca_token_set_precision": 0.36134577861995854,
      "eval_python_code_alpaca_token_set_recall": 0.39210593101354335,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 59724
    },
    {
      "epoch": 3.362080612474668,
      "eval_wikibio_bleu_score": 4.1217981895010185,
      "eval_wikibio_bleu_score_sem": 0.1409163407264881,
      "eval_wikibio_emb_cos_sim": 0.9919973015785217,
      "eval_wikibio_emb_cos_sim_sem": 0.00021640902252642824,
      "eval_wikibio_emb_top1_equal": 0.6766666769981384,
      "eval_wikibio_emb_top1_equal_sem": 0.02705060760513538,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.956798553466797,
      "eval_wikibio_n_ngrams_match_1": 3.354,
      "eval_wikibio_n_ngrams_match_2": 0.936,
      "eval_wikibio_n_ngrams_match_3": 0.214,
      "eval_wikibio_num_pred_words": 19.98,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.2993765147876242,
      "eval_wikibio_runtime": 6.6458,
      "eval_wikibio_samples_per_second": 75.236,
      "eval_wikibio_steps_per_second": 0.301,
      "eval_wikibio_token_set_f1": 0.21903003414651956,
      "eval_wikibio_token_set_f1_sem": 0.005529863823059263,
      "eval_wikibio_token_set_precision": 0.25673838775962077,
      "eval_wikibio_token_set_recall": 0.19769556488575102,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 59724
    },
    {
      "epoch": 3.362080612474668,
      "eval_nq_5round_bleu_score": 14.024864134317639,
      "eval_nq_5round_bleu_score_sem": 0.6191661915194967,
      "eval_nq_5round_emb_cos_sim": 0.994175136089325,
      "eval_nq_5round_emb_cos_sim_sem": 0.00032528433387896545,
      "eval_nq_5round_emb_top1_equal": 0.6399999856948853,
      "eval_nq_5round_emb_top1_equal_sem": 0.02775911810844162,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.106,
      "eval_nq_5round_n_ngrams_match_2": 4.498,
      "eval_nq_5round_n_ngrams_match_3": 2.262,
      "eval_nq_5round_num_pred_words": 23.668,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4573233294017618,
      "eval_nq_5round_token_set_f1": 0.47621212783984257,
      "eval_nq_5round_token_set_f1_sem": 0.007228701675615152,
      "eval_nq_5round_token_set_precision": 0.4537508035188671,
      "eval_nq_5round_token_set_recall": 0.5107790901391581,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.90901419849154,
      "eval_nq_bleu_score_sem": 0.5205916541510857,
      "eval_nq_emb_cos_sim": 0.9941496849060059,
      "eval_nq_emb_cos_sim_sem": 0.0001966256120506404,
      "eval_nq_emb_top1_equal": 0.6766666769981384,
      "eval_nq_emb_top1_equal_sem": 0.02705060760513538,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 2.0085997581481934,
      "eval_nq_n_ngrams_match_1": 10.336,
      "eval_nq_n_ngrams_match_2": 3.872,
      "eval_nq_n_ngrams_match_3": 1.848,
      "eval_nq_num_pred_words": 23.902,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4212618271470049,
      "eval_nq_runtime": 18.9818,
      "eval_nq_samples_per_second": 26.341,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.4450014012386401,
      "eval_nq_token_set_f1_sem": 0.006469510388525012,
      "eval_nq_token_set_precision": 0.4227837056967317,
      "eval_nq_token_set_recall": 0.4809013364262706,
      "eval_nq_true_num_tokens": 32.0,
      "step": 59724
    },
    {
      "epoch": 3.362249493357352,
      "grad_norm": 0.38193047046661377,
      "learning_rate": 0.001,
      "loss": 2.2599,
      "step": 59727
    },
    {
      "epoch": 3.3646701193424904,
      "grad_norm": 0.4081433117389679,
      "learning_rate": 0.001,
      "loss": 2.2693,
      "step": 59770
    },
    {
      "epoch": 3.367090745327629,
      "grad_norm": 0.37456750869750977,
      "learning_rate": 0.001,
      "loss": 2.2625,
      "step": 59813
    },
    {
      "epoch": 3.3695113713127673,
      "grad_norm": 0.4225941002368927,
      "learning_rate": 0.001,
      "loss": 2.2704,
      "step": 59856
    },
    {
      "epoch": 3.371931997297906,
      "grad_norm": 0.42181047797203064,
      "learning_rate": 0.001,
      "loss": 2.2761,
      "step": 59899
    },
    {
      "epoch": 3.3743526232830443,
      "grad_norm": 0.4240216314792633,
      "learning_rate": 0.001,
      "loss": 2.2736,
      "step": 59942
    },
    {
      "epoch": 3.376773249268183,
      "grad_norm": 0.41545671224594116,
      "learning_rate": 0.001,
      "loss": 2.265,
      "step": 59985
    },
    {
      "epoch": 3.3791938752533213,
      "grad_norm": 0.4138946831226349,
      "learning_rate": 0.001,
      "loss": 2.2855,
      "step": 60028
    },
    {
      "epoch": 3.38161450123846,
      "grad_norm": 0.38716936111450195,
      "learning_rate": 0.001,
      "loss": 2.2726,
      "step": 60071
    },
    {
      "epoch": 3.3840351272235982,
      "grad_norm": 0.4821769595146179,
      "learning_rate": 0.001,
      "loss": 2.2713,
      "step": 60114
    },
    {
      "epoch": 3.386455753208737,
      "grad_norm": 0.4017936885356903,
      "learning_rate": 0.001,
      "loss": 2.2719,
      "step": 60157
    },
    {
      "epoch": 3.388876379193875,
      "grad_norm": 0.44197648763656616,
      "learning_rate": 0.001,
      "loss": 2.2683,
      "step": 60200
    },
    {
      "epoch": 3.391297005179014,
      "grad_norm": 0.39021047949790955,
      "learning_rate": 0.001,
      "loss": 2.2756,
      "step": 60243
    },
    {
      "epoch": 3.393717631164152,
      "grad_norm": 0.386944979429245,
      "learning_rate": 0.001,
      "loss": 2.2716,
      "step": 60286
    },
    {
      "epoch": 3.396138257149291,
      "grad_norm": 0.42880868911743164,
      "learning_rate": 0.001,
      "loss": 2.2693,
      "step": 60329
    },
    {
      "epoch": 3.398558883134429,
      "grad_norm": 0.37624773383140564,
      "learning_rate": 0.001,
      "loss": 2.266,
      "step": 60372
    },
    {
      "epoch": 3.400979509119568,
      "grad_norm": 0.35299769043922424,
      "learning_rate": 0.001,
      "loss": 2.2532,
      "step": 60415
    },
    {
      "epoch": 3.403400135104706,
      "grad_norm": 0.47689005732536316,
      "learning_rate": 0.001,
      "loss": 2.2607,
      "step": 60458
    },
    {
      "epoch": 3.4058207610898448,
      "grad_norm": 0.3746825158596039,
      "learning_rate": 0.001,
      "loss": 2.2771,
      "step": 60501
    },
    {
      "epoch": 3.408241387074983,
      "grad_norm": 0.40578556060791016,
      "learning_rate": 0.001,
      "loss": 2.2602,
      "step": 60544
    },
    {
      "epoch": 3.4106620130601217,
      "grad_norm": 0.4499170184135437,
      "learning_rate": 0.001,
      "loss": 2.2646,
      "step": 60587
    },
    {
      "epoch": 3.41308263904526,
      "grad_norm": 0.38451746106147766,
      "learning_rate": 0.001,
      "loss": 2.2723,
      "step": 60630
    },
    {
      "epoch": 3.4155032650303987,
      "grad_norm": 0.4169243276119232,
      "learning_rate": 0.001,
      "loss": 2.2732,
      "step": 60673
    },
    {
      "epoch": 3.417923891015537,
      "grad_norm": 0.4175083041191101,
      "learning_rate": 0.001,
      "loss": 2.2632,
      "step": 60716
    },
    {
      "epoch": 3.4203445170006757,
      "grad_norm": 0.4085213840007782,
      "learning_rate": 0.001,
      "loss": 2.2778,
      "step": 60759
    },
    {
      "epoch": 3.422765142985814,
      "grad_norm": 0.4703681766986847,
      "learning_rate": 0.001,
      "loss": 2.2712,
      "step": 60802
    },
    {
      "epoch": 3.4251857689709526,
      "grad_norm": 0.4026029706001282,
      "learning_rate": 0.001,
      "loss": 2.2727,
      "step": 60845
    },
    {
      "epoch": 3.427606394956091,
      "grad_norm": 0.37225058674812317,
      "learning_rate": 0.001,
      "loss": 2.2723,
      "step": 60888
    },
    {
      "epoch": 3.4300270209412296,
      "grad_norm": 0.43151357769966125,
      "learning_rate": 0.001,
      "loss": 2.2742,
      "step": 60931
    },
    {
      "epoch": 3.432447646926368,
      "grad_norm": 0.37508532404899597,
      "learning_rate": 0.001,
      "loss": 2.2779,
      "step": 60974
    },
    {
      "epoch": 3.4348682729115065,
      "grad_norm": 0.3926122188568115,
      "learning_rate": 0.001,
      "loss": 2.2935,
      "step": 61017
    },
    {
      "epoch": 3.437288898896645,
      "grad_norm": 0.42195338010787964,
      "learning_rate": 0.001,
      "loss": 2.2552,
      "step": 61060
    },
    {
      "epoch": 3.4397095248817835,
      "grad_norm": 0.3882333040237427,
      "learning_rate": 0.001,
      "loss": 2.269,
      "step": 61103
    },
    {
      "epoch": 3.4421301508669218,
      "grad_norm": 0.4180922508239746,
      "learning_rate": 0.001,
      "loss": 2.2521,
      "step": 61146
    },
    {
      "epoch": 3.4445507768520605,
      "grad_norm": 0.4202941358089447,
      "learning_rate": 0.001,
      "loss": 2.2635,
      "step": 61189
    },
    {
      "epoch": 3.4469714028371987,
      "grad_norm": 0.4511958658695221,
      "learning_rate": 0.001,
      "loss": 2.2711,
      "step": 61232
    },
    {
      "epoch": 3.4493920288223374,
      "grad_norm": 0.37935033440589905,
      "learning_rate": 0.001,
      "loss": 2.2678,
      "step": 61275
    },
    {
      "epoch": 3.4518126548074757,
      "grad_norm": 0.4097431004047394,
      "learning_rate": 0.001,
      "loss": 2.2592,
      "step": 61318
    },
    {
      "epoch": 3.4542332807926144,
      "grad_norm": 0.3910957872867584,
      "learning_rate": 0.001,
      "loss": 2.2628,
      "step": 61361
    },
    {
      "epoch": 3.4566539067777526,
      "grad_norm": 0.36142823100090027,
      "learning_rate": 0.001,
      "loss": 2.2713,
      "step": 61404
    },
    {
      "epoch": 3.4590745327628913,
      "grad_norm": 0.395727276802063,
      "learning_rate": 0.001,
      "loss": 2.268,
      "step": 61447
    },
    {
      "epoch": 3.4614951587480296,
      "grad_norm": 0.42044541239738464,
      "learning_rate": 0.001,
      "loss": 2.2602,
      "step": 61490
    },
    {
      "epoch": 3.4639157847331683,
      "grad_norm": 0.42336758971214294,
      "learning_rate": 0.001,
      "loss": 2.2673,
      "step": 61533
    },
    {
      "epoch": 3.4663364107183066,
      "grad_norm": 0.5023131370544434,
      "learning_rate": 0.001,
      "loss": 2.2582,
      "step": 61576
    },
    {
      "epoch": 3.4687570367034453,
      "grad_norm": 0.430204838514328,
      "learning_rate": 0.001,
      "loss": 2.2693,
      "step": 61619
    },
    {
      "epoch": 3.4711776626885835,
      "grad_norm": 0.4816841185092926,
      "learning_rate": 0.001,
      "loss": 2.2655,
      "step": 61662
    },
    {
      "epoch": 3.4735982886737222,
      "grad_norm": 0.3531249463558197,
      "learning_rate": 0.001,
      "loss": 2.2689,
      "step": 61705
    },
    {
      "epoch": 3.4760189146588605,
      "grad_norm": 0.36051568388938904,
      "learning_rate": 0.001,
      "loss": 2.2684,
      "step": 61748
    },
    {
      "epoch": 3.478439540643999,
      "grad_norm": 0.4140871465206146,
      "learning_rate": 0.001,
      "loss": 2.2617,
      "step": 61791
    },
    {
      "epoch": 3.4808601666291374,
      "grad_norm": 0.40363451838493347,
      "learning_rate": 0.001,
      "loss": 2.2626,
      "step": 61834
    },
    {
      "epoch": 3.482154920063049,
      "eval_ag_news_bleu_score": 4.592406937386719,
      "eval_ag_news_bleu_score_sem": 0.17408615727706403,
      "eval_ag_news_emb_cos_sim": 0.9918776750564575,
      "eval_ag_news_emb_cos_sim_sem": 0.0003151860445549676,
      "eval_ag_news_emb_top1_equal": 0.6466666460037231,
      "eval_ag_news_emb_top1_equal_sem": 0.027643749338232177,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.5662732124328613,
      "eval_ag_news_n_ngrams_match_1": 5.316,
      "eval_ag_news_n_ngrams_match_2": 1.09,
      "eval_ag_news_n_ngrams_match_3": 0.29,
      "eval_ag_news_num_pred_words": 22.61,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.29085105172794645,
      "eval_ag_news_runtime": 15.656,
      "eval_ag_news_samples_per_second": 31.937,
      "eval_ag_news_steps_per_second": 0.128,
      "eval_ag_news_token_set_f1": 0.24766104354684895,
      "eval_ag_news_token_set_f1_sem": 0.005361691491385178,
      "eval_ag_news_token_set_precision": 0.24149751120457544,
      "eval_ag_news_token_set_recall": 0.26354939196717103,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 61857
    },
    {
      "epoch": 3.482154920063049,
      "eval_anthropic_toxic_prompts_bleu_score": 6.115252061256007,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.25174270058962817,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9919776916503906,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002631127671340802,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6399999856948853,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02775911810844162,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8027114868164062,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.394,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.794,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.744,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.36,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.31651045070611183,
      "eval_anthropic_toxic_prompts_runtime": 6.7643,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.918,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.296,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3447530375594758,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006540558727777392,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4223494607097239,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.31559978562243807,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 61857
    },
    {
      "epoch": 3.482154920063049,
      "eval_arxiv_bleu_score": 3.9432209121702235,
      "eval_arxiv_bleu_score_sem": 0.12038345777730715,
      "eval_arxiv_emb_cos_sim": 0.9915634393692017,
      "eval_arxiv_emb_cos_sim_sem": 0.0002444167460318688,
      "eval_arxiv_emb_top1_equal": 0.54666668176651,
      "eval_arxiv_emb_top1_equal_sem": 0.028789527267967893,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5611441135406494,
      "eval_arxiv_n_ngrams_match_1": 5.084,
      "eval_arxiv_n_ngrams_match_2": 0.862,
      "eval_arxiv_n_ngrams_match_3": 0.152,
      "eval_arxiv_num_pred_words": 18.282,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.27350966286025624,
      "eval_arxiv_runtime": 6.8313,
      "eval_arxiv_samples_per_second": 73.192,
      "eval_arxiv_steps_per_second": 0.293,
      "eval_arxiv_token_set_f1": 0.24692706457702912,
      "eval_arxiv_token_set_f1_sem": 0.005168240322098627,
      "eval_arxiv_token_set_precision": 0.21790142490591496,
      "eval_arxiv_token_set_recall": 0.2986079689614618,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 61857
    },
    {
      "epoch": 3.482154920063049,
      "eval_python_code_alpaca_bleu_score": 6.551281133521306,
      "eval_python_code_alpaca_bleu_score_sem": 0.2606007540865453,
      "eval_python_code_alpaca_emb_cos_sim": 0.9905633926391602,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00033792639812161245,
      "eval_python_code_alpaca_emb_top1_equal": 0.6433333158493042,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02770216499443815,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.828425407409668,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.092,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.628,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.604,
      "eval_python_code_alpaca_num_pred_words": 19.976,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3399403371562746,
      "eval_python_code_alpaca_runtime": 6.9423,
      "eval_python_code_alpaca_samples_per_second": 72.022,
      "eval_python_code_alpaca_steps_per_second": 0.288,
      "eval_python_code_alpaca_token_set_f1": 0.3639751491339863,
      "eval_python_code_alpaca_token_set_f1_sem": 0.00605227793547412,
      "eval_python_code_alpaca_token_set_precision": 0.3677042202618977,
      "eval_python_code_alpaca_token_set_recall": 0.39138288966892343,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 61857
    },
    {
      "epoch": 3.482154920063049,
      "eval_wikibio_bleu_score": 4.048569539500962,
      "eval_wikibio_bleu_score_sem": 0.1400994563245404,
      "eval_wikibio_emb_cos_sim": 0.9913551807403564,
      "eval_wikibio_emb_cos_sim_sem": 0.0003050146012358774,
      "eval_wikibio_emb_top1_equal": 0.6433333158493042,
      "eval_wikibio_emb_top1_equal_sem": 0.027702163273800266,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.915915012359619,
      "eval_wikibio_n_ngrams_match_1": 3.32,
      "eval_wikibio_n_ngrams_match_2": 0.9,
      "eval_wikibio_n_ngrams_match_3": 0.192,
      "eval_wikibio_num_pred_words": 19.458,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3029667064221614,
      "eval_wikibio_runtime": 7.1275,
      "eval_wikibio_samples_per_second": 70.151,
      "eval_wikibio_steps_per_second": 0.281,
      "eval_wikibio_token_set_f1": 0.21653499765551248,
      "eval_wikibio_token_set_f1_sem": 0.005635535031552854,
      "eval_wikibio_token_set_precision": 0.2522457893415271,
      "eval_wikibio_token_set_recall": 0.195491354208153,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 61857
    },
    {
      "epoch": 3.482154920063049,
      "eval_nq_5round_bleu_score": 14.39838738355668,
      "eval_nq_5round_bleu_score_sem": 0.6347399331223192,
      "eval_nq_5round_emb_cos_sim": 0.9944877624511719,
      "eval_nq_5round_emb_cos_sim_sem": 0.00020293498953878768,
      "eval_nq_5round_emb_top1_equal": 0.6499999761581421,
      "eval_nq_5round_emb_top1_equal_sem": 0.027583864257272155,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.284,
      "eval_nq_5round_n_ngrams_match_2": 4.62,
      "eval_nq_5round_n_ngrams_match_3": 2.342,
      "eval_nq_5round_num_pred_words": 23.826,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.46243942618344275,
      "eval_nq_5round_token_set_f1": 0.48413138936793376,
      "eval_nq_5round_token_set_f1_sem": 0.007067534955695961,
      "eval_nq_5round_token_set_precision": 0.462372141575449,
      "eval_nq_5round_token_set_recall": 0.5163080744866972,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 11.940486965054774,
      "eval_nq_bleu_score_sem": 0.5434811868939892,
      "eval_nq_emb_cos_sim": 0.9941467642784119,
      "eval_nq_emb_cos_sim_sem": 0.00020678597876334394,
      "eval_nq_emb_top1_equal": 0.6333333253860474,
      "eval_nq_emb_top1_equal_sem": 0.02786867456387452,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.995790958404541,
      "eval_nq_n_ngrams_match_1": 10.39,
      "eval_nq_n_ngrams_match_2": 3.884,
      "eval_nq_n_ngrams_match_3": 1.816,
      "eval_nq_num_pred_words": 23.908,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.42302369519928396,
      "eval_nq_runtime": 19.7763,
      "eval_nq_samples_per_second": 25.283,
      "eval_nq_steps_per_second": 0.101,
      "eval_nq_token_set_f1": 0.447904950689386,
      "eval_nq_token_set_f1_sem": 0.006418325069570695,
      "eval_nq_token_set_precision": 0.42817569700047325,
      "eval_nq_token_set_recall": 0.4789726001491619,
      "eval_nq_true_num_tokens": 32.0,
      "step": 61857
    },
    {
      "epoch": 3.483280792614276,
      "grad_norm": 0.3743383586406708,
      "learning_rate": 0.001,
      "loss": 2.2698,
      "step": 61877
    },
    {
      "epoch": 3.4857014185994144,
      "grad_norm": 0.3993799090385437,
      "learning_rate": 0.001,
      "loss": 2.2599,
      "step": 61920
    },
    {
      "epoch": 3.488122044584553,
      "grad_norm": 0.4400752782821655,
      "learning_rate": 0.001,
      "loss": 2.2505,
      "step": 61963
    },
    {
      "epoch": 3.4905426705696914,
      "grad_norm": 0.3988088369369507,
      "learning_rate": 0.001,
      "loss": 2.2569,
      "step": 62006
    },
    {
      "epoch": 3.49296329655483,
      "grad_norm": 0.39118778705596924,
      "learning_rate": 0.001,
      "loss": 2.2663,
      "step": 62049
    },
    {
      "epoch": 3.4953839225399683,
      "grad_norm": 0.4103110432624817,
      "learning_rate": 0.001,
      "loss": 2.257,
      "step": 62092
    },
    {
      "epoch": 3.497804548525107,
      "grad_norm": 0.3916657567024231,
      "learning_rate": 0.001,
      "loss": 2.2621,
      "step": 62135
    },
    {
      "epoch": 3.5002251745102453,
      "grad_norm": 0.4620775878429413,
      "learning_rate": 0.001,
      "loss": 2.2638,
      "step": 62178
    },
    {
      "epoch": 3.502645800495384,
      "grad_norm": 0.441123366355896,
      "learning_rate": 0.001,
      "loss": 2.2638,
      "step": 62221
    },
    {
      "epoch": 3.5050664264805222,
      "grad_norm": 0.36567312479019165,
      "learning_rate": 0.001,
      "loss": 2.2651,
      "step": 62264
    },
    {
      "epoch": 3.507487052465661,
      "grad_norm": 0.46143287420272827,
      "learning_rate": 0.001,
      "loss": 2.2595,
      "step": 62307
    },
    {
      "epoch": 3.509907678450799,
      "grad_norm": 0.4032467007637024,
      "learning_rate": 0.001,
      "loss": 2.2633,
      "step": 62350
    },
    {
      "epoch": 3.512328304435938,
      "grad_norm": 0.37715962529182434,
      "learning_rate": 0.001,
      "loss": 2.2558,
      "step": 62393
    },
    {
      "epoch": 3.514748930421076,
      "grad_norm": 0.35154837369918823,
      "learning_rate": 0.001,
      "loss": 2.2681,
      "step": 62436
    },
    {
      "epoch": 3.517169556406215,
      "grad_norm": 0.4070186913013458,
      "learning_rate": 0.001,
      "loss": 2.2671,
      "step": 62479
    },
    {
      "epoch": 3.5195901823913536,
      "grad_norm": 0.47789594531059265,
      "learning_rate": 0.001,
      "loss": 2.2567,
      "step": 62522
    },
    {
      "epoch": 3.522010808376492,
      "grad_norm": 0.37221869826316833,
      "learning_rate": 0.001,
      "loss": 2.2582,
      "step": 62565
    },
    {
      "epoch": 3.52443143436163,
      "grad_norm": 0.3691619336605072,
      "learning_rate": 0.001,
      "loss": 2.2673,
      "step": 62608
    },
    {
      "epoch": 3.526852060346769,
      "grad_norm": 0.4336853325366974,
      "learning_rate": 0.001,
      "loss": 2.2608,
      "step": 62651
    },
    {
      "epoch": 3.5292726863319075,
      "grad_norm": 0.3684805929660797,
      "learning_rate": 0.001,
      "loss": 2.2461,
      "step": 62694
    },
    {
      "epoch": 3.5316933123170458,
      "grad_norm": 0.41532576084136963,
      "learning_rate": 0.001,
      "loss": 2.2589,
      "step": 62737
    },
    {
      "epoch": 3.534113938302184,
      "grad_norm": 0.4495520293712616,
      "learning_rate": 0.001,
      "loss": 2.2591,
      "step": 62780
    },
    {
      "epoch": 3.5365345642873227,
      "grad_norm": 0.4130357503890991,
      "learning_rate": 0.001,
      "loss": 2.2604,
      "step": 62823
    },
    {
      "epoch": 3.5389551902724614,
      "grad_norm": 0.4609544575214386,
      "learning_rate": 0.001,
      "loss": 2.2561,
      "step": 62866
    },
    {
      "epoch": 3.5413758162575997,
      "grad_norm": 0.4158463180065155,
      "learning_rate": 0.001,
      "loss": 2.2701,
      "step": 62909
    },
    {
      "epoch": 3.543796442242738,
      "grad_norm": 0.39426344633102417,
      "learning_rate": 0.001,
      "loss": 2.2585,
      "step": 62952
    },
    {
      "epoch": 3.5462170682278766,
      "grad_norm": 0.38473188877105713,
      "learning_rate": 0.001,
      "loss": 2.2585,
      "step": 62995
    },
    {
      "epoch": 3.5486376942130153,
      "grad_norm": 0.3969265818595886,
      "learning_rate": 0.001,
      "loss": 2.251,
      "step": 63038
    },
    {
      "epoch": 3.5510583201981536,
      "grad_norm": 0.3863164186477661,
      "learning_rate": 0.001,
      "loss": 2.2641,
      "step": 63081
    },
    {
      "epoch": 3.553478946183292,
      "grad_norm": 0.4394087791442871,
      "learning_rate": 0.001,
      "loss": 2.2623,
      "step": 63124
    },
    {
      "epoch": 3.5558995721684306,
      "grad_norm": 0.4093898832798004,
      "learning_rate": 0.001,
      "loss": 2.2533,
      "step": 63167
    },
    {
      "epoch": 3.5583201981535693,
      "grad_norm": 0.4008769690990448,
      "learning_rate": 0.001,
      "loss": 2.2688,
      "step": 63210
    },
    {
      "epoch": 3.5607408241387075,
      "grad_norm": 0.37627992033958435,
      "learning_rate": 0.001,
      "loss": 2.254,
      "step": 63253
    },
    {
      "epoch": 3.5631614501238458,
      "grad_norm": 0.3811394274234772,
      "learning_rate": 0.001,
      "loss": 2.2571,
      "step": 63296
    },
    {
      "epoch": 3.5655820761089845,
      "grad_norm": 0.3896208107471466,
      "learning_rate": 0.001,
      "loss": 2.2538,
      "step": 63339
    },
    {
      "epoch": 3.568002702094123,
      "grad_norm": 0.48971933126449585,
      "learning_rate": 0.001,
      "loss": 2.2598,
      "step": 63382
    },
    {
      "epoch": 3.5704233280792614,
      "grad_norm": 0.37416917085647583,
      "learning_rate": 0.001,
      "loss": 2.264,
      "step": 63425
    },
    {
      "epoch": 3.5728439540643997,
      "grad_norm": 0.4368751049041748,
      "learning_rate": 0.001,
      "loss": 2.2464,
      "step": 63468
    },
    {
      "epoch": 3.5752645800495384,
      "grad_norm": 0.4109216630458832,
      "learning_rate": 0.001,
      "loss": 2.2603,
      "step": 63511
    },
    {
      "epoch": 3.577685206034677,
      "grad_norm": 0.35259512066841125,
      "learning_rate": 0.001,
      "loss": 2.263,
      "step": 63554
    },
    {
      "epoch": 3.5801058320198154,
      "grad_norm": 0.3911515176296234,
      "learning_rate": 0.001,
      "loss": 2.2707,
      "step": 63597
    },
    {
      "epoch": 3.5825264580049536,
      "grad_norm": 0.46080145239830017,
      "learning_rate": 0.001,
      "loss": 2.257,
      "step": 63640
    },
    {
      "epoch": 3.5849470839900923,
      "grad_norm": 0.47051459550857544,
      "learning_rate": 0.001,
      "loss": 2.2609,
      "step": 63683
    },
    {
      "epoch": 3.587367709975231,
      "grad_norm": 0.3768419623374939,
      "learning_rate": 0.001,
      "loss": 2.2541,
      "step": 63726
    },
    {
      "epoch": 3.5897883359603693,
      "grad_norm": 0.38839519023895264,
      "learning_rate": 0.001,
      "loss": 2.2681,
      "step": 63769
    },
    {
      "epoch": 3.5922089619455075,
      "grad_norm": 0.39023858308792114,
      "learning_rate": 0.001,
      "loss": 2.248,
      "step": 63812
    },
    {
      "epoch": 3.5946295879306462,
      "grad_norm": 0.3818659782409668,
      "learning_rate": 0.001,
      "loss": 2.2436,
      "step": 63855
    },
    {
      "epoch": 3.597050213915785,
      "grad_norm": 0.3908728063106537,
      "learning_rate": 0.001,
      "loss": 2.2657,
      "step": 63898
    },
    {
      "epoch": 3.599470839900923,
      "grad_norm": 0.40444472432136536,
      "learning_rate": 0.001,
      "loss": 2.2594,
      "step": 63941
    },
    {
      "epoch": 3.6018914658860615,
      "grad_norm": 0.4378105401992798,
      "learning_rate": 0.001,
      "loss": 2.2556,
      "step": 63984
    },
    {
      "epoch": 3.6022292276514296,
      "eval_ag_news_bleu_score": 4.48246606979602,
      "eval_ag_news_bleu_score_sem": 0.1745150357306402,
      "eval_ag_news_emb_cos_sim": 0.9922311902046204,
      "eval_ag_news_emb_cos_sim_sem": 0.0002923678858939467,
      "eval_ag_news_emb_top1_equal": 0.6366666555404663,
      "eval_ag_news_emb_top1_equal_sem": 0.02781461728343201,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.552459955215454,
      "eval_ag_news_n_ngrams_match_1": 5.258,
      "eval_ag_news_n_ngrams_match_2": 1.018,
      "eval_ag_news_n_ngrams_match_3": 0.284,
      "eval_ag_news_num_pred_words": 22.642,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.29414681109099594,
      "eval_ag_news_runtime": 19.0366,
      "eval_ag_news_samples_per_second": 26.265,
      "eval_ag_news_steps_per_second": 0.105,
      "eval_ag_news_token_set_f1": 0.24526590905102363,
      "eval_ag_news_token_set_f1_sem": 0.005312648572028589,
      "eval_ag_news_token_set_precision": 0.23830783418668502,
      "eval_ag_news_token_set_recall": 0.2620635294722279,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 63990
    },
    {
      "epoch": 3.6022292276514296,
      "eval_anthropic_toxic_prompts_bleu_score": 6.291981712606197,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2602047136101575,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9921036958694458,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002643629449825863,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6433333158493042,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027702163273800266,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.7977445125579834,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.438,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.842,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.808,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.646,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3187473014355515,
      "eval_anthropic_toxic_prompts_runtime": 6.9821,
      "eval_anthropic_toxic_prompts_samples_per_second": 71.611,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.286,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.35170200712427213,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0066700899774542344,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.43167686011513184,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3198746985938593,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 63990
    },
    {
      "epoch": 3.6022292276514296,
      "eval_arxiv_bleu_score": 4.052823523626996,
      "eval_arxiv_bleu_score_sem": 0.11938612152231615,
      "eval_arxiv_emb_cos_sim": 0.9915632605552673,
      "eval_arxiv_emb_cos_sim_sem": 0.000273649120109689,
      "eval_arxiv_emb_top1_equal": 0.5799999833106995,
      "eval_arxiv_emb_top1_equal_sem": 0.02854322483723857,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.582068681716919,
      "eval_arxiv_n_ngrams_match_1": 5.21,
      "eval_arxiv_n_ngrams_match_2": 0.93,
      "eval_arxiv_n_ngrams_match_3": 0.178,
      "eval_arxiv_num_pred_words": 18.524,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.27608161764119843,
      "eval_arxiv_runtime": 7.234,
      "eval_arxiv_samples_per_second": 69.118,
      "eval_arxiv_steps_per_second": 0.276,
      "eval_arxiv_token_set_f1": 0.2549872991777447,
      "eval_arxiv_token_set_f1_sem": 0.005122658205452698,
      "eval_arxiv_token_set_precision": 0.22389603499398186,
      "eval_arxiv_token_set_recall": 0.3162243529401762,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 63990
    },
    {
      "epoch": 3.6022292276514296,
      "eval_python_code_alpaca_bleu_score": 6.724081171279952,
      "eval_python_code_alpaca_bleu_score_sem": 0.24419942096285832,
      "eval_python_code_alpaca_emb_cos_sim": 0.9907143115997314,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00030316507681898036,
      "eval_python_code_alpaca_emb_top1_equal": 0.5633333325386047,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.028682842557165564,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.836130142211914,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.41,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.674,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.598,
      "eval_python_code_alpaca_num_pred_words": 20.456,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35591482081232684,
      "eval_python_code_alpaca_runtime": 6.8064,
      "eval_python_code_alpaca_samples_per_second": 73.46,
      "eval_python_code_alpaca_steps_per_second": 0.294,
      "eval_python_code_alpaca_token_set_f1": 0.3807816399065913,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005944103581252124,
      "eval_python_code_alpaca_token_set_precision": 0.3837111674921902,
      "eval_python_code_alpaca_token_set_recall": 0.40434788949724604,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 63990
    },
    {
      "epoch": 3.6022292276514296,
      "eval_wikibio_bleu_score": 4.148522119484862,
      "eval_wikibio_bleu_score_sem": 0.15571879183204393,
      "eval_wikibio_emb_cos_sim": 0.9917064905166626,
      "eval_wikibio_emb_cos_sim_sem": 0.00025553308839965886,
      "eval_wikibio_emb_top1_equal": 0.6600000262260437,
      "eval_wikibio_emb_top1_equal_sem": 0.027395285786318915,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.8989930152893066,
      "eval_wikibio_n_ngrams_match_1": 3.356,
      "eval_wikibio_n_ngrams_match_2": 0.944,
      "eval_wikibio_n_ngrams_match_3": 0.226,
      "eval_wikibio_num_pred_words": 20.092,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3050591069404228,
      "eval_wikibio_runtime": 6.6651,
      "eval_wikibio_samples_per_second": 75.017,
      "eval_wikibio_steps_per_second": 0.3,
      "eval_wikibio_token_set_f1": 0.2168275824423016,
      "eval_wikibio_token_set_f1_sem": 0.005651311118956313,
      "eval_wikibio_token_set_precision": 0.25412110320595765,
      "eval_wikibio_token_set_recall": 0.1965006738838555,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 63990
    },
    {
      "epoch": 3.6022292276514296,
      "eval_nq_5round_bleu_score": 14.602900065691788,
      "eval_nq_5round_bleu_score_sem": 0.6578845170882258,
      "eval_nq_5round_emb_cos_sim": 0.9944454431533813,
      "eval_nq_5round_emb_cos_sim_sem": 0.00022207220639126058,
      "eval_nq_5round_emb_top1_equal": 0.6466666460037231,
      "eval_nq_5round_emb_top1_equal_sem": 0.027643749338232177,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 11.26,
      "eval_nq_5round_n_ngrams_match_2": 4.61,
      "eval_nq_5round_n_ngrams_match_3": 2.388,
      "eval_nq_5round_num_pred_words": 23.842,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.46409625666175036,
      "eval_nq_5round_token_set_f1": 0.4837141064844089,
      "eval_nq_5round_token_set_f1_sem": 0.0071125875675422095,
      "eval_nq_5round_token_set_precision": 0.4639777027763633,
      "eval_nq_5round_token_set_recall": 0.5126488942736313,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 12.067567631917889,
      "eval_nq_bleu_score_sem": 0.5296336024255047,
      "eval_nq_emb_cos_sim": 0.9938352704048157,
      "eval_nq_emb_cos_sim_sem": 0.00033329527436835735,
      "eval_nq_emb_top1_equal": 0.6666666865348816,
      "eval_nq_emb_top1_equal_sem": 0.027262027544015993,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.9894274473190308,
      "eval_nq_n_ngrams_match_1": 10.43,
      "eval_nq_n_ngrams_match_2": 3.95,
      "eval_nq_n_ngrams_match_3": 1.87,
      "eval_nq_num_pred_words": 23.852,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4240822279517773,
      "eval_nq_runtime": 19.0441,
      "eval_nq_samples_per_second": 26.255,
      "eval_nq_steps_per_second": 0.105,
      "eval_nq_token_set_f1": 0.4473951072908268,
      "eval_nq_token_set_f1_sem": 0.006273596216807243,
      "eval_nq_token_set_precision": 0.4289306694034396,
      "eval_nq_token_set_recall": 0.47661755190736893,
      "eval_nq_true_num_tokens": 32.0,
      "step": 63990
    },
    {
      "epoch": 3.6043120918712,
      "grad_norm": 0.4319443106651306,
      "learning_rate": 0.001,
      "loss": 2.2555,
      "step": 64027
    },
    {
      "epoch": 3.606732717856339,
      "grad_norm": 0.39805087447166443,
      "learning_rate": 0.001,
      "loss": 2.256,
      "step": 64070
    },
    {
      "epoch": 3.609153343841477,
      "grad_norm": 0.40591850876808167,
      "learning_rate": 0.001,
      "loss": 2.2553,
      "step": 64113
    },
    {
      "epoch": 3.6115739698266154,
      "grad_norm": 0.3850516676902771,
      "learning_rate": 0.001,
      "loss": 2.2528,
      "step": 64156
    },
    {
      "epoch": 3.613994595811754,
      "grad_norm": 0.43691325187683105,
      "learning_rate": 0.001,
      "loss": 2.2496,
      "step": 64199
    },
    {
      "epoch": 3.616415221796893,
      "grad_norm": 0.4252815246582031,
      "learning_rate": 0.001,
      "loss": 2.2494,
      "step": 64242
    },
    {
      "epoch": 3.618835847782031,
      "grad_norm": 0.40993407368659973,
      "learning_rate": 0.001,
      "loss": 2.2546,
      "step": 64285
    },
    {
      "epoch": 3.6212564737671693,
      "grad_norm": 0.38057827949523926,
      "learning_rate": 0.001,
      "loss": 2.2542,
      "step": 64328
    },
    {
      "epoch": 3.623677099752308,
      "grad_norm": 0.40795156359672546,
      "learning_rate": 0.001,
      "loss": 2.2498,
      "step": 64371
    },
    {
      "epoch": 3.6260977257374467,
      "grad_norm": 0.40774789452552795,
      "learning_rate": 0.001,
      "loss": 2.256,
      "step": 64414
    },
    {
      "epoch": 3.628518351722585,
      "grad_norm": 0.41117823123931885,
      "learning_rate": 0.001,
      "loss": 2.2499,
      "step": 64457
    },
    {
      "epoch": 3.6309389777077232,
      "grad_norm": 0.36719903349876404,
      "learning_rate": 0.001,
      "loss": 2.2505,
      "step": 64500
    },
    {
      "epoch": 3.633359603692862,
      "grad_norm": 0.38249558210372925,
      "learning_rate": 0.001,
      "loss": 2.2487,
      "step": 64543
    },
    {
      "epoch": 3.6357802296780006,
      "grad_norm": 0.40703195333480835,
      "learning_rate": 0.001,
      "loss": 2.2512,
      "step": 64586
    },
    {
      "epoch": 3.638200855663139,
      "grad_norm": 0.3870728611946106,
      "learning_rate": 0.001,
      "loss": 2.2409,
      "step": 64629
    },
    {
      "epoch": 3.640621481648277,
      "grad_norm": 0.4289146065711975,
      "learning_rate": 0.001,
      "loss": 2.2638,
      "step": 64672
    },
    {
      "epoch": 3.643042107633416,
      "grad_norm": 0.38264304399490356,
      "learning_rate": 0.001,
      "loss": 2.2505,
      "step": 64715
    },
    {
      "epoch": 3.6454627336185546,
      "grad_norm": 0.41105905175209045,
      "learning_rate": 0.001,
      "loss": 2.252,
      "step": 64758
    },
    {
      "epoch": 3.647883359603693,
      "grad_norm": 0.3864805996417999,
      "learning_rate": 0.001,
      "loss": 2.2582,
      "step": 64801
    },
    {
      "epoch": 3.650303985588831,
      "grad_norm": 0.4083375334739685,
      "learning_rate": 0.001,
      "loss": 2.2536,
      "step": 64844
    },
    {
      "epoch": 3.6527246115739698,
      "grad_norm": 0.40528222918510437,
      "learning_rate": 0.001,
      "loss": 2.2465,
      "step": 64887
    },
    {
      "epoch": 3.6551452375591085,
      "grad_norm": 0.39493197202682495,
      "learning_rate": 0.001,
      "loss": 2.25,
      "step": 64930
    },
    {
      "epoch": 3.6575658635442467,
      "grad_norm": 0.34618285298347473,
      "learning_rate": 0.001,
      "loss": 2.2499,
      "step": 64973
    },
    {
      "epoch": 3.6599864895293854,
      "grad_norm": 0.4158504009246826,
      "learning_rate": 0.001,
      "loss": 2.2541,
      "step": 65016
    },
    {
      "epoch": 3.6624071155145237,
      "grad_norm": 0.4326108694076538,
      "learning_rate": 0.001,
      "loss": 2.255,
      "step": 65059
    },
    {
      "epoch": 3.6648277414996624,
      "grad_norm": 0.4262112081050873,
      "learning_rate": 0.001,
      "loss": 2.2544,
      "step": 65102
    },
    {
      "epoch": 3.6672483674848007,
      "grad_norm": 0.40814971923828125,
      "learning_rate": 0.001,
      "loss": 2.248,
      "step": 65145
    },
    {
      "epoch": 3.6696689934699394,
      "grad_norm": 0.3596804738044739,
      "learning_rate": 0.001,
      "loss": 2.2439,
      "step": 65188
    },
    {
      "epoch": 3.6720896194550776,
      "grad_norm": 0.36804327368736267,
      "learning_rate": 0.001,
      "loss": 2.2381,
      "step": 65231
    },
    {
      "epoch": 3.6745102454402163,
      "grad_norm": 0.3943122625350952,
      "learning_rate": 0.001,
      "loss": 2.2496,
      "step": 65274
    },
    {
      "epoch": 3.6769308714253546,
      "grad_norm": 0.5054014921188354,
      "learning_rate": 0.001,
      "loss": 2.2508,
      "step": 65317
    },
    {
      "epoch": 3.6793514974104933,
      "grad_norm": 0.42513057589530945,
      "learning_rate": 0.001,
      "loss": 2.2524,
      "step": 65360
    },
    {
      "epoch": 3.6817721233956315,
      "grad_norm": 0.4183689057826996,
      "learning_rate": 0.001,
      "loss": 2.2643,
      "step": 65403
    },
    {
      "epoch": 3.6841927493807702,
      "grad_norm": 0.3729482591152191,
      "learning_rate": 0.001,
      "loss": 2.2517,
      "step": 65446
    },
    {
      "epoch": 3.6866133753659085,
      "grad_norm": 0.38989531993865967,
      "learning_rate": 0.001,
      "loss": 2.2551,
      "step": 65489
    },
    {
      "epoch": 3.689034001351047,
      "grad_norm": 0.38615795969963074,
      "learning_rate": 0.001,
      "loss": 2.2639,
      "step": 65532
    },
    {
      "epoch": 3.6914546273361855,
      "grad_norm": 0.4919602870941162,
      "learning_rate": 0.001,
      "loss": 2.253,
      "step": 65575
    },
    {
      "epoch": 3.693875253321324,
      "grad_norm": 0.447257399559021,
      "learning_rate": 0.001,
      "loss": 2.2425,
      "step": 65618
    },
    {
      "epoch": 3.6962958793064624,
      "grad_norm": 0.38409489393234253,
      "learning_rate": 0.001,
      "loss": 2.2525,
      "step": 65661
    },
    {
      "epoch": 3.698716505291601,
      "grad_norm": 0.45774224400520325,
      "learning_rate": 0.001,
      "loss": 2.2454,
      "step": 65704
    },
    {
      "epoch": 3.7011371312767394,
      "grad_norm": 0.45903587341308594,
      "learning_rate": 0.001,
      "loss": 2.2466,
      "step": 65747
    },
    {
      "epoch": 3.703557757261878,
      "grad_norm": 0.414654940366745,
      "learning_rate": 0.001,
      "loss": 2.2536,
      "step": 65790
    },
    {
      "epoch": 3.7059783832470163,
      "grad_norm": 0.3891694247722626,
      "learning_rate": 0.001,
      "loss": 2.2429,
      "step": 65833
    },
    {
      "epoch": 3.708399009232155,
      "grad_norm": 0.37660324573516846,
      "learning_rate": 0.001,
      "loss": 2.2576,
      "step": 65876
    },
    {
      "epoch": 3.7108196352172933,
      "grad_norm": 0.369788259267807,
      "learning_rate": 0.001,
      "loss": 2.2406,
      "step": 65919
    },
    {
      "epoch": 3.713240261202432,
      "grad_norm": 0.4078064262866974,
      "learning_rate": 0.001,
      "loss": 2.2542,
      "step": 65962
    },
    {
      "epoch": 3.7156608871875703,
      "grad_norm": 0.457574725151062,
      "learning_rate": 0.001,
      "loss": 2.2543,
      "step": 66005
    },
    {
      "epoch": 3.718081513172709,
      "grad_norm": 0.437721312046051,
      "learning_rate": 0.001,
      "loss": 2.26,
      "step": 66048
    },
    {
      "epoch": 3.720502139157847,
      "grad_norm": 0.3985402286052704,
      "learning_rate": 0.001,
      "loss": 2.2531,
      "step": 66091
    },
    {
      "epoch": 3.722303535239811,
      "eval_ag_news_bleu_score": 4.601042464797805,
      "eval_ag_news_bleu_score_sem": 0.1792550453624814,
      "eval_ag_news_emb_cos_sim": 0.9922434687614441,
      "eval_ag_news_emb_cos_sim_sem": 0.0002693482244055769,
      "eval_ag_news_emb_top1_equal": 0.6366666555404663,
      "eval_ag_news_emb_top1_equal_sem": 0.02781461728343201,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.550832509994507,
      "eval_ag_news_n_ngrams_match_1": 5.344,
      "eval_ag_news_n_ngrams_match_2": 1.06,
      "eval_ag_news_n_ngrams_match_3": 0.296,
      "eval_ag_news_num_pred_words": 22.522,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.29741953665839227,
      "eval_ag_news_runtime": 15.9867,
      "eval_ag_news_samples_per_second": 31.276,
      "eval_ag_news_steps_per_second": 0.125,
      "eval_ag_news_token_set_f1": 0.24920261103946864,
      "eval_ag_news_token_set_f1_sem": 0.005104213017162403,
      "eval_ag_news_token_set_precision": 0.24145795072128798,
      "eval_ag_news_token_set_recall": 0.26748040816140856,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 66123
    },
    {
      "epoch": 3.722303535239811,
      "eval_anthropic_toxic_prompts_bleu_score": 6.405169673872124,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.26803415466180025,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9923840761184692,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002618701708462562,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6700000166893005,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027193103952244484,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8199164867401123,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.546,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.904,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.838,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.596,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3232802255344478,
      "eval_anthropic_toxic_prompts_runtime": 6.7862,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.678,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.295,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.35654070033510177,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006483572332884966,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4385241033973912,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.32351180732552753,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 66123
    },
    {
      "epoch": 3.722303535239811,
      "eval_arxiv_bleu_score": 4.211824136630103,
      "eval_arxiv_bleu_score_sem": 0.13963007331112923,
      "eval_arxiv_emb_cos_sim": 0.9915816187858582,
      "eval_arxiv_emb_cos_sim_sem": 0.00031829585556794685,
      "eval_arxiv_emb_top1_equal": 0.6066666841506958,
      "eval_arxiv_emb_top1_equal_sem": 0.028250091805185003,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5353646278381348,
      "eval_arxiv_n_ngrams_match_1": 5.3,
      "eval_arxiv_n_ngrams_match_2": 0.928,
      "eval_arxiv_n_ngrams_match_3": 0.202,
      "eval_arxiv_num_pred_words": 18.266,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2828765810909646,
      "eval_arxiv_runtime": 6.765,
      "eval_arxiv_samples_per_second": 73.91,
      "eval_arxiv_steps_per_second": 0.296,
      "eval_arxiv_token_set_f1": 0.26006744801579385,
      "eval_arxiv_token_set_f1_sem": 0.00506764075912393,
      "eval_arxiv_token_set_precision": 0.22756070083292165,
      "eval_arxiv_token_set_recall": 0.31820006975159415,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 66123
    },
    {
      "epoch": 3.722303535239811,
      "eval_python_code_alpaca_bleu_score": 6.7757870836731975,
      "eval_python_code_alpaca_bleu_score_sem": 0.24030725026282113,
      "eval_python_code_alpaca_emb_cos_sim": 0.9909369945526123,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0002892820025015599,
      "eval_python_code_alpaca_emb_top1_equal": 0.6333333253860474,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02786867456387452,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.815061092376709,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.57,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.782,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.654,
      "eval_python_code_alpaca_num_pred_words": 20.814,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3578641276412843,
      "eval_python_code_alpaca_runtime": 6.8794,
      "eval_python_code_alpaca_samples_per_second": 72.681,
      "eval_python_code_alpaca_steps_per_second": 0.291,
      "eval_python_code_alpaca_token_set_f1": 0.38621542554330957,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006017060535228305,
      "eval_python_code_alpaca_token_set_precision": 0.3945489091767777,
      "eval_python_code_alpaca_token_set_recall": 0.4080277988725056,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 66123
    },
    {
      "epoch": 3.722303535239811,
      "eval_wikibio_bleu_score": 4.097580934136046,
      "eval_wikibio_bleu_score_sem": 0.14628524154417674,
      "eval_wikibio_emb_cos_sim": 0.991817057132721,
      "eval_wikibio_emb_cos_sim_sem": 0.0002858851944663058,
      "eval_wikibio_emb_top1_equal": 0.6933333277702332,
      "eval_wikibio_emb_top1_equal_sem": 0.026666666188047156,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9272868633270264,
      "eval_wikibio_n_ngrams_match_1": 3.372,
      "eval_wikibio_n_ngrams_match_2": 0.94,
      "eval_wikibio_n_ngrams_match_3": 0.202,
      "eval_wikibio_num_pred_words": 19.898,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3050286347885166,
      "eval_wikibio_runtime": 6.6778,
      "eval_wikibio_samples_per_second": 74.875,
      "eval_wikibio_steps_per_second": 0.3,
      "eval_wikibio_token_set_f1": 0.21987314262439253,
      "eval_wikibio_token_set_f1_sem": 0.0055816264795409375,
      "eval_wikibio_token_set_precision": 0.25774943744309997,
      "eval_wikibio_token_set_recall": 0.19772400439243099,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 66123
    },
    {
      "epoch": 3.722303535239811,
      "eval_nq_5round_bleu_score": 14.609973242791778,
      "eval_nq_5round_bleu_score_sem": 0.6769758891963797,
      "eval_nq_5round_emb_cos_sim": 0.9942792057991028,
      "eval_nq_5round_emb_cos_sim_sem": 0.0002472441035819891,
      "eval_nq_5round_emb_top1_equal": 0.6200000047683716,
      "eval_nq_5round_emb_top1_equal_sem": 0.028070624111834433,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 11.302,
      "eval_nq_5round_n_ngrams_match_2": 4.66,
      "eval_nq_5round_n_ngrams_match_3": 2.416,
      "eval_nq_5round_num_pred_words": 23.808,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.46448080431807337,
      "eval_nq_5round_token_set_f1": 0.4857693682578421,
      "eval_nq_5round_token_set_f1_sem": 0.006924242802318776,
      "eval_nq_5round_token_set_precision": 0.46437851095491234,
      "eval_nq_5round_token_set_recall": 0.5176550679702236,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 12.699813765873973,
      "eval_nq_bleu_score_sem": 0.5898993129042084,
      "eval_nq_emb_cos_sim": 0.9939388036727905,
      "eval_nq_emb_cos_sim_sem": 0.0003241309150265406,
      "eval_nq_emb_top1_equal": 0.6299999952316284,
      "eval_nq_emb_top1_equal_sem": 0.027921293391044915,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.972907304763794,
      "eval_nq_n_ngrams_match_1": 10.678,
      "eval_nq_n_ngrams_match_2": 4.098,
      "eval_nq_n_ngrams_match_3": 1.988,
      "eval_nq_num_pred_words": 24.022,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4312552830873624,
      "eval_nq_runtime": 19.2613,
      "eval_nq_samples_per_second": 25.959,
      "eval_nq_steps_per_second": 0.104,
      "eval_nq_token_set_f1": 0.45460040666962265,
      "eval_nq_token_set_f1_sem": 0.006616065368422897,
      "eval_nq_token_set_precision": 0.4358559131744938,
      "eval_nq_token_set_recall": 0.4839590213266591,
      "eval_nq_true_num_tokens": 32.0,
      "step": 66123
    },
    {
      "epoch": 3.722922765142986,
      "grad_norm": 0.3772100508213043,
      "learning_rate": 0.001,
      "loss": 2.246,
      "step": 66134
    },
    {
      "epoch": 3.725343391128124,
      "grad_norm": 0.46081700921058655,
      "learning_rate": 0.001,
      "loss": 2.2478,
      "step": 66177
    },
    {
      "epoch": 3.727764017113263,
      "grad_norm": 0.4067345857620239,
      "learning_rate": 0.001,
      "loss": 2.2597,
      "step": 66220
    },
    {
      "epoch": 3.730184643098401,
      "grad_norm": 0.3949984014034271,
      "learning_rate": 0.001,
      "loss": 2.241,
      "step": 66263
    },
    {
      "epoch": 3.73260526908354,
      "grad_norm": 0.39068567752838135,
      "learning_rate": 0.001,
      "loss": 2.2471,
      "step": 66306
    },
    {
      "epoch": 3.735025895068678,
      "grad_norm": 0.4905417263507843,
      "learning_rate": 0.001,
      "loss": 2.2466,
      "step": 66349
    },
    {
      "epoch": 3.737446521053817,
      "grad_norm": 0.4255632758140564,
      "learning_rate": 0.001,
      "loss": 2.2521,
      "step": 66392
    },
    {
      "epoch": 3.739867147038955,
      "grad_norm": 0.37613245844841003,
      "learning_rate": 0.001,
      "loss": 2.2494,
      "step": 66435
    },
    {
      "epoch": 3.7422877730240938,
      "grad_norm": 0.40929317474365234,
      "learning_rate": 0.001,
      "loss": 2.2509,
      "step": 66478
    },
    {
      "epoch": 3.744708399009232,
      "grad_norm": 0.4932287037372589,
      "learning_rate": 0.001,
      "loss": 2.2421,
      "step": 66521
    },
    {
      "epoch": 3.7471290249943707,
      "grad_norm": 0.3778914511203766,
      "learning_rate": 0.001,
      "loss": 2.2563,
      "step": 66564
    },
    {
      "epoch": 3.749549650979509,
      "grad_norm": 0.4170365035533905,
      "learning_rate": 0.001,
      "loss": 2.2578,
      "step": 66607
    },
    {
      "epoch": 3.7519702769646477,
      "grad_norm": 0.4574374854564667,
      "learning_rate": 0.001,
      "loss": 2.2476,
      "step": 66650
    },
    {
      "epoch": 3.754390902949786,
      "grad_norm": 0.40007832646369934,
      "learning_rate": 0.001,
      "loss": 2.2568,
      "step": 66693
    },
    {
      "epoch": 3.7568115289349246,
      "grad_norm": 0.4511682391166687,
      "learning_rate": 0.001,
      "loss": 2.2524,
      "step": 66736
    },
    {
      "epoch": 3.759232154920063,
      "grad_norm": 0.4066614508628845,
      "learning_rate": 0.001,
      "loss": 2.2451,
      "step": 66779
    },
    {
      "epoch": 3.7616527809052016,
      "grad_norm": 0.4364764392375946,
      "learning_rate": 0.001,
      "loss": 2.2411,
      "step": 66822
    },
    {
      "epoch": 3.76407340689034,
      "grad_norm": 0.3626806139945984,
      "learning_rate": 0.001,
      "loss": 2.247,
      "step": 66865
    },
    {
      "epoch": 3.7664940328754786,
      "grad_norm": 0.34122198820114136,
      "learning_rate": 0.001,
      "loss": 2.2532,
      "step": 66908
    },
    {
      "epoch": 3.768914658860617,
      "grad_norm": 0.3830035626888275,
      "learning_rate": 0.001,
      "loss": 2.2553,
      "step": 66951
    },
    {
      "epoch": 3.7713352848457555,
      "grad_norm": 0.40592166781425476,
      "learning_rate": 0.001,
      "loss": 2.2399,
      "step": 66994
    },
    {
      "epoch": 3.773755910830894,
      "grad_norm": 0.42617878317832947,
      "learning_rate": 0.001,
      "loss": 2.2461,
      "step": 67037
    },
    {
      "epoch": 3.7761765368160325,
      "grad_norm": 0.38133493065834045,
      "learning_rate": 0.001,
      "loss": 2.2396,
      "step": 67080
    },
    {
      "epoch": 3.7785971628011708,
      "grad_norm": 0.44496282935142517,
      "learning_rate": 0.001,
      "loss": 2.2433,
      "step": 67123
    },
    {
      "epoch": 3.7810177887863095,
      "grad_norm": 0.45126819610595703,
      "learning_rate": 0.001,
      "loss": 2.2464,
      "step": 67166
    },
    {
      "epoch": 3.7834384147714477,
      "grad_norm": 0.43426409363746643,
      "learning_rate": 0.001,
      "loss": 2.2455,
      "step": 67209
    },
    {
      "epoch": 3.7858590407565864,
      "grad_norm": 0.38092532753944397,
      "learning_rate": 0.001,
      "loss": 2.2376,
      "step": 67252
    },
    {
      "epoch": 3.7882796667417247,
      "grad_norm": 0.3561946749687195,
      "learning_rate": 0.001,
      "loss": 2.2433,
      "step": 67295
    },
    {
      "epoch": 3.7907002927268634,
      "grad_norm": 0.38651469349861145,
      "learning_rate": 0.001,
      "loss": 2.2333,
      "step": 67338
    },
    {
      "epoch": 3.7931209187120016,
      "grad_norm": 0.3680992126464844,
      "learning_rate": 0.001,
      "loss": 2.2441,
      "step": 67381
    },
    {
      "epoch": 3.7955415446971403,
      "grad_norm": 0.38792797923088074,
      "learning_rate": 0.001,
      "loss": 2.244,
      "step": 67424
    },
    {
      "epoch": 3.797962170682279,
      "grad_norm": 0.3867528438568115,
      "learning_rate": 0.001,
      "loss": 2.2501,
      "step": 67467
    },
    {
      "epoch": 3.8003827966674173,
      "grad_norm": 0.43207862973213196,
      "learning_rate": 0.001,
      "loss": 2.2558,
      "step": 67510
    },
    {
      "epoch": 3.8028034226525556,
      "grad_norm": 0.389087051153183,
      "learning_rate": 0.001,
      "loss": 2.2392,
      "step": 67553
    },
    {
      "epoch": 3.8052240486376943,
      "grad_norm": 0.40484532713890076,
      "learning_rate": 0.001,
      "loss": 2.2345,
      "step": 67596
    },
    {
      "epoch": 3.807644674622833,
      "grad_norm": 0.3830552101135254,
      "learning_rate": 0.001,
      "loss": 2.2436,
      "step": 67639
    },
    {
      "epoch": 3.810065300607971,
      "grad_norm": 0.4312455356121063,
      "learning_rate": 0.001,
      "loss": 2.2411,
      "step": 67682
    },
    {
      "epoch": 3.8124859265931095,
      "grad_norm": 0.37447845935821533,
      "learning_rate": 0.001,
      "loss": 2.2398,
      "step": 67725
    },
    {
      "epoch": 3.814906552578248,
      "grad_norm": 0.3865653872489929,
      "learning_rate": 0.001,
      "loss": 2.2384,
      "step": 67768
    },
    {
      "epoch": 3.817327178563387,
      "grad_norm": 0.4321003258228302,
      "learning_rate": 0.001,
      "loss": 2.2474,
      "step": 67811
    },
    {
      "epoch": 3.819747804548525,
      "grad_norm": 0.4194558262825012,
      "learning_rate": 0.001,
      "loss": 2.2461,
      "step": 67854
    },
    {
      "epoch": 3.8221684305336634,
      "grad_norm": 0.35703593492507935,
      "learning_rate": 0.001,
      "loss": 2.2383,
      "step": 67897
    },
    {
      "epoch": 3.824589056518802,
      "grad_norm": 0.4862925410270691,
      "learning_rate": 0.001,
      "loss": 2.2439,
      "step": 67940
    },
    {
      "epoch": 3.827009682503941,
      "grad_norm": 0.3823133409023285,
      "learning_rate": 0.001,
      "loss": 2.2468,
      "step": 67983
    },
    {
      "epoch": 3.829430308489079,
      "grad_norm": 0.3958582580089569,
      "learning_rate": 0.001,
      "loss": 2.2329,
      "step": 68026
    },
    {
      "epoch": 3.8318509344742173,
      "grad_norm": 0.38222670555114746,
      "learning_rate": 0.001,
      "loss": 2.2353,
      "step": 68069
    },
    {
      "epoch": 3.834271560459356,
      "grad_norm": 0.37466466426849365,
      "learning_rate": 0.001,
      "loss": 2.2346,
      "step": 68112
    },
    {
      "epoch": 3.8366921864444947,
      "grad_norm": 0.3637560307979584,
      "learning_rate": 0.001,
      "loss": 2.2465,
      "step": 68155
    },
    {
      "epoch": 3.839112812429633,
      "grad_norm": 0.42377546429634094,
      "learning_rate": 0.001,
      "loss": 2.2455,
      "step": 68198
    },
    {
      "epoch": 3.8415334384147712,
      "grad_norm": 0.43142008781433105,
      "learning_rate": 0.001,
      "loss": 2.2357,
      "step": 68241
    },
    {
      "epoch": 3.842377842828192,
      "eval_ag_news_bleu_score": 4.615577936890429,
      "eval_ag_news_bleu_score_sem": 0.18759293889000944,
      "eval_ag_news_emb_cos_sim": 0.9923074841499329,
      "eval_ag_news_emb_cos_sim_sem": 0.0002596418641298903,
      "eval_ag_news_emb_top1_equal": 0.6433333158493042,
      "eval_ag_news_emb_top1_equal_sem": 0.027702163273800266,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.5243749618530273,
      "eval_ag_news_n_ngrams_match_1": 5.43,
      "eval_ag_news_n_ngrams_match_2": 1.122,
      "eval_ag_news_n_ngrams_match_3": 0.328,
      "eval_ag_news_num_pred_words": 22.674,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2978474118242943,
      "eval_ag_news_runtime": 15.167,
      "eval_ag_news_samples_per_second": 32.966,
      "eval_ag_news_steps_per_second": 0.132,
      "eval_ag_news_token_set_f1": 0.2511871219138249,
      "eval_ag_news_token_set_f1_sem": 0.005604325567912411,
      "eval_ag_news_token_set_precision": 0.24514191604378796,
      "eval_ag_news_token_set_recall": 0.26644481001173315,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 68256
    },
    {
      "epoch": 3.842377842828192,
      "eval_anthropic_toxic_prompts_bleu_score": 6.727957806855518,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.28558460276118425,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9920673966407776,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00026209697242751245,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6399999856948853,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02775911810844162,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.808535575866699,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.564,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 2.022,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.93,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.546,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.324916355584111,
      "eval_anthropic_toxic_prompts_runtime": 7.5153,
      "eval_anthropic_toxic_prompts_samples_per_second": 66.531,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.266,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3592399321227402,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006667122600970091,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.43424364043087904,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.33217969078169135,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 68256
    },
    {
      "epoch": 3.842377842828192,
      "eval_arxiv_bleu_score": 3.987866884037571,
      "eval_arxiv_bleu_score_sem": 0.12789215772481743,
      "eval_arxiv_emb_cos_sim": 0.9916191101074219,
      "eval_arxiv_emb_cos_sim_sem": 0.0002900187043661384,
      "eval_arxiv_emb_top1_equal": 0.6033333539962769,
      "eval_arxiv_emb_top1_equal_sem": 0.028291497235256893,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5443716049194336,
      "eval_arxiv_n_ngrams_match_1": 5.1,
      "eval_arxiv_n_ngrams_match_2": 0.856,
      "eval_arxiv_n_ngrams_match_3": 0.17,
      "eval_arxiv_num_pred_words": 18.036,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2800106745382506,
      "eval_arxiv_runtime": 6.8744,
      "eval_arxiv_samples_per_second": 72.734,
      "eval_arxiv_steps_per_second": 0.291,
      "eval_arxiv_token_set_f1": 0.24997580211423806,
      "eval_arxiv_token_set_f1_sem": 0.0050420726161552335,
      "eval_arxiv_token_set_precision": 0.21882607842950513,
      "eval_arxiv_token_set_recall": 0.3073163477363343,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 68256
    },
    {
      "epoch": 3.842377842828192,
      "eval_python_code_alpaca_bleu_score": 6.76477509164536,
      "eval_python_code_alpaca_bleu_score_sem": 0.2603838638366174,
      "eval_python_code_alpaca_emb_cos_sim": 0.990666925907135,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00034164628280479887,
      "eval_python_code_alpaca_emb_top1_equal": 0.6299999952316284,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027921293391044915,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.8340251445770264,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.474,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.732,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.67,
      "eval_python_code_alpaca_num_pred_words": 21.03,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3532521571335152,
      "eval_python_code_alpaca_runtime": 6.7963,
      "eval_python_code_alpaca_samples_per_second": 73.57,
      "eval_python_code_alpaca_steps_per_second": 0.294,
      "eval_python_code_alpaca_token_set_f1": 0.3808250774611362,
      "eval_python_code_alpaca_token_set_f1_sem": 0.0059741754592590405,
      "eval_python_code_alpaca_token_set_precision": 0.3877096834975299,
      "eval_python_code_alpaca_token_set_recall": 0.4013715087889681,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 68256
    },
    {
      "epoch": 3.842377842828192,
      "eval_wikibio_bleu_score": 4.266700142828701,
      "eval_wikibio_bleu_score_sem": 0.14194163983955455,
      "eval_wikibio_emb_cos_sim": 0.9917196035385132,
      "eval_wikibio_emb_cos_sim_sem": 0.0002717958049124509,
      "eval_wikibio_emb_top1_equal": 0.6833333373069763,
      "eval_wikibio_emb_top1_equal_sem": 0.02690183265038281,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.8978681564331055,
      "eval_wikibio_n_ngrams_match_1": 3.522,
      "eval_wikibio_n_ngrams_match_2": 1.024,
      "eval_wikibio_n_ngrams_match_3": 0.25,
      "eval_wikibio_num_pred_words": 20.436,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.31899979704338477,
      "eval_wikibio_runtime": 6.7273,
      "eval_wikibio_samples_per_second": 74.324,
      "eval_wikibio_steps_per_second": 0.297,
      "eval_wikibio_token_set_f1": 0.22703131950788802,
      "eval_wikibio_token_set_f1_sem": 0.005574441077597451,
      "eval_wikibio_token_set_precision": 0.26852556481851403,
      "eval_wikibio_token_set_recall": 0.20336156849910408,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 68256
    },
    {
      "epoch": 3.842377842828192,
      "eval_nq_5round_bleu_score": 14.582972530624378,
      "eval_nq_5round_bleu_score_sem": 0.6604115643721938,
      "eval_nq_5round_emb_cos_sim": 0.9946908950805664,
      "eval_nq_5round_emb_cos_sim_sem": 0.00030798143799457494,
      "eval_nq_5round_emb_top1_equal": 0.653333306312561,
      "eval_nq_5round_emb_top1_equal_sem": 0.027522495986455002,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.378,
      "eval_nq_5round_n_ngrams_match_2": 4.738,
      "eval_nq_5round_n_ngrams_match_3": 2.418,
      "eval_nq_5round_num_pred_words": 24.136,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4694950222148263,
      "eval_nq_5round_token_set_f1": 0.4851541119324355,
      "eval_nq_5round_token_set_f1_sem": 0.0070431708832202145,
      "eval_nq_5round_token_set_precision": 0.46716015194791305,
      "eval_nq_5round_token_set_recall": 0.51200434035121,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 12.733934543107255,
      "eval_nq_bleu_score_sem": 0.5451515093173931,
      "eval_nq_emb_cos_sim": 0.9939138889312744,
      "eval_nq_emb_cos_sim_sem": 0.0003195839412198635,
      "eval_nq_emb_top1_equal": 0.6600000262260437,
      "eval_nq_emb_top1_equal_sem": 0.0273952875069568,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.9651803970336914,
      "eval_nq_n_ngrams_match_1": 10.638,
      "eval_nq_n_ngrams_match_2": 4.188,
      "eval_nq_n_ngrams_match_3": 2.034,
      "eval_nq_num_pred_words": 24.038,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.43397047782106957,
      "eval_nq_runtime": 39.3695,
      "eval_nq_samples_per_second": 12.7,
      "eval_nq_steps_per_second": 0.051,
      "eval_nq_token_set_f1": 0.4571692649120633,
      "eval_nq_token_set_f1_sem": 0.006334125359251041,
      "eval_nq_token_set_precision": 0.4363877707944214,
      "eval_nq_token_set_recall": 0.48834004125827907,
      "eval_nq_true_num_tokens": 32.0,
      "step": 68256
    },
    {
      "epoch": 3.84395406439991,
      "grad_norm": 0.3874695301055908,
      "learning_rate": 0.001,
      "loss": 2.2357,
      "step": 68284
    },
    {
      "epoch": 3.8463746903850486,
      "grad_norm": 0.4291534125804901,
      "learning_rate": 0.001,
      "loss": 2.2462,
      "step": 68327
    },
    {
      "epoch": 3.848795316370187,
      "grad_norm": 0.3927708566188812,
      "learning_rate": 0.001,
      "loss": 2.2408,
      "step": 68370
    },
    {
      "epoch": 3.851215942355325,
      "grad_norm": 0.34048938751220703,
      "learning_rate": 0.001,
      "loss": 2.2346,
      "step": 68413
    },
    {
      "epoch": 3.853636568340464,
      "grad_norm": 0.38909339904785156,
      "learning_rate": 0.001,
      "loss": 2.244,
      "step": 68456
    },
    {
      "epoch": 3.8560571943256026,
      "grad_norm": 0.38770362734794617,
      "learning_rate": 0.001,
      "loss": 2.2429,
      "step": 68499
    },
    {
      "epoch": 3.858477820310741,
      "grad_norm": 0.41383472084999084,
      "learning_rate": 0.001,
      "loss": 2.2448,
      "step": 68542
    },
    {
      "epoch": 3.860898446295879,
      "grad_norm": 0.3831659257411957,
      "learning_rate": 0.001,
      "loss": 2.2411,
      "step": 68585
    },
    {
      "epoch": 3.863319072281018,
      "grad_norm": 0.45087695121765137,
      "learning_rate": 0.001,
      "loss": 2.2374,
      "step": 68628
    },
    {
      "epoch": 3.8657396982661565,
      "grad_norm": 0.37601619958877563,
      "learning_rate": 0.001,
      "loss": 2.242,
      "step": 68671
    },
    {
      "epoch": 3.8681603242512947,
      "grad_norm": 0.44605857133865356,
      "learning_rate": 0.001,
      "loss": 2.2466,
      "step": 68714
    },
    {
      "epoch": 3.870580950236433,
      "grad_norm": 0.41611409187316895,
      "learning_rate": 0.001,
      "loss": 2.2427,
      "step": 68757
    },
    {
      "epoch": 3.8730015762215717,
      "grad_norm": 0.4255372881889343,
      "learning_rate": 0.001,
      "loss": 2.2353,
      "step": 68800
    },
    {
      "epoch": 3.8754222022067104,
      "grad_norm": 0.4432355761528015,
      "learning_rate": 0.001,
      "loss": 2.2402,
      "step": 68843
    },
    {
      "epoch": 3.8778428281918487,
      "grad_norm": 0.42589330673217773,
      "learning_rate": 0.001,
      "loss": 2.227,
      "step": 68886
    },
    {
      "epoch": 3.880263454176987,
      "grad_norm": 0.4654923975467682,
      "learning_rate": 0.001,
      "loss": 2.2329,
      "step": 68929
    },
    {
      "epoch": 3.8826840801621256,
      "grad_norm": 0.38589924573898315,
      "learning_rate": 0.001,
      "loss": 2.2413,
      "step": 68972
    },
    {
      "epoch": 3.8851047061472643,
      "grad_norm": 0.4162467122077942,
      "learning_rate": 0.001,
      "loss": 2.2422,
      "step": 69015
    },
    {
      "epoch": 3.8875253321324026,
      "grad_norm": 0.4429599642753601,
      "learning_rate": 0.001,
      "loss": 2.2391,
      "step": 69058
    },
    {
      "epoch": 3.889945958117541,
      "grad_norm": 0.38080981373786926,
      "learning_rate": 0.001,
      "loss": 2.2306,
      "step": 69101
    },
    {
      "epoch": 3.8923665841026795,
      "grad_norm": 0.4183889925479889,
      "learning_rate": 0.001,
      "loss": 2.2383,
      "step": 69144
    },
    {
      "epoch": 3.8947872100878183,
      "grad_norm": 0.464759886264801,
      "learning_rate": 0.001,
      "loss": 2.2331,
      "step": 69187
    },
    {
      "epoch": 3.8972078360729565,
      "grad_norm": 0.4447712004184723,
      "learning_rate": 0.001,
      "loss": 2.2324,
      "step": 69230
    },
    {
      "epoch": 3.8996284620580948,
      "grad_norm": 0.38259056210517883,
      "learning_rate": 0.001,
      "loss": 2.229,
      "step": 69273
    },
    {
      "epoch": 3.9020490880432335,
      "grad_norm": 0.4382227957248688,
      "learning_rate": 0.001,
      "loss": 2.2319,
      "step": 69316
    },
    {
      "epoch": 3.904469714028372,
      "grad_norm": 0.38661229610443115,
      "learning_rate": 0.001,
      "loss": 2.2421,
      "step": 69359
    },
    {
      "epoch": 3.9068903400135104,
      "grad_norm": 0.41434431076049805,
      "learning_rate": 0.001,
      "loss": 2.2396,
      "step": 69402
    },
    {
      "epoch": 3.9093109659986487,
      "grad_norm": 0.4047320783138275,
      "learning_rate": 0.001,
      "loss": 2.2403,
      "step": 69445
    },
    {
      "epoch": 3.9117315919837874,
      "grad_norm": 0.4954921305179596,
      "learning_rate": 0.001,
      "loss": 2.2419,
      "step": 69488
    },
    {
      "epoch": 3.914152217968926,
      "grad_norm": 0.4574543535709381,
      "learning_rate": 0.001,
      "loss": 2.2383,
      "step": 69531
    },
    {
      "epoch": 3.9165728439540644,
      "grad_norm": 0.3961266577243805,
      "learning_rate": 0.001,
      "loss": 2.2322,
      "step": 69574
    },
    {
      "epoch": 3.9189934699392026,
      "grad_norm": 0.43269333243370056,
      "learning_rate": 0.001,
      "loss": 2.2364,
      "step": 69617
    },
    {
      "epoch": 3.9214140959243413,
      "grad_norm": 0.41284045577049255,
      "learning_rate": 0.001,
      "loss": 2.2323,
      "step": 69660
    },
    {
      "epoch": 3.92383472190948,
      "grad_norm": 0.47512444853782654,
      "learning_rate": 0.001,
      "loss": 2.2374,
      "step": 69703
    },
    {
      "epoch": 3.9262553478946183,
      "grad_norm": 0.3947232663631439,
      "learning_rate": 0.001,
      "loss": 2.2276,
      "step": 69746
    },
    {
      "epoch": 3.928675973879757,
      "grad_norm": 0.47019749879837036,
      "learning_rate": 0.001,
      "loss": 2.2413,
      "step": 69789
    },
    {
      "epoch": 3.9310965998648952,
      "grad_norm": 0.4600406587123871,
      "learning_rate": 0.001,
      "loss": 2.2354,
      "step": 69832
    },
    {
      "epoch": 3.933517225850034,
      "grad_norm": 0.4310091435909271,
      "learning_rate": 0.001,
      "loss": 2.2326,
      "step": 69875
    },
    {
      "epoch": 3.935937851835172,
      "grad_norm": 0.48462119698524475,
      "learning_rate": 0.001,
      "loss": 2.2353,
      "step": 69918
    },
    {
      "epoch": 3.938358477820311,
      "grad_norm": 0.4250897467136383,
      "learning_rate": 0.001,
      "loss": 2.2374,
      "step": 69961
    },
    {
      "epoch": 3.940779103805449,
      "grad_norm": 0.3932710886001587,
      "learning_rate": 0.001,
      "loss": 2.24,
      "step": 70004
    },
    {
      "epoch": 3.943199729790588,
      "grad_norm": 0.4196799099445343,
      "learning_rate": 0.001,
      "loss": 2.2313,
      "step": 70047
    },
    {
      "epoch": 3.945620355775726,
      "grad_norm": 0.3967357575893402,
      "learning_rate": 0.001,
      "loss": 2.2208,
      "step": 70090
    },
    {
      "epoch": 3.948040981760865,
      "grad_norm": 0.40873631834983826,
      "learning_rate": 0.001,
      "loss": 2.2332,
      "step": 70133
    },
    {
      "epoch": 3.950461607746003,
      "grad_norm": 0.3522254526615143,
      "learning_rate": 0.001,
      "loss": 2.2298,
      "step": 70176
    },
    {
      "epoch": 3.952882233731142,
      "grad_norm": 0.40759965777397156,
      "learning_rate": 0.001,
      "loss": 2.2397,
      "step": 70219
    },
    {
      "epoch": 3.95530285971628,
      "grad_norm": 0.42094868421554565,
      "learning_rate": 0.001,
      "loss": 2.2488,
      "step": 70262
    },
    {
      "epoch": 3.9577234857014187,
      "grad_norm": 0.43499842286109924,
      "learning_rate": 0.001,
      "loss": 2.2335,
      "step": 70305
    },
    {
      "epoch": 3.960144111686557,
      "grad_norm": 0.4416908621788025,
      "learning_rate": 0.001,
      "loss": 2.2362,
      "step": 70348
    },
    {
      "epoch": 3.962452150416573,
      "eval_ag_news_bleu_score": 4.576253965139951,
      "eval_ag_news_bleu_score_sem": 0.1791266395236134,
      "eval_ag_news_emb_cos_sim": 0.9924857020378113,
      "eval_ag_news_emb_cos_sim_sem": 0.0002506414224316154,
      "eval_ag_news_emb_top1_equal": 0.6399999856948853,
      "eval_ag_news_emb_top1_equal_sem": 0.027759119829079505,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.517878532409668,
      "eval_ag_news_n_ngrams_match_1": 5.348,
      "eval_ag_news_n_ngrams_match_2": 1.084,
      "eval_ag_news_n_ngrams_match_3": 0.29,
      "eval_ag_news_num_pred_words": 22.662,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.3006184363796428,
      "eval_ag_news_runtime": 19.3098,
      "eval_ag_news_samples_per_second": 25.894,
      "eval_ag_news_steps_per_second": 0.104,
      "eval_ag_news_token_set_f1": 0.24931272591210346,
      "eval_ag_news_token_set_f1_sem": 0.005311080321235085,
      "eval_ag_news_token_set_precision": 0.2433557263429853,
      "eval_ag_news_token_set_recall": 0.2635363545663245,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 70389
    },
    {
      "epoch": 3.962452150416573,
      "eval_anthropic_toxic_prompts_bleu_score": 6.200875529980909,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.25778183395483484,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9922908544540405,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002522689845613086,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6366666555404663,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027814619004069896,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8030874729156494,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.542,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.826,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.808,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.798,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.32363228743783445,
      "eval_anthropic_toxic_prompts_runtime": 6.8035,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.492,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.294,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3594380474228453,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006978381116461572,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.43376987296847686,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.33297082075990775,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 70389
    },
    {
      "epoch": 3.962452150416573,
      "eval_arxiv_bleu_score": 4.193391450533035,
      "eval_arxiv_bleu_score_sem": 0.12323017518999617,
      "eval_arxiv_emb_cos_sim": 0.9918512105941772,
      "eval_arxiv_emb_cos_sim_sem": 0.00024511814793483165,
      "eval_arxiv_emb_top1_equal": 0.6033333539962769,
      "eval_arxiv_emb_top1_equal_sem": 0.028291497235256893,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5138356685638428,
      "eval_arxiv_n_ngrams_match_1": 5.38,
      "eval_arxiv_n_ngrams_match_2": 0.926,
      "eval_arxiv_n_ngrams_match_3": 0.186,
      "eval_arxiv_num_pred_words": 18.948,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2892982121352721,
      "eval_arxiv_runtime": 6.7221,
      "eval_arxiv_samples_per_second": 74.381,
      "eval_arxiv_steps_per_second": 0.298,
      "eval_arxiv_token_set_f1": 0.26367295689085685,
      "eval_arxiv_token_set_f1_sem": 0.004920283294499103,
      "eval_arxiv_token_set_precision": 0.2342965414233344,
      "eval_arxiv_token_set_recall": 0.3149099211828988,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 70389
    },
    {
      "epoch": 3.962452150416573,
      "eval_python_code_alpaca_bleu_score": 6.816909061724511,
      "eval_python_code_alpaca_bleu_score_sem": 0.2472636973839729,
      "eval_python_code_alpaca_emb_cos_sim": 0.9903557300567627,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004797776673335296,
      "eval_python_code_alpaca_emb_top1_equal": 0.653333306312561,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027522499427730773,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.8140904903411865,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.724,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.806,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.69,
      "eval_python_code_alpaca_num_pred_words": 21.524,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35829968117919103,
      "eval_python_code_alpaca_runtime": 6.7519,
      "eval_python_code_alpaca_samples_per_second": 74.053,
      "eval_python_code_alpaca_steps_per_second": 0.296,
      "eval_python_code_alpaca_token_set_f1": 0.3905375279223901,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006036532515664744,
      "eval_python_code_alpaca_token_set_precision": 0.40036433907482444,
      "eval_python_code_alpaca_token_set_recall": 0.40786573967006096,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 70389
    },
    {
      "epoch": 3.962452150416573,
      "eval_wikibio_bleu_score": 4.25946105832277,
      "eval_wikibio_bleu_score_sem": 0.14633561237695522,
      "eval_wikibio_emb_cos_sim": 0.9917640686035156,
      "eval_wikibio_emb_cos_sim_sem": 0.00026880864311873323,
      "eval_wikibio_emb_top1_equal": 0.6499999761581421,
      "eval_wikibio_emb_top1_equal_sem": 0.027583864257272155,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.830842971801758,
      "eval_wikibio_n_ngrams_match_1": 3.438,
      "eval_wikibio_n_ngrams_match_2": 1.004,
      "eval_wikibio_n_ngrams_match_3": 0.25,
      "eval_wikibio_num_pred_words": 19.994,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.31099112730304534,
      "eval_wikibio_runtime": 6.6019,
      "eval_wikibio_samples_per_second": 75.736,
      "eval_wikibio_steps_per_second": 0.303,
      "eval_wikibio_token_set_f1": 0.22218364501268054,
      "eval_wikibio_token_set_f1_sem": 0.005745896074432162,
      "eval_wikibio_token_set_precision": 0.25934033952130947,
      "eval_wikibio_token_set_recall": 0.20101831220229827,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 70389
    },
    {
      "epoch": 3.962452150416573,
      "eval_nq_5round_bleu_score": 15.656212619425634,
      "eval_nq_5round_bleu_score_sem": 0.7181266912293499,
      "eval_nq_5round_emb_cos_sim": 0.9945778846740723,
      "eval_nq_5round_emb_cos_sim_sem": 0.00032635069920838086,
      "eval_nq_5round_emb_top1_equal": 0.7066666483879089,
      "eval_nq_5round_emb_top1_equal_sem": 0.026330093931941276,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.632,
      "eval_nq_5round_n_ngrams_match_2": 4.948,
      "eval_nq_5round_n_ngrams_match_3": 2.642,
      "eval_nq_5round_num_pred_words": 24.048,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4762724785026299,
      "eval_nq_5round_token_set_f1": 0.4980724554067563,
      "eval_nq_5round_token_set_f1_sem": 0.0073666794837598144,
      "eval_nq_5round_token_set_precision": 0.4779865468213734,
      "eval_nq_5round_token_set_recall": 0.5272954699811426,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 12.572881114764558,
      "eval_nq_bleu_score_sem": 0.543334179377087,
      "eval_nq_emb_cos_sim": 0.9938811659812927,
      "eval_nq_emb_cos_sim_sem": 0.00032604714104646,
      "eval_nq_emb_top1_equal": 0.6733333468437195,
      "eval_nq_emb_top1_equal_sem": 0.027122635227651973,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.9547473192214966,
      "eval_nq_n_ngrams_match_1": 10.674,
      "eval_nq_n_ngrams_match_2": 4.152,
      "eval_nq_n_ngrams_match_3": 1.958,
      "eval_nq_num_pred_words": 23.964,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4362051888376509,
      "eval_nq_runtime": 18.9559,
      "eval_nq_samples_per_second": 26.377,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.46004095519474697,
      "eval_nq_token_set_f1_sem": 0.006358126269220885,
      "eval_nq_token_set_precision": 0.43859541201293323,
      "eval_nq_token_set_recall": 0.492216101292679,
      "eval_nq_true_num_tokens": 32.0,
      "step": 70389
    },
    {
      "epoch": 3.9625647376716957,
      "grad_norm": 0.44031986594200134,
      "learning_rate": 0.001,
      "loss": 2.2335,
      "step": 70391
    },
    {
      "epoch": 3.964985363656834,
      "grad_norm": 0.46846669912338257,
      "learning_rate": 0.001,
      "loss": 2.2416,
      "step": 70434
    },
    {
      "epoch": 3.9674059896419727,
      "grad_norm": 0.368499219417572,
      "learning_rate": 0.001,
      "loss": 2.2239,
      "step": 70477
    },
    {
      "epoch": 3.969826615627111,
      "grad_norm": 0.43030786514282227,
      "learning_rate": 0.001,
      "loss": 2.2301,
      "step": 70520
    },
    {
      "epoch": 3.9722472416122496,
      "grad_norm": 0.36223894357681274,
      "learning_rate": 0.001,
      "loss": 2.2309,
      "step": 70563
    },
    {
      "epoch": 3.974667867597388,
      "grad_norm": 0.35419762134552,
      "learning_rate": 0.001,
      "loss": 2.2257,
      "step": 70606
    },
    {
      "epoch": 3.9770884935825266,
      "grad_norm": 0.3968651294708252,
      "learning_rate": 0.001,
      "loss": 2.2326,
      "step": 70649
    },
    {
      "epoch": 3.979509119567665,
      "grad_norm": 0.4082814157009125,
      "learning_rate": 0.001,
      "loss": 2.2308,
      "step": 70692
    },
    {
      "epoch": 3.9819297455528035,
      "grad_norm": 0.4301328659057617,
      "learning_rate": 0.001,
      "loss": 2.2421,
      "step": 70735
    },
    {
      "epoch": 3.984350371537942,
      "grad_norm": 0.45578697323799133,
      "learning_rate": 0.001,
      "loss": 2.2275,
      "step": 70778
    },
    {
      "epoch": 3.9867709975230805,
      "grad_norm": 0.39755237102508545,
      "learning_rate": 0.001,
      "loss": 2.2356,
      "step": 70821
    },
    {
      "epoch": 3.9891916235082188,
      "grad_norm": 0.4156053960323334,
      "learning_rate": 0.001,
      "loss": 2.2429,
      "step": 70864
    },
    {
      "epoch": 3.9916122494933575,
      "grad_norm": 0.41138726472854614,
      "learning_rate": 0.001,
      "loss": 2.2416,
      "step": 70907
    },
    {
      "epoch": 3.9940328754784957,
      "grad_norm": 0.3499782085418701,
      "learning_rate": 0.001,
      "loss": 2.2358,
      "step": 70950
    },
    {
      "epoch": 3.9964535014636344,
      "grad_norm": 0.3797292411327362,
      "learning_rate": 0.001,
      "loss": 2.2378,
      "step": 70993
    },
    {
      "epoch": 3.9988741274487727,
      "grad_norm": 0.452372670173645,
      "learning_rate": 0.001,
      "loss": 2.236,
      "step": 71036
    },
    {
      "epoch": 4.001294753433911,
      "grad_norm": 0.47517263889312744,
      "learning_rate": 0.001,
      "loss": 2.1992,
      "step": 71079
    },
    {
      "epoch": 4.00371537941905,
      "grad_norm": 0.4066734313964844,
      "learning_rate": 0.001,
      "loss": 2.1874,
      "step": 71122
    },
    {
      "epoch": 4.006136005404188,
      "grad_norm": 0.3964747190475464,
      "learning_rate": 0.001,
      "loss": 2.1668,
      "step": 71165
    },
    {
      "epoch": 4.008556631389327,
      "grad_norm": 0.4290848970413208,
      "learning_rate": 0.001,
      "loss": 2.1768,
      "step": 71208
    },
    {
      "epoch": 4.010977257374465,
      "grad_norm": 0.4303681254386902,
      "learning_rate": 0.001,
      "loss": 2.1789,
      "step": 71251
    },
    {
      "epoch": 4.013397883359604,
      "grad_norm": 0.40318870544433594,
      "learning_rate": 0.001,
      "loss": 2.1725,
      "step": 71294
    },
    {
      "epoch": 4.015818509344742,
      "grad_norm": 0.44062340259552,
      "learning_rate": 0.001,
      "loss": 2.172,
      "step": 71337
    },
    {
      "epoch": 4.018239135329881,
      "grad_norm": 0.425620973110199,
      "learning_rate": 0.001,
      "loss": 2.1622,
      "step": 71380
    },
    {
      "epoch": 4.020659761315019,
      "grad_norm": 0.4030260145664215,
      "learning_rate": 0.001,
      "loss": 2.1751,
      "step": 71423
    },
    {
      "epoch": 4.0230803873001575,
      "grad_norm": 0.47269511222839355,
      "learning_rate": 0.001,
      "loss": 2.1754,
      "step": 71466
    },
    {
      "epoch": 4.025501013285296,
      "grad_norm": 0.3643563985824585,
      "learning_rate": 0.001,
      "loss": 2.1713,
      "step": 71509
    },
    {
      "epoch": 4.027921639270435,
      "grad_norm": 0.4113730192184448,
      "learning_rate": 0.001,
      "loss": 2.1874,
      "step": 71552
    },
    {
      "epoch": 4.030342265255573,
      "grad_norm": 0.42347851395606995,
      "learning_rate": 0.001,
      "loss": 2.1852,
      "step": 71595
    },
    {
      "epoch": 4.032762891240711,
      "grad_norm": 0.42524150013923645,
      "learning_rate": 0.001,
      "loss": 2.1787,
      "step": 71638
    },
    {
      "epoch": 4.03518351722585,
      "grad_norm": 0.4400519132614136,
      "learning_rate": 0.001,
      "loss": 2.1686,
      "step": 71681
    },
    {
      "epoch": 4.037604143210989,
      "grad_norm": 0.38585224747657776,
      "learning_rate": 0.001,
      "loss": 2.1732,
      "step": 71724
    },
    {
      "epoch": 4.040024769196127,
      "grad_norm": 0.4121893346309662,
      "learning_rate": 0.001,
      "loss": 2.1827,
      "step": 71767
    },
    {
      "epoch": 4.042445395181265,
      "grad_norm": 0.4725170135498047,
      "learning_rate": 0.001,
      "loss": 2.1801,
      "step": 71810
    },
    {
      "epoch": 4.044866021166404,
      "grad_norm": 0.42665234208106995,
      "learning_rate": 0.001,
      "loss": 2.1617,
      "step": 71853
    },
    {
      "epoch": 4.047286647151543,
      "grad_norm": 0.4308958351612091,
      "learning_rate": 0.001,
      "loss": 2.183,
      "step": 71896
    },
    {
      "epoch": 4.049707273136681,
      "grad_norm": 0.3869113326072693,
      "learning_rate": 0.001,
      "loss": 2.1839,
      "step": 71939
    },
    {
      "epoch": 4.052127899121819,
      "grad_norm": 0.5202300548553467,
      "learning_rate": 0.001,
      "loss": 2.1814,
      "step": 71982
    },
    {
      "epoch": 4.0545485251069575,
      "grad_norm": 0.4508724808692932,
      "learning_rate": 0.001,
      "loss": 2.1855,
      "step": 72025
    },
    {
      "epoch": 4.056969151092097,
      "grad_norm": 0.3946329951286316,
      "learning_rate": 0.001,
      "loss": 2.1798,
      "step": 72068
    },
    {
      "epoch": 4.059389777077235,
      "grad_norm": 0.39734598994255066,
      "learning_rate": 0.001,
      "loss": 2.1761,
      "step": 72111
    },
    {
      "epoch": 4.061810403062373,
      "grad_norm": 0.5285067558288574,
      "learning_rate": 0.001,
      "loss": 2.1766,
      "step": 72154
    },
    {
      "epoch": 4.064231029047511,
      "grad_norm": 0.42348983883857727,
      "learning_rate": 0.001,
      "loss": 2.1811,
      "step": 72197
    },
    {
      "epoch": 4.066651655032651,
      "grad_norm": 0.41594165563583374,
      "learning_rate": 0.001,
      "loss": 2.1758,
      "step": 72240
    },
    {
      "epoch": 4.069072281017789,
      "grad_norm": 0.4239123463630676,
      "learning_rate": 0.001,
      "loss": 2.1751,
      "step": 72283
    },
    {
      "epoch": 4.071492907002927,
      "grad_norm": 0.4109530448913574,
      "learning_rate": 0.001,
      "loss": 2.183,
      "step": 72326
    },
    {
      "epoch": 4.073913532988065,
      "grad_norm": 0.4017280340194702,
      "learning_rate": 0.001,
      "loss": 2.1891,
      "step": 72369
    },
    {
      "epoch": 4.0763341589732045,
      "grad_norm": 0.4297613501548767,
      "learning_rate": 0.001,
      "loss": 2.1743,
      "step": 72412
    },
    {
      "epoch": 4.078754784958343,
      "grad_norm": 0.424167275428772,
      "learning_rate": 0.001,
      "loss": 2.1864,
      "step": 72455
    },
    {
      "epoch": 4.081175410943481,
      "grad_norm": 0.44717347621917725,
      "learning_rate": 0.001,
      "loss": 2.1758,
      "step": 72498
    },
    {
      "epoch": 4.082526458004954,
      "eval_ag_news_bleu_score": 4.647247347920624,
      "eval_ag_news_bleu_score_sem": 0.17372609717791015,
      "eval_ag_news_emb_cos_sim": 0.992713212966919,
      "eval_ag_news_emb_cos_sim_sem": 0.0002411073947941285,
      "eval_ag_news_emb_top1_equal": 0.6933333277702332,
      "eval_ag_news_emb_top1_equal_sem": 0.026666666188047156,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.5136289596557617,
      "eval_ag_news_n_ngrams_match_1": 5.45,
      "eval_ag_news_n_ngrams_match_2": 1.14,
      "eval_ag_news_n_ngrams_match_3": 0.312,
      "eval_ag_news_num_pred_words": 22.584,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.30574494611248587,
      "eval_ag_news_runtime": 17.0614,
      "eval_ag_news_samples_per_second": 29.306,
      "eval_ag_news_steps_per_second": 0.117,
      "eval_ag_news_token_set_f1": 0.25343823633052537,
      "eval_ag_news_token_set_f1_sem": 0.005389352996391564,
      "eval_ag_news_token_set_precision": 0.2460291261608415,
      "eval_ag_news_token_set_recall": 0.2674061885064578,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 72522
    },
    {
      "epoch": 4.082526458004954,
      "eval_anthropic_toxic_prompts_bleu_score": 6.781292732043332,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.28746934183948236,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9925810098648071,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00022355924079866515,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6100000143051147,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02820730814416699,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.801724672317505,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.662,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 2.016,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.9,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.328,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3311997502173668,
      "eval_anthropic_toxic_prompts_runtime": 6.9379,
      "eval_anthropic_toxic_prompts_samples_per_second": 72.068,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.288,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3600357319004318,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006557562059363866,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.44375599495739654,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3258402418804334,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 72522
    },
    {
      "epoch": 4.082526458004954,
      "eval_arxiv_bleu_score": 4.08904324137283,
      "eval_arxiv_bleu_score_sem": 0.1342508684592907,
      "eval_arxiv_emb_cos_sim": 0.9917317032814026,
      "eval_arxiv_emb_cos_sim_sem": 0.0002727622119347171,
      "eval_arxiv_emb_top1_equal": 0.6166666746139526,
      "eval_arxiv_emb_top1_equal_sem": 0.028117578599086417,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5064775943756104,
      "eval_arxiv_n_ngrams_match_1": 5.198,
      "eval_arxiv_n_ngrams_match_2": 0.894,
      "eval_arxiv_n_ngrams_match_3": 0.182,
      "eval_arxiv_num_pred_words": 18.108,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2823828616646764,
      "eval_arxiv_runtime": 7.4167,
      "eval_arxiv_samples_per_second": 67.415,
      "eval_arxiv_steps_per_second": 0.27,
      "eval_arxiv_token_set_f1": 0.2559904437820558,
      "eval_arxiv_token_set_f1_sem": 0.005203871885148068,
      "eval_arxiv_token_set_precision": 0.22422863246015498,
      "eval_arxiv_token_set_recall": 0.313130813847764,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 72522
    },
    {
      "epoch": 4.082526458004954,
      "eval_python_code_alpaca_bleu_score": 6.897884566434168,
      "eval_python_code_alpaca_bleu_score_sem": 0.27615830010813935,
      "eval_python_code_alpaca_emb_cos_sim": 0.9903109073638916,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003989746302920987,
      "eval_python_code_alpaca_emb_top1_equal": 0.6433333158493042,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027702163273800266,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.828421115875244,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.36,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.716,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.676,
      "eval_python_code_alpaca_num_pred_words": 20.76,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3503584254346439,
      "eval_python_code_alpaca_runtime": 6.6737,
      "eval_python_code_alpaca_samples_per_second": 74.921,
      "eval_python_code_alpaca_steps_per_second": 0.3,
      "eval_python_code_alpaca_token_set_f1": 0.37714784409734714,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005894864285970998,
      "eval_python_code_alpaca_token_set_precision": 0.3836773564382008,
      "eval_python_code_alpaca_token_set_recall": 0.39660805099650526,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 72522
    },
    {
      "epoch": 4.082526458004954,
      "eval_wikibio_bleu_score": 4.205047197078186,
      "eval_wikibio_bleu_score_sem": 0.14065101242967828,
      "eval_wikibio_emb_cos_sim": 0.9922627210617065,
      "eval_wikibio_emb_cos_sim_sem": 0.00023875350839710254,
      "eval_wikibio_emb_top1_equal": 0.6333333253860474,
      "eval_wikibio_emb_top1_equal_sem": 0.02786867456387452,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.9311001300811768,
      "eval_wikibio_n_ngrams_match_1": 3.392,
      "eval_wikibio_n_ngrams_match_2": 0.932,
      "eval_wikibio_n_ngrams_match_3": 0.198,
      "eval_wikibio_num_pred_words": 19.862,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3172238884408407,
      "eval_wikibio_runtime": 6.4994,
      "eval_wikibio_samples_per_second": 76.93,
      "eval_wikibio_steps_per_second": 0.308,
      "eval_wikibio_token_set_f1": 0.22386995741731827,
      "eval_wikibio_token_set_f1_sem": 0.005271239120517767,
      "eval_wikibio_token_set_precision": 0.2597946590992203,
      "eval_wikibio_token_set_recall": 0.20366787282130572,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 72522
    },
    {
      "epoch": 4.082526458004954,
      "eval_nq_5round_bleu_score": 15.23429645363025,
      "eval_nq_5round_bleu_score_sem": 0.6688628436030286,
      "eval_nq_5round_emb_cos_sim": 0.9946694374084473,
      "eval_nq_5round_emb_cos_sim_sem": 0.00023220752975724288,
      "eval_nq_5round_emb_top1_equal": 0.6566666960716248,
      "eval_nq_5round_emb_top1_equal_sem": 0.027459642805142834,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.462,
      "eval_nq_5round_n_ngrams_match_2": 4.886,
      "eval_nq_5round_n_ngrams_match_3": 2.52,
      "eval_nq_5round_num_pred_words": 23.854,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.47714750043092247,
      "eval_nq_5round_token_set_f1": 0.4915702825056438,
      "eval_nq_5round_token_set_f1_sem": 0.0072344334102137356,
      "eval_nq_5round_token_set_precision": 0.4723588566764508,
      "eval_nq_5round_token_set_recall": 0.5197480052538208,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 12.437979836769175,
      "eval_nq_bleu_score_sem": 0.5603274296846881,
      "eval_nq_emb_cos_sim": 0.9941980242729187,
      "eval_nq_emb_cos_sim_sem": 0.0002981642847700551,
      "eval_nq_emb_top1_equal": 0.6566666960716248,
      "eval_nq_emb_top1_equal_sem": 0.027459642805142834,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.9386626482009888,
      "eval_nq_n_ngrams_match_1": 10.632,
      "eval_nq_n_ngrams_match_2": 4.05,
      "eval_nq_n_ngrams_match_3": 1.954,
      "eval_nq_num_pred_words": 23.856,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.43498464803961867,
      "eval_nq_runtime": 18.6756,
      "eval_nq_samples_per_second": 26.773,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.4546540279691461,
      "eval_nq_token_set_f1_sem": 0.006548517753536076,
      "eval_nq_token_set_precision": 0.43507546076158343,
      "eval_nq_token_set_recall": 0.4838808619082172,
      "eval_nq_true_num_tokens": 32.0,
      "step": 72522
    },
    {
      "epoch": 4.083596036928619,
      "grad_norm": 0.39113959670066833,
      "learning_rate": 0.001,
      "loss": 2.1775,
      "step": 72541
    },
    {
      "epoch": 4.086016662913758,
      "grad_norm": 0.4426662027835846,
      "learning_rate": 0.001,
      "loss": 2.1783,
      "step": 72584
    },
    {
      "epoch": 4.088437288898897,
      "grad_norm": 0.44767001271247864,
      "learning_rate": 0.001,
      "loss": 2.1701,
      "step": 72627
    },
    {
      "epoch": 4.090857914884035,
      "grad_norm": 0.40469425916671753,
      "learning_rate": 0.001,
      "loss": 2.1782,
      "step": 72670
    },
    {
      "epoch": 4.093278540869173,
      "grad_norm": 0.4129607379436493,
      "learning_rate": 0.001,
      "loss": 2.1884,
      "step": 72713
    },
    {
      "epoch": 4.095699166854312,
      "grad_norm": 0.43493518233299255,
      "learning_rate": 0.001,
      "loss": 2.1777,
      "step": 72756
    },
    {
      "epoch": 4.098119792839451,
      "grad_norm": 0.44783562421798706,
      "learning_rate": 0.001,
      "loss": 2.185,
      "step": 72799
    },
    {
      "epoch": 4.100540418824589,
      "grad_norm": 0.40518832206726074,
      "learning_rate": 0.001,
      "loss": 2.1857,
      "step": 72842
    },
    {
      "epoch": 4.102961044809727,
      "grad_norm": 0.38911405205726624,
      "learning_rate": 0.001,
      "loss": 2.1862,
      "step": 72885
    },
    {
      "epoch": 4.105381670794866,
      "grad_norm": 0.4029066860675812,
      "learning_rate": 0.001,
      "loss": 2.1853,
      "step": 72928
    },
    {
      "epoch": 4.1078022967800045,
      "grad_norm": 0.43000924587249756,
      "learning_rate": 0.001,
      "loss": 2.1773,
      "step": 72971
    },
    {
      "epoch": 4.110222922765143,
      "grad_norm": 0.5197097063064575,
      "learning_rate": 0.001,
      "loss": 2.1848,
      "step": 73014
    },
    {
      "epoch": 4.112643548750281,
      "grad_norm": 0.4448365569114685,
      "learning_rate": 0.001,
      "loss": 2.1877,
      "step": 73057
    },
    {
      "epoch": 4.11506417473542,
      "grad_norm": 0.4318884313106537,
      "learning_rate": 0.001,
      "loss": 2.1761,
      "step": 73100
    },
    {
      "epoch": 4.1174848007205584,
      "grad_norm": 0.41710707545280457,
      "learning_rate": 0.001,
      "loss": 2.1758,
      "step": 73143
    },
    {
      "epoch": 4.119905426705697,
      "grad_norm": 0.3947029411792755,
      "learning_rate": 0.001,
      "loss": 2.1775,
      "step": 73186
    },
    {
      "epoch": 4.122326052690835,
      "grad_norm": 0.4171319305896759,
      "learning_rate": 0.001,
      "loss": 2.183,
      "step": 73229
    },
    {
      "epoch": 4.124746678675974,
      "grad_norm": 0.4211637079715729,
      "learning_rate": 0.001,
      "loss": 2.1761,
      "step": 73272
    },
    {
      "epoch": 4.127167304661112,
      "grad_norm": 0.4738789200782776,
      "learning_rate": 0.001,
      "loss": 2.1809,
      "step": 73315
    },
    {
      "epoch": 4.129587930646251,
      "grad_norm": 0.4061814844608307,
      "learning_rate": 0.001,
      "loss": 2.1854,
      "step": 73358
    },
    {
      "epoch": 4.132008556631389,
      "grad_norm": 0.4318654239177704,
      "learning_rate": 0.001,
      "loss": 2.1839,
      "step": 73401
    },
    {
      "epoch": 4.134429182616528,
      "grad_norm": 0.44027504324913025,
      "learning_rate": 0.001,
      "loss": 2.1757,
      "step": 73444
    },
    {
      "epoch": 4.136849808601666,
      "grad_norm": 0.3901550769805908,
      "learning_rate": 0.001,
      "loss": 2.1867,
      "step": 73487
    },
    {
      "epoch": 4.1392704345868045,
      "grad_norm": 0.4633273482322693,
      "learning_rate": 0.001,
      "loss": 2.1857,
      "step": 73530
    },
    {
      "epoch": 4.141691060571944,
      "grad_norm": 0.3734920024871826,
      "learning_rate": 0.001,
      "loss": 2.1739,
      "step": 73573
    },
    {
      "epoch": 4.144111686557082,
      "grad_norm": 0.3872428834438324,
      "learning_rate": 0.001,
      "loss": 2.1841,
      "step": 73616
    },
    {
      "epoch": 4.14653231254222,
      "grad_norm": 0.38338446617126465,
      "learning_rate": 0.001,
      "loss": 2.1822,
      "step": 73659
    },
    {
      "epoch": 4.1489529385273585,
      "grad_norm": 0.4594787657260895,
      "learning_rate": 0.001,
      "loss": 2.1663,
      "step": 73702
    },
    {
      "epoch": 4.151373564512497,
      "grad_norm": 0.47575464844703674,
      "learning_rate": 0.001,
      "loss": 2.1855,
      "step": 73745
    },
    {
      "epoch": 4.153794190497636,
      "grad_norm": 0.4698750972747803,
      "learning_rate": 0.001,
      "loss": 2.1831,
      "step": 73788
    },
    {
      "epoch": 4.156214816482774,
      "grad_norm": 0.44433438777923584,
      "learning_rate": 0.001,
      "loss": 2.1742,
      "step": 73831
    },
    {
      "epoch": 4.158635442467912,
      "grad_norm": 0.3767596185207367,
      "learning_rate": 0.001,
      "loss": 2.1696,
      "step": 73874
    },
    {
      "epoch": 4.1610560684530515,
      "grad_norm": 0.3475441038608551,
      "learning_rate": 0.001,
      "loss": 2.1889,
      "step": 73917
    },
    {
      "epoch": 4.16347669443819,
      "grad_norm": 0.4567931890487671,
      "learning_rate": 0.001,
      "loss": 2.1916,
      "step": 73960
    },
    {
      "epoch": 4.165897320423328,
      "grad_norm": 0.4352571368217468,
      "learning_rate": 0.001,
      "loss": 2.1812,
      "step": 74003
    },
    {
      "epoch": 4.168317946408466,
      "grad_norm": 0.42780518531799316,
      "learning_rate": 0.001,
      "loss": 2.1883,
      "step": 74046
    },
    {
      "epoch": 4.1707385723936055,
      "grad_norm": 0.41725656390190125,
      "learning_rate": 0.001,
      "loss": 2.1869,
      "step": 74089
    },
    {
      "epoch": 4.173159198378744,
      "grad_norm": 0.40519753098487854,
      "learning_rate": 0.001,
      "loss": 2.1851,
      "step": 74132
    },
    {
      "epoch": 4.175579824363882,
      "grad_norm": 0.4546051323413849,
      "learning_rate": 0.001,
      "loss": 2.1868,
      "step": 74175
    },
    {
      "epoch": 4.17800045034902,
      "grad_norm": 0.4248879551887512,
      "learning_rate": 0.001,
      "loss": 2.18,
      "step": 74218
    },
    {
      "epoch": 4.180421076334159,
      "grad_norm": 0.47496265172958374,
      "learning_rate": 0.001,
      "loss": 2.1745,
      "step": 74261
    },
    {
      "epoch": 4.182841702319298,
      "grad_norm": 0.39850249886512756,
      "learning_rate": 0.001,
      "loss": 2.1907,
      "step": 74304
    },
    {
      "epoch": 4.185262328304436,
      "grad_norm": 0.44187888503074646,
      "learning_rate": 0.001,
      "loss": 2.1739,
      "step": 74347
    },
    {
      "epoch": 4.187682954289574,
      "grad_norm": 0.42159518599510193,
      "learning_rate": 0.001,
      "loss": 2.1784,
      "step": 74390
    },
    {
      "epoch": 4.190103580274713,
      "grad_norm": 0.40323901176452637,
      "learning_rate": 0.001,
      "loss": 2.1747,
      "step": 74433
    },
    {
      "epoch": 4.192524206259852,
      "grad_norm": 0.450204074382782,
      "learning_rate": 0.001,
      "loss": 2.1886,
      "step": 74476
    },
    {
      "epoch": 4.19494483224499,
      "grad_norm": 0.38516291975975037,
      "learning_rate": 0.001,
      "loss": 2.1832,
      "step": 74519
    },
    {
      "epoch": 4.197365458230128,
      "grad_norm": 0.389541894197464,
      "learning_rate": 0.001,
      "loss": 2.1784,
      "step": 74562
    },
    {
      "epoch": 4.199786084215267,
      "grad_norm": 0.4036976397037506,
      "learning_rate": 0.001,
      "loss": 2.18,
      "step": 74605
    },
    {
      "epoch": 4.2022067102004055,
      "grad_norm": 0.3888542950153351,
      "learning_rate": 0.001,
      "loss": 2.1808,
      "step": 74648
    },
    {
      "epoch": 4.202600765593335,
      "eval_ag_news_bleu_score": 4.58819824455895,
      "eval_ag_news_bleu_score_sem": 0.17395986026038643,
      "eval_ag_news_emb_cos_sim": 0.9927927851676941,
      "eval_ag_news_emb_cos_sim_sem": 0.00025220029347073134,
      "eval_ag_news_emb_top1_equal": 0.6433333158493042,
      "eval_ag_news_emb_top1_equal_sem": 0.02770216499443815,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.4968700408935547,
      "eval_ag_news_n_ngrams_match_1": 5.364,
      "eval_ag_news_n_ngrams_match_2": 1.098,
      "eval_ag_news_n_ngrams_match_3": 0.286,
      "eval_ag_news_num_pred_words": 22.786,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.2977633353251623,
      "eval_ag_news_runtime": 15.9852,
      "eval_ag_news_samples_per_second": 31.279,
      "eval_ag_news_steps_per_second": 0.125,
      "eval_ag_news_token_set_f1": 0.24837148187248292,
      "eval_ag_news_token_set_f1_sem": 0.005349703011887137,
      "eval_ag_news_token_set_precision": 0.2431822515362826,
      "eval_ag_news_token_set_recall": 0.2631669271888629,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 74655
    },
    {
      "epoch": 4.202600765593335,
      "eval_anthropic_toxic_prompts_bleu_score": 6.4731029817010866,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2667799984318804,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9923665523529053,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00024557634842669397,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6433333158493042,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027702163273800266,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.791635274887085,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.514,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.888,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.84,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.32,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.327731666060577,
      "eval_anthropic_toxic_prompts_runtime": 6.7755,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.795,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.295,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.35762213171878665,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0066428870047521935,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4408565187936976,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.32488059493131005,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 74655
    },
    {
      "epoch": 4.202600765593335,
      "eval_arxiv_bleu_score": 4.188797877823886,
      "eval_arxiv_bleu_score_sem": 0.12949040323411268,
      "eval_arxiv_emb_cos_sim": 0.9916254878044128,
      "eval_arxiv_emb_cos_sim_sem": 0.00026270744937222394,
      "eval_arxiv_emb_top1_equal": 0.5866666436195374,
      "eval_arxiv_emb_top1_equal_sem": 0.028478053956694904,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5428216457366943,
      "eval_arxiv_n_ngrams_match_1": 5.348,
      "eval_arxiv_n_ngrams_match_2": 0.936,
      "eval_arxiv_n_ngrams_match_3": 0.2,
      "eval_arxiv_num_pred_words": 18.48,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2898348851326209,
      "eval_arxiv_runtime": 6.621,
      "eval_arxiv_samples_per_second": 75.517,
      "eval_arxiv_steps_per_second": 0.302,
      "eval_arxiv_token_set_f1": 0.2621218419223152,
      "eval_arxiv_token_set_f1_sem": 0.005004823673912536,
      "eval_arxiv_token_set_precision": 0.23237597408257737,
      "eval_arxiv_token_set_recall": 0.3152321817611955,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 74655
    },
    {
      "epoch": 4.202600765593335,
      "eval_python_code_alpaca_bleu_score": 7.047603370812633,
      "eval_python_code_alpaca_bleu_score_sem": 0.26157382354967623,
      "eval_python_code_alpaca_emb_cos_sim": 0.9903501868247986,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00039194068950195436,
      "eval_python_code_alpaca_emb_top1_equal": 0.6600000262260437,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.0273952875069568,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.831798791885376,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.394,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.768,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.664,
      "eval_python_code_alpaca_num_pred_words": 20.162,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35961572493480887,
      "eval_python_code_alpaca_runtime": 6.6059,
      "eval_python_code_alpaca_samples_per_second": 75.69,
      "eval_python_code_alpaca_steps_per_second": 0.303,
      "eval_python_code_alpaca_token_set_f1": 0.379464329674013,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005853881645266885,
      "eval_python_code_alpaca_token_set_precision": 0.3835399652768176,
      "eval_python_code_alpaca_token_set_recall": 0.40236280017039605,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 74655
    },
    {
      "epoch": 4.202600765593335,
      "eval_wikibio_bleu_score": 4.444849977091168,
      "eval_wikibio_bleu_score_sem": 0.1515431350529477,
      "eval_wikibio_emb_cos_sim": 0.9921970963478088,
      "eval_wikibio_emb_cos_sim_sem": 0.00026177569707236636,
      "eval_wikibio_emb_top1_equal": 0.6833333373069763,
      "eval_wikibio_emb_top1_equal_sem": 0.02690183265038281,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.904026746749878,
      "eval_wikibio_n_ngrams_match_1": 3.438,
      "eval_wikibio_n_ngrams_match_2": 1.018,
      "eval_wikibio_n_ngrams_match_3": 0.246,
      "eval_wikibio_num_pred_words": 19.672,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.32133745034427885,
      "eval_wikibio_runtime": 6.4571,
      "eval_wikibio_samples_per_second": 77.434,
      "eval_wikibio_steps_per_second": 0.31,
      "eval_wikibio_token_set_f1": 0.2280739949424819,
      "eval_wikibio_token_set_f1_sem": 0.005705810191275505,
      "eval_wikibio_token_set_precision": 0.2622342943977397,
      "eval_wikibio_token_set_recall": 0.2091210162004135,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 74655
    },
    {
      "epoch": 4.202600765593335,
      "eval_nq_5round_bleu_score": 15.644334083342912,
      "eval_nq_5round_bleu_score_sem": 0.6928685239027337,
      "eval_nq_5round_emb_cos_sim": 0.9948338866233826,
      "eval_nq_5round_emb_cos_sim_sem": 0.00021018068977117534,
      "eval_nq_5round_emb_top1_equal": 0.6466666460037231,
      "eval_nq_5round_emb_top1_equal_sem": 0.027643749338232177,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.58,
      "eval_nq_5round_n_ngrams_match_2": 4.98,
      "eval_nq_5round_n_ngrams_match_3": 2.602,
      "eval_nq_5round_num_pred_words": 23.802,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.47932377836105233,
      "eval_nq_5round_token_set_f1": 0.4954173615182504,
      "eval_nq_5round_token_set_f1_sem": 0.007088911800296254,
      "eval_nq_5round_token_set_precision": 0.47573615985883466,
      "eval_nq_5round_token_set_recall": 0.5243123235566521,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.244424257759766,
      "eval_nq_bleu_score_sem": 0.6050857965044132,
      "eval_nq_emb_cos_sim": 0.9942433834075928,
      "eval_nq_emb_cos_sim_sem": 0.00021139988269526256,
      "eval_nq_emb_top1_equal": 0.7366666793823242,
      "eval_nq_emb_top1_equal_sem": 0.025471400992092292,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.9295889139175415,
      "eval_nq_n_ngrams_match_1": 10.852,
      "eval_nq_n_ngrams_match_2": 4.264,
      "eval_nq_n_ngrams_match_3": 2.132,
      "eval_nq_num_pred_words": 23.94,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4412943130370709,
      "eval_nq_runtime": 18.7654,
      "eval_nq_samples_per_second": 26.645,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.46450383816666463,
      "eval_nq_token_set_f1_sem": 0.006713509190920723,
      "eval_nq_token_set_precision": 0.4452815182145735,
      "eval_nq_token_set_recall": 0.4930726515404147,
      "eval_nq_true_num_tokens": 32.0,
      "step": 74655
    },
    {
      "epoch": 4.204627336185544,
      "grad_norm": 0.4230063855648041,
      "learning_rate": 0.001,
      "loss": 2.1773,
      "step": 74691
    },
    {
      "epoch": 4.207047962170682,
      "grad_norm": 0.44431838393211365,
      "learning_rate": 0.001,
      "loss": 2.1751,
      "step": 74734
    },
    {
      "epoch": 4.209468588155821,
      "grad_norm": 0.3925401568412781,
      "learning_rate": 0.001,
      "loss": 2.1864,
      "step": 74777
    },
    {
      "epoch": 4.211889214140959,
      "grad_norm": 0.44740474224090576,
      "learning_rate": 0.001,
      "loss": 2.1904,
      "step": 74820
    },
    {
      "epoch": 4.214309840126098,
      "grad_norm": 0.41982796788215637,
      "learning_rate": 0.001,
      "loss": 2.18,
      "step": 74863
    },
    {
      "epoch": 4.216730466111236,
      "grad_norm": 0.39618435502052307,
      "learning_rate": 0.001,
      "loss": 2.1774,
      "step": 74906
    },
    {
      "epoch": 4.219151092096375,
      "grad_norm": 0.5284149050712585,
      "learning_rate": 0.001,
      "loss": 2.1698,
      "step": 74949
    },
    {
      "epoch": 4.221571718081513,
      "grad_norm": 0.3851787745952606,
      "learning_rate": 0.001,
      "loss": 2.1699,
      "step": 74992
    },
    {
      "epoch": 4.223992344066652,
      "grad_norm": 0.3993995785713196,
      "learning_rate": 0.001,
      "loss": 2.1757,
      "step": 75035
    },
    {
      "epoch": 4.22641297005179,
      "grad_norm": 0.4293125569820404,
      "learning_rate": 0.001,
      "loss": 2.1821,
      "step": 75078
    },
    {
      "epoch": 4.228833596036929,
      "grad_norm": 0.3981727361679077,
      "learning_rate": 0.001,
      "loss": 2.1826,
      "step": 75121
    },
    {
      "epoch": 4.231254222022067,
      "grad_norm": 0.42949017882347107,
      "learning_rate": 0.001,
      "loss": 2.162,
      "step": 75164
    },
    {
      "epoch": 4.2336748480072055,
      "grad_norm": 0.45193880796432495,
      "learning_rate": 0.001,
      "loss": 2.1795,
      "step": 75207
    },
    {
      "epoch": 4.236095473992344,
      "grad_norm": 0.4484451413154602,
      "learning_rate": 0.001,
      "loss": 2.1759,
      "step": 75250
    },
    {
      "epoch": 4.238516099977483,
      "grad_norm": 0.42133063077926636,
      "learning_rate": 0.001,
      "loss": 2.1755,
      "step": 75293
    },
    {
      "epoch": 4.240936725962621,
      "grad_norm": 0.4464724063873291,
      "learning_rate": 0.001,
      "loss": 2.1858,
      "step": 75336
    },
    {
      "epoch": 4.243357351947759,
      "grad_norm": 0.3699142634868622,
      "learning_rate": 0.001,
      "loss": 2.1954,
      "step": 75379
    },
    {
      "epoch": 4.245777977932898,
      "grad_norm": 0.43838825821876526,
      "learning_rate": 0.001,
      "loss": 2.1761,
      "step": 75422
    },
    {
      "epoch": 4.248198603918037,
      "grad_norm": 0.39645111560821533,
      "learning_rate": 0.001,
      "loss": 2.1866,
      "step": 75465
    },
    {
      "epoch": 4.250619229903175,
      "grad_norm": 0.44165900349617004,
      "learning_rate": 0.001,
      "loss": 2.178,
      "step": 75508
    },
    {
      "epoch": 4.253039855888313,
      "grad_norm": 0.4866761863231659,
      "learning_rate": 0.001,
      "loss": 2.1852,
      "step": 75551
    },
    {
      "epoch": 4.255460481873452,
      "grad_norm": 0.3914507031440735,
      "learning_rate": 0.001,
      "loss": 2.1763,
      "step": 75594
    },
    {
      "epoch": 4.257881107858591,
      "grad_norm": 0.4168241322040558,
      "learning_rate": 0.001,
      "loss": 2.1779,
      "step": 75637
    },
    {
      "epoch": 4.260301733843729,
      "grad_norm": 0.43648380041122437,
      "learning_rate": 0.001,
      "loss": 2.186,
      "step": 75680
    },
    {
      "epoch": 4.262722359828867,
      "grad_norm": 0.4170612096786499,
      "learning_rate": 0.001,
      "loss": 2.1927,
      "step": 75723
    },
    {
      "epoch": 4.2651429858140055,
      "grad_norm": 0.4293377101421356,
      "learning_rate": 0.001,
      "loss": 2.1899,
      "step": 75766
    },
    {
      "epoch": 4.267563611799145,
      "grad_norm": 0.40237805247306824,
      "learning_rate": 0.001,
      "loss": 2.1664,
      "step": 75809
    },
    {
      "epoch": 4.269984237784283,
      "grad_norm": 0.43678808212280273,
      "learning_rate": 0.001,
      "loss": 2.1901,
      "step": 75852
    },
    {
      "epoch": 4.272404863769421,
      "grad_norm": 0.4242008626461029,
      "learning_rate": 0.001,
      "loss": 2.1763,
      "step": 75895
    },
    {
      "epoch": 4.2748254897545594,
      "grad_norm": 0.3467586040496826,
      "learning_rate": 0.001,
      "loss": 2.1806,
      "step": 75938
    },
    {
      "epoch": 4.277246115739699,
      "grad_norm": 0.44172635674476624,
      "learning_rate": 0.001,
      "loss": 2.1748,
      "step": 75981
    },
    {
      "epoch": 4.279666741724837,
      "grad_norm": 0.43868333101272583,
      "learning_rate": 0.001,
      "loss": 2.1729,
      "step": 76024
    },
    {
      "epoch": 4.282087367709975,
      "grad_norm": 0.48360127210617065,
      "learning_rate": 0.001,
      "loss": 2.1764,
      "step": 76067
    },
    {
      "epoch": 4.284507993695113,
      "grad_norm": 0.46181339025497437,
      "learning_rate": 0.001,
      "loss": 2.1829,
      "step": 76110
    },
    {
      "epoch": 4.2869286196802525,
      "grad_norm": 0.45860251784324646,
      "learning_rate": 0.001,
      "loss": 2.1953,
      "step": 76153
    },
    {
      "epoch": 4.289349245665391,
      "grad_norm": 0.45454955101013184,
      "learning_rate": 0.001,
      "loss": 2.1786,
      "step": 76196
    },
    {
      "epoch": 4.291769871650529,
      "grad_norm": 0.45579618215560913,
      "learning_rate": 0.001,
      "loss": 2.1771,
      "step": 76239
    },
    {
      "epoch": 4.294190497635667,
      "grad_norm": 0.41455990076065063,
      "learning_rate": 0.001,
      "loss": 2.1785,
      "step": 76282
    },
    {
      "epoch": 4.296611123620806,
      "grad_norm": 0.39138489961624146,
      "learning_rate": 0.001,
      "loss": 2.1811,
      "step": 76325
    },
    {
      "epoch": 4.299031749605945,
      "grad_norm": 0.4097137153148651,
      "learning_rate": 0.001,
      "loss": 2.1848,
      "step": 76368
    },
    {
      "epoch": 4.301452375591083,
      "grad_norm": 0.45801806449890137,
      "learning_rate": 0.001,
      "loss": 2.1715,
      "step": 76411
    },
    {
      "epoch": 4.303873001576221,
      "grad_norm": 0.47055187821388245,
      "learning_rate": 0.001,
      "loss": 2.1808,
      "step": 76454
    },
    {
      "epoch": 4.30629362756136,
      "grad_norm": 0.4613226056098938,
      "learning_rate": 0.001,
      "loss": 2.1781,
      "step": 76497
    },
    {
      "epoch": 4.308714253546499,
      "grad_norm": 0.3969699442386627,
      "learning_rate": 0.001,
      "loss": 2.1861,
      "step": 76540
    },
    {
      "epoch": 4.311134879531637,
      "grad_norm": 0.43469348549842834,
      "learning_rate": 0.001,
      "loss": 2.1761,
      "step": 76583
    },
    {
      "epoch": 4.313555505516775,
      "grad_norm": 0.43797942996025085,
      "learning_rate": 0.001,
      "loss": 2.1782,
      "step": 76626
    },
    {
      "epoch": 4.315976131501914,
      "grad_norm": 0.43207988142967224,
      "learning_rate": 0.001,
      "loss": 2.1804,
      "step": 76669
    },
    {
      "epoch": 4.3183967574870525,
      "grad_norm": 0.4193178713321686,
      "learning_rate": 0.001,
      "loss": 2.1788,
      "step": 76712
    },
    {
      "epoch": 4.320817383472191,
      "grad_norm": 0.44262999296188354,
      "learning_rate": 0.001,
      "loss": 2.1872,
      "step": 76755
    },
    {
      "epoch": 4.322675073181716,
      "eval_ag_news_bleu_score": 4.606567029053476,
      "eval_ag_news_bleu_score_sem": 0.1793310594650978,
      "eval_ag_news_emb_cos_sim": 0.9920088052749634,
      "eval_ag_news_emb_cos_sim_sem": 0.00029907517433555336,
      "eval_ag_news_emb_top1_equal": 0.6666666865348816,
      "eval_ag_news_emb_top1_equal_sem": 0.027262025823378108,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.515170097351074,
      "eval_ag_news_n_ngrams_match_1": 5.398,
      "eval_ag_news_n_ngrams_match_2": 1.078,
      "eval_ag_news_n_ngrams_match_3": 0.294,
      "eval_ag_news_num_pred_words": 22.356,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.30417533314585643,
      "eval_ag_news_runtime": 15.9152,
      "eval_ag_news_samples_per_second": 31.417,
      "eval_ag_news_steps_per_second": 0.126,
      "eval_ag_news_token_set_f1": 0.25350307632437785,
      "eval_ag_news_token_set_f1_sem": 0.005135671181241633,
      "eval_ag_news_token_set_precision": 0.2454117421293947,
      "eval_ag_news_token_set_recall": 0.27514876285219003,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 76788
    },
    {
      "epoch": 4.322675073181716,
      "eval_anthropic_toxic_prompts_bleu_score": 6.6612755505011485,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.280644421274709,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.992136538028717,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00024320923462592476,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.653333306312561,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027522495986455002,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8285083770751953,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.588,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.974,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.89,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.326,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3275529793561688,
      "eval_anthropic_toxic_prompts_runtime": 6.8248,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.262,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.293,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.35965233083114867,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006868085583581941,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4406602477778815,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3259381065519936,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 76788
    },
    {
      "epoch": 4.322675073181716,
      "eval_arxiv_bleu_score": 4.1817186098547925,
      "eval_arxiv_bleu_score_sem": 0.14841141446351216,
      "eval_arxiv_emb_cos_sim": 0.9915327429771423,
      "eval_arxiv_emb_cos_sim_sem": 0.00030068819169812235,
      "eval_arxiv_emb_top1_equal": 0.5766666531562805,
      "eval_arxiv_emb_top1_equal_sem": 0.028573804013736142,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.511140823364258,
      "eval_arxiv_n_ngrams_match_1": 5.33,
      "eval_arxiv_n_ngrams_match_2": 0.908,
      "eval_arxiv_n_ngrams_match_3": 0.21,
      "eval_arxiv_num_pred_words": 18.658,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2876660552921547,
      "eval_arxiv_runtime": 6.7514,
      "eval_arxiv_samples_per_second": 74.059,
      "eval_arxiv_steps_per_second": 0.296,
      "eval_arxiv_token_set_f1": 0.26220304126260996,
      "eval_arxiv_token_set_f1_sem": 0.004953546206774743,
      "eval_arxiv_token_set_precision": 0.23268135491759828,
      "eval_arxiv_token_set_recall": 0.3159038302645744,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 76788
    },
    {
      "epoch": 4.322675073181716,
      "eval_python_code_alpaca_bleu_score": 7.042511502200503,
      "eval_python_code_alpaca_bleu_score_sem": 0.2753309442981591,
      "eval_python_code_alpaca_emb_cos_sim": 0.990702748298645,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003055215979427682,
      "eval_python_code_alpaca_emb_top1_equal": 0.6633333563804626,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027329419768069584,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.838787794113159,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.386,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.744,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.686,
      "eval_python_code_alpaca_num_pred_words": 20.196,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35641733313766744,
      "eval_python_code_alpaca_runtime": 6.6823,
      "eval_python_code_alpaca_samples_per_second": 74.824,
      "eval_python_code_alpaca_steps_per_second": 0.299,
      "eval_python_code_alpaca_token_set_f1": 0.3772278796476912,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006213129484284061,
      "eval_python_code_alpaca_token_set_precision": 0.3823589324316424,
      "eval_python_code_alpaca_token_set_recall": 0.3995609833859027,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 76788
    },
    {
      "epoch": 4.322675073181716,
      "eval_wikibio_bleu_score": 4.276713063750169,
      "eval_wikibio_bleu_score_sem": 0.14257884350710476,
      "eval_wikibio_emb_cos_sim": 0.991728663444519,
      "eval_wikibio_emb_cos_sim_sem": 0.000283357792492534,
      "eval_wikibio_emb_top1_equal": 0.6333333253860474,
      "eval_wikibio_emb_top1_equal_sem": 0.02786867456387452,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.8776588439941406,
      "eval_wikibio_n_ngrams_match_1": 3.44,
      "eval_wikibio_n_ngrams_match_2": 0.98,
      "eval_wikibio_n_ngrams_match_3": 0.216,
      "eval_wikibio_num_pred_words": 19.466,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.31991855494455756,
      "eval_wikibio_runtime": 6.6782,
      "eval_wikibio_samples_per_second": 74.87,
      "eval_wikibio_steps_per_second": 0.299,
      "eval_wikibio_token_set_f1": 0.22724951872909685,
      "eval_wikibio_token_set_f1_sem": 0.005667709715090538,
      "eval_wikibio_token_set_precision": 0.2631474261863829,
      "eval_wikibio_token_set_recall": 0.2064616562238559,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 76788
    },
    {
      "epoch": 4.322675073181716,
      "eval_nq_5round_bleu_score": 15.780762078397057,
      "eval_nq_5round_bleu_score_sem": 0.7290403818940213,
      "eval_nq_5round_emb_cos_sim": 0.9947339296340942,
      "eval_nq_5round_emb_cos_sim_sem": 0.0003091514717565798,
      "eval_nq_5round_emb_top1_equal": 0.6933333277702332,
      "eval_nq_5round_emb_top1_equal_sem": 0.026666666188047156,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 11.606,
      "eval_nq_5round_n_ngrams_match_2": 4.878,
      "eval_nq_5round_n_ngrams_match_3": 2.68,
      "eval_nq_5round_num_pred_words": 23.716,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4789179463519862,
      "eval_nq_5round_token_set_f1": 0.4963615808864314,
      "eval_nq_5round_token_set_f1_sem": 0.007475776971963147,
      "eval_nq_5round_token_set_precision": 0.4756899583394874,
      "eval_nq_5round_token_set_recall": 0.5267995202154553,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.313241423152425,
      "eval_nq_bleu_score_sem": 0.5998005910807084,
      "eval_nq_emb_cos_sim": 0.9941312670707703,
      "eval_nq_emb_cos_sim_sem": 0.0003008112710768553,
      "eval_nq_emb_top1_equal": 0.6466666460037231,
      "eval_nq_emb_top1_equal_sem": 0.027643749338232177,
      "eval_nq_exact_match": 0.002,
      "eval_nq_exact_match_sem": 0.002,
      "eval_nq_loss": 1.9178656339645386,
      "eval_nq_n_ngrams_match_1": 10.738,
      "eval_nq_n_ngrams_match_2": 4.264,
      "eval_nq_n_ngrams_match_3": 2.16,
      "eval_nq_num_pred_words": 23.856,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.44028545233184657,
      "eval_nq_runtime": 18.7738,
      "eval_nq_samples_per_second": 26.633,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.4622349898121296,
      "eval_nq_token_set_f1_sem": 0.006670147784211084,
      "eval_nq_token_set_precision": 0.44144354605668357,
      "eval_nq_token_set_recall": 0.49353175918023146,
      "eval_nq_true_num_tokens": 32.0,
      "step": 76788
    },
    {
      "epoch": 4.323238009457329,
      "grad_norm": 0.4994538128376007,
      "learning_rate": 0.001,
      "loss": 2.1838,
      "step": 76798
    },
    {
      "epoch": 4.325658635442468,
      "grad_norm": 0.4435728192329407,
      "learning_rate": 0.001,
      "loss": 2.1671,
      "step": 76841
    },
    {
      "epoch": 4.3280792614276065,
      "grad_norm": 0.4568573236465454,
      "learning_rate": 0.001,
      "loss": 2.1774,
      "step": 76884
    },
    {
      "epoch": 4.330499887412745,
      "grad_norm": 0.48163506388664246,
      "learning_rate": 0.001,
      "loss": 2.1805,
      "step": 76927
    },
    {
      "epoch": 4.332920513397883,
      "grad_norm": 0.37533774971961975,
      "learning_rate": 0.001,
      "loss": 2.1836,
      "step": 76970
    },
    {
      "epoch": 4.335341139383022,
      "grad_norm": 0.5097411274909973,
      "learning_rate": 0.001,
      "loss": 2.1809,
      "step": 77013
    },
    {
      "epoch": 4.33776176536816,
      "grad_norm": 0.46210816502571106,
      "learning_rate": 0.001,
      "loss": 2.1771,
      "step": 77056
    },
    {
      "epoch": 4.340182391353299,
      "grad_norm": 0.41375377774238586,
      "learning_rate": 0.001,
      "loss": 2.1891,
      "step": 77099
    },
    {
      "epoch": 4.342603017338437,
      "grad_norm": 0.4179149270057678,
      "learning_rate": 0.001,
      "loss": 2.1764,
      "step": 77142
    },
    {
      "epoch": 4.345023643323576,
      "grad_norm": 0.42984268069267273,
      "learning_rate": 0.001,
      "loss": 2.1782,
      "step": 77185
    },
    {
      "epoch": 4.347444269308714,
      "grad_norm": 0.4545655846595764,
      "learning_rate": 0.001,
      "loss": 2.1731,
      "step": 77228
    },
    {
      "epoch": 4.349864895293853,
      "grad_norm": 0.38662680983543396,
      "learning_rate": 0.001,
      "loss": 2.1769,
      "step": 77271
    },
    {
      "epoch": 4.352285521278991,
      "grad_norm": 0.5055582523345947,
      "learning_rate": 0.001,
      "loss": 2.1838,
      "step": 77314
    },
    {
      "epoch": 4.35470614726413,
      "grad_norm": 0.42296192049980164,
      "learning_rate": 0.001,
      "loss": 2.1768,
      "step": 77357
    },
    {
      "epoch": 4.357126773249268,
      "grad_norm": 0.447129487991333,
      "learning_rate": 0.001,
      "loss": 2.1697,
      "step": 77400
    },
    {
      "epoch": 4.3595473992344065,
      "grad_norm": 0.39407289028167725,
      "learning_rate": 0.001,
      "loss": 2.1892,
      "step": 77443
    },
    {
      "epoch": 4.361968025219545,
      "grad_norm": 0.4214778244495392,
      "learning_rate": 0.001,
      "loss": 2.1792,
      "step": 77486
    },
    {
      "epoch": 4.364388651204684,
      "grad_norm": 0.4254453778266907,
      "learning_rate": 0.001,
      "loss": 2.1886,
      "step": 77529
    },
    {
      "epoch": 4.366809277189822,
      "grad_norm": 0.5075379014015198,
      "learning_rate": 0.001,
      "loss": 2.1736,
      "step": 77572
    },
    {
      "epoch": 4.36922990317496,
      "grad_norm": 0.44919353723526,
      "learning_rate": 0.001,
      "loss": 2.177,
      "step": 77615
    },
    {
      "epoch": 4.371650529160099,
      "grad_norm": 0.42347481846809387,
      "learning_rate": 0.001,
      "loss": 2.1728,
      "step": 77658
    },
    {
      "epoch": 4.374071155145238,
      "grad_norm": 0.46982407569885254,
      "learning_rate": 0.001,
      "loss": 2.1826,
      "step": 77701
    },
    {
      "epoch": 4.376491781130376,
      "grad_norm": 0.4112502932548523,
      "learning_rate": 0.001,
      "loss": 2.1925,
      "step": 77744
    },
    {
      "epoch": 4.378912407115514,
      "grad_norm": 0.38474413752555847,
      "learning_rate": 0.001,
      "loss": 2.168,
      "step": 77787
    },
    {
      "epoch": 4.381333033100653,
      "grad_norm": 0.4048967659473419,
      "learning_rate": 0.001,
      "loss": 2.1782,
      "step": 77830
    },
    {
      "epoch": 4.383753659085792,
      "grad_norm": 0.4422946870326996,
      "learning_rate": 0.001,
      "loss": 2.181,
      "step": 77873
    },
    {
      "epoch": 4.38617428507093,
      "grad_norm": 0.370259553194046,
      "learning_rate": 0.001,
      "loss": 2.1657,
      "step": 77916
    },
    {
      "epoch": 4.388594911056068,
      "grad_norm": 0.5181688070297241,
      "learning_rate": 0.001,
      "loss": 2.1705,
      "step": 77959
    },
    {
      "epoch": 4.391015537041207,
      "grad_norm": 0.37978219985961914,
      "learning_rate": 0.001,
      "loss": 2.1802,
      "step": 78002
    },
    {
      "epoch": 4.393436163026346,
      "grad_norm": 0.497893750667572,
      "learning_rate": 0.001,
      "loss": 2.1799,
      "step": 78045
    },
    {
      "epoch": 4.395856789011484,
      "grad_norm": 0.3534224033355713,
      "learning_rate": 0.001,
      "loss": 2.1816,
      "step": 78088
    },
    {
      "epoch": 4.398277414996622,
      "grad_norm": 0.40843045711517334,
      "learning_rate": 0.001,
      "loss": 2.1677,
      "step": 78131
    },
    {
      "epoch": 4.40069804098176,
      "grad_norm": 0.45104557275772095,
      "learning_rate": 0.001,
      "loss": 2.1805,
      "step": 78174
    },
    {
      "epoch": 4.4031186669669,
      "grad_norm": 0.5691792368888855,
      "learning_rate": 0.001,
      "loss": 2.172,
      "step": 78217
    },
    {
      "epoch": 4.405539292952038,
      "grad_norm": 0.407808393239975,
      "learning_rate": 0.001,
      "loss": 2.1759,
      "step": 78260
    },
    {
      "epoch": 4.407959918937176,
      "grad_norm": 0.44516539573669434,
      "learning_rate": 0.001,
      "loss": 2.176,
      "step": 78303
    },
    {
      "epoch": 4.410380544922315,
      "grad_norm": 0.4218983054161072,
      "learning_rate": 0.001,
      "loss": 2.1864,
      "step": 78346
    },
    {
      "epoch": 4.4128011709074535,
      "grad_norm": 0.4233691692352295,
      "learning_rate": 0.001,
      "loss": 2.1774,
      "step": 78389
    },
    {
      "epoch": 4.415221796892592,
      "grad_norm": 0.417263925075531,
      "learning_rate": 0.001,
      "loss": 2.1804,
      "step": 78432
    },
    {
      "epoch": 4.41764242287773,
      "grad_norm": 0.4415801465511322,
      "learning_rate": 0.001,
      "loss": 2.1863,
      "step": 78475
    },
    {
      "epoch": 4.420063048862868,
      "grad_norm": 0.4235725700855255,
      "learning_rate": 0.001,
      "loss": 2.179,
      "step": 78518
    },
    {
      "epoch": 4.422483674848007,
      "grad_norm": 0.46427398920059204,
      "learning_rate": 0.001,
      "loss": 2.1779,
      "step": 78561
    },
    {
      "epoch": 4.424904300833146,
      "grad_norm": 0.3985130488872528,
      "learning_rate": 0.001,
      "loss": 2.1666,
      "step": 78604
    },
    {
      "epoch": 4.427324926818284,
      "grad_norm": 0.4967668354511261,
      "learning_rate": 0.001,
      "loss": 2.1792,
      "step": 78647
    },
    {
      "epoch": 4.429745552803423,
      "grad_norm": 0.4257707893848419,
      "learning_rate": 0.001,
      "loss": 2.1731,
      "step": 78690
    },
    {
      "epoch": 4.432166178788561,
      "grad_norm": 0.42839157581329346,
      "learning_rate": 0.001,
      "loss": 2.1722,
      "step": 78733
    },
    {
      "epoch": 4.4345868047737,
      "grad_norm": 0.4544007182121277,
      "learning_rate": 0.001,
      "loss": 2.172,
      "step": 78776
    },
    {
      "epoch": 4.437007430758838,
      "grad_norm": 0.4200676679611206,
      "learning_rate": 0.001,
      "loss": 2.171,
      "step": 78819
    },
    {
      "epoch": 4.439428056743976,
      "grad_norm": 0.4226556420326233,
      "learning_rate": 0.001,
      "loss": 2.1829,
      "step": 78862
    },
    {
      "epoch": 4.441848682729115,
      "grad_norm": 0.4018550515174866,
      "learning_rate": 0.001,
      "loss": 2.1851,
      "step": 78905
    },
    {
      "epoch": 4.442749380770097,
      "eval_ag_news_bleu_score": 4.821668057538916,
      "eval_ag_news_bleu_score_sem": 0.19037177690801724,
      "eval_ag_news_emb_cos_sim": 0.992409884929657,
      "eval_ag_news_emb_cos_sim_sem": 0.00026354596460667173,
      "eval_ag_news_emb_top1_equal": 0.6466666460037231,
      "eval_ag_news_emb_top1_equal_sem": 0.027643749338232177,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.4859116077423096,
      "eval_ag_news_n_ngrams_match_1": 5.428,
      "eval_ag_news_n_ngrams_match_2": 1.164,
      "eval_ag_news_n_ngrams_match_3": 0.32,
      "eval_ag_news_num_pred_words": 22.62,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.3035432935548004,
      "eval_ag_news_runtime": 15.4963,
      "eval_ag_news_samples_per_second": 32.266,
      "eval_ag_news_steps_per_second": 0.129,
      "eval_ag_news_token_set_f1": 0.2518705425774005,
      "eval_ag_news_token_set_f1_sem": 0.005217315336701915,
      "eval_ag_news_token_set_precision": 0.24548744981290535,
      "eval_ag_news_token_set_recall": 0.26622470192214387,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 78921
    },
    {
      "epoch": 4.442749380770097,
      "eval_anthropic_toxic_prompts_bleu_score": 6.594125639577043,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.28043646068661243,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9920156002044678,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00028919898172359413,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6766666769981384,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02705060760513538,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8072762489318848,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.576,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.978,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.878,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.664,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3271948268940454,
      "eval_anthropic_toxic_prompts_runtime": 6.7268,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.33,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.297,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.35775011226865183,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006850095681982673,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.43808091776697133,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.32584364431540735,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 78921
    },
    {
      "epoch": 4.442749380770097,
      "eval_arxiv_bleu_score": 4.289064162651348,
      "eval_arxiv_bleu_score_sem": 0.15667089590202926,
      "eval_arxiv_emb_cos_sim": 0.9919295310974121,
      "eval_arxiv_emb_cos_sim_sem": 0.00025500668074662455,
      "eval_arxiv_emb_top1_equal": 0.5833333134651184,
      "eval_arxiv_emb_top1_equal_sem": 0.02851131216637989,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5194473266601562,
      "eval_arxiv_n_ngrams_match_1": 5.328,
      "eval_arxiv_n_ngrams_match_2": 0.934,
      "eval_arxiv_n_ngrams_match_3": 0.22,
      "eval_arxiv_num_pred_words": 18.522,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.28822626853359823,
      "eval_arxiv_runtime": 6.6231,
      "eval_arxiv_samples_per_second": 75.494,
      "eval_arxiv_steps_per_second": 0.302,
      "eval_arxiv_token_set_f1": 0.26170795362058724,
      "eval_arxiv_token_set_f1_sem": 0.005390912521244125,
      "eval_arxiv_token_set_precision": 0.23202134374698868,
      "eval_arxiv_token_set_recall": 0.31277241577629417,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 78921
    },
    {
      "epoch": 4.442749380770097,
      "eval_python_code_alpaca_bleu_score": 6.709051538395968,
      "eval_python_code_alpaca_bleu_score_sem": 0.26307441053498803,
      "eval_python_code_alpaca_emb_cos_sim": 0.9903188347816467,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004382033996691466,
      "eval_python_code_alpaca_emb_top1_equal": 0.6433333158493042,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027702163273800266,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.8120052814483643,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.438,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.69,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.638,
      "eval_python_code_alpaca_num_pred_words": 20.344,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3603532797709731,
      "eval_python_code_alpaca_runtime": 7.2476,
      "eval_python_code_alpaca_samples_per_second": 68.989,
      "eval_python_code_alpaca_steps_per_second": 0.276,
      "eval_python_code_alpaca_token_set_f1": 0.3782946300372311,
      "eval_python_code_alpaca_token_set_f1_sem": 0.0059943483630002445,
      "eval_python_code_alpaca_token_set_precision": 0.3843040397548341,
      "eval_python_code_alpaca_token_set_recall": 0.3984642232041459,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 78921
    },
    {
      "epoch": 4.442749380770097,
      "eval_wikibio_bleu_score": 4.407115137350689,
      "eval_wikibio_bleu_score_sem": 0.19502550040901495,
      "eval_wikibio_emb_cos_sim": 0.9913690686225891,
      "eval_wikibio_emb_cos_sim_sem": 0.0003331853955084007,
      "eval_wikibio_emb_top1_equal": 0.6333333253860474,
      "eval_wikibio_emb_top1_equal_sem": 0.02786867456387452,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.896735429763794,
      "eval_wikibio_n_ngrams_match_1": 3.448,
      "eval_wikibio_n_ngrams_match_2": 1.02,
      "eval_wikibio_n_ngrams_match_3": 0.25,
      "eval_wikibio_num_pred_words": 19.662,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3239332442768289,
      "eval_wikibio_runtime": 6.5624,
      "eval_wikibio_samples_per_second": 76.191,
      "eval_wikibio_steps_per_second": 0.305,
      "eval_wikibio_token_set_f1": 0.227132010387748,
      "eval_wikibio_token_set_f1_sem": 0.005841394957587318,
      "eval_wikibio_token_set_precision": 0.26200100220788475,
      "eval_wikibio_token_set_recall": 0.20892928072787229,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 78921
    },
    {
      "epoch": 4.442749380770097,
      "eval_nq_5round_bleu_score": 15.873132283286242,
      "eval_nq_5round_bleu_score_sem": 0.7004190740732367,
      "eval_nq_5round_emb_cos_sim": 0.9945316910743713,
      "eval_nq_5round_emb_cos_sim_sem": 0.00032039296364555855,
      "eval_nq_5round_emb_top1_equal": 0.6833333373069763,
      "eval_nq_5round_emb_top1_equal_sem": 0.02690183265038281,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.708,
      "eval_nq_5round_n_ngrams_match_2": 4.996,
      "eval_nq_5round_n_ngrams_match_3": 2.676,
      "eval_nq_5round_num_pred_words": 24.008,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.48145395462016793,
      "eval_nq_5round_token_set_f1": 0.5018120932049799,
      "eval_nq_5round_token_set_f1_sem": 0.007141804514870429,
      "eval_nq_5round_token_set_precision": 0.48333029091814694,
      "eval_nq_5round_token_set_recall": 0.5291216510493538,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.069874083855538,
      "eval_nq_bleu_score_sem": 0.6112512633141012,
      "eval_nq_emb_cos_sim": 0.9939588904380798,
      "eval_nq_emb_cos_sim_sem": 0.0003221683930935198,
      "eval_nq_emb_top1_equal": 0.6566666960716248,
      "eval_nq_emb_top1_equal_sem": 0.027459642805142834,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.914692997932434,
      "eval_nq_n_ngrams_match_1": 10.87,
      "eval_nq_n_ngrams_match_2": 4.194,
      "eval_nq_n_ngrams_match_3": 2.07,
      "eval_nq_num_pred_words": 24.038,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4454297559448049,
      "eval_nq_runtime": 18.947,
      "eval_nq_samples_per_second": 26.389,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.4648038570565959,
      "eval_nq_token_set_f1_sem": 0.00657901542064565,
      "eval_nq_token_set_precision": 0.44808480296390313,
      "eval_nq_token_set_recall": 0.4901229060416104,
      "eval_nq_true_num_tokens": 32.0,
      "step": 78921
    },
    {
      "epoch": 4.4442693087142535,
      "grad_norm": 0.4159271717071533,
      "learning_rate": 0.001,
      "loss": 2.178,
      "step": 78948
    },
    {
      "epoch": 4.446689934699392,
      "grad_norm": 0.4909856915473938,
      "learning_rate": 0.001,
      "loss": 2.1829,
      "step": 78991
    },
    {
      "epoch": 4.449110560684531,
      "grad_norm": 0.36039721965789795,
      "learning_rate": 0.001,
      "loss": 2.166,
      "step": 79034
    },
    {
      "epoch": 4.451531186669669,
      "grad_norm": 0.42206457257270813,
      "learning_rate": 0.001,
      "loss": 2.1752,
      "step": 79077
    },
    {
      "epoch": 4.453951812654807,
      "grad_norm": 0.40058308839797974,
      "learning_rate": 0.001,
      "loss": 2.1755,
      "step": 79120
    },
    {
      "epoch": 4.456372438639946,
      "grad_norm": 0.4033859074115753,
      "learning_rate": 0.001,
      "loss": 2.181,
      "step": 79163
    },
    {
      "epoch": 4.458793064625085,
      "grad_norm": 0.40485793352127075,
      "learning_rate": 0.001,
      "loss": 2.1812,
      "step": 79206
    },
    {
      "epoch": 4.461213690610223,
      "grad_norm": 0.4741356670856476,
      "learning_rate": 0.001,
      "loss": 2.1708,
      "step": 79249
    },
    {
      "epoch": 4.463634316595361,
      "grad_norm": 0.4663408100605011,
      "learning_rate": 0.001,
      "loss": 2.1772,
      "step": 79292
    },
    {
      "epoch": 4.4660549425805,
      "grad_norm": 0.4639168977737427,
      "learning_rate": 0.001,
      "loss": 2.184,
      "step": 79335
    },
    {
      "epoch": 4.468475568565639,
      "grad_norm": 0.4147396981716156,
      "learning_rate": 0.001,
      "loss": 2.17,
      "step": 79378
    },
    {
      "epoch": 4.470896194550777,
      "grad_norm": 0.38119080662727356,
      "learning_rate": 0.001,
      "loss": 2.174,
      "step": 79421
    },
    {
      "epoch": 4.473316820535915,
      "grad_norm": 0.4268666207790375,
      "learning_rate": 0.001,
      "loss": 2.1737,
      "step": 79464
    },
    {
      "epoch": 4.4757374465210535,
      "grad_norm": 0.46784624457359314,
      "learning_rate": 0.001,
      "loss": 2.1845,
      "step": 79507
    },
    {
      "epoch": 4.478158072506193,
      "grad_norm": 0.36700618267059326,
      "learning_rate": 0.001,
      "loss": 2.1697,
      "step": 79550
    },
    {
      "epoch": 4.480578698491331,
      "grad_norm": 0.4395718574523926,
      "learning_rate": 0.001,
      "loss": 2.169,
      "step": 79593
    },
    {
      "epoch": 4.482999324476469,
      "grad_norm": 0.4249113202095032,
      "learning_rate": 0.001,
      "loss": 2.1792,
      "step": 79636
    },
    {
      "epoch": 4.4854199504616075,
      "grad_norm": 0.42579829692840576,
      "learning_rate": 0.001,
      "loss": 2.1856,
      "step": 79679
    },
    {
      "epoch": 4.487840576446747,
      "grad_norm": 0.4080027639865875,
      "learning_rate": 0.001,
      "loss": 2.1718,
      "step": 79722
    },
    {
      "epoch": 4.490261202431885,
      "grad_norm": 0.42231154441833496,
      "learning_rate": 0.001,
      "loss": 2.1722,
      "step": 79765
    },
    {
      "epoch": 4.492681828417023,
      "grad_norm": 0.3950883746147156,
      "learning_rate": 0.001,
      "loss": 2.1719,
      "step": 79808
    },
    {
      "epoch": 4.495102454402161,
      "grad_norm": 0.49257394671440125,
      "learning_rate": 0.001,
      "loss": 2.1803,
      "step": 79851
    },
    {
      "epoch": 4.4975230803873005,
      "grad_norm": 0.42617207765579224,
      "learning_rate": 0.001,
      "loss": 2.1683,
      "step": 79894
    },
    {
      "epoch": 4.499943706372439,
      "grad_norm": 0.43404704332351685,
      "learning_rate": 0.001,
      "loss": 2.1882,
      "step": 79937
    },
    {
      "epoch": 4.502364332357577,
      "grad_norm": 0.40333661437034607,
      "learning_rate": 0.001,
      "loss": 2.1787,
      "step": 79980
    },
    {
      "epoch": 4.504784958342715,
      "grad_norm": 0.3583446741104126,
      "learning_rate": 0.001,
      "loss": 2.1743,
      "step": 80023
    },
    {
      "epoch": 4.5072055843278545,
      "grad_norm": 0.4144943356513977,
      "learning_rate": 0.001,
      "loss": 2.1887,
      "step": 80066
    },
    {
      "epoch": 4.509626210312993,
      "grad_norm": 0.3918212056159973,
      "learning_rate": 0.001,
      "loss": 2.1678,
      "step": 80109
    },
    {
      "epoch": 4.512046836298131,
      "grad_norm": 0.4500170350074768,
      "learning_rate": 0.001,
      "loss": 2.1761,
      "step": 80152
    },
    {
      "epoch": 4.514467462283269,
      "grad_norm": 0.47215333580970764,
      "learning_rate": 0.001,
      "loss": 2.1786,
      "step": 80195
    },
    {
      "epoch": 4.516888088268408,
      "grad_norm": 0.43470123410224915,
      "learning_rate": 0.001,
      "loss": 2.1662,
      "step": 80238
    },
    {
      "epoch": 4.519308714253547,
      "grad_norm": 0.44340115785598755,
      "learning_rate": 0.001,
      "loss": 2.1639,
      "step": 80281
    },
    {
      "epoch": 4.521729340238685,
      "grad_norm": 0.44923287630081177,
      "learning_rate": 0.001,
      "loss": 2.1734,
      "step": 80324
    },
    {
      "epoch": 4.524149966223823,
      "grad_norm": 0.47960561513900757,
      "learning_rate": 0.001,
      "loss": 2.1642,
      "step": 80367
    },
    {
      "epoch": 4.526570592208962,
      "grad_norm": 0.43645933270454407,
      "learning_rate": 0.001,
      "loss": 2.1717,
      "step": 80410
    },
    {
      "epoch": 4.5289912181941006,
      "grad_norm": 0.3931477665901184,
      "learning_rate": 0.001,
      "loss": 2.1665,
      "step": 80453
    },
    {
      "epoch": 4.531411844179239,
      "grad_norm": 0.41931137442588806,
      "learning_rate": 0.001,
      "loss": 2.175,
      "step": 80496
    },
    {
      "epoch": 4.533832470164377,
      "grad_norm": 0.4167011082172394,
      "learning_rate": 0.001,
      "loss": 2.1736,
      "step": 80539
    },
    {
      "epoch": 4.536253096149516,
      "grad_norm": 0.42888668179512024,
      "learning_rate": 0.001,
      "loss": 2.1676,
      "step": 80582
    },
    {
      "epoch": 4.5386737221346545,
      "grad_norm": 0.46068641543388367,
      "learning_rate": 0.001,
      "loss": 2.1732,
      "step": 80625
    },
    {
      "epoch": 4.541094348119793,
      "grad_norm": 0.4141801595687866,
      "learning_rate": 0.001,
      "loss": 2.1655,
      "step": 80668
    },
    {
      "epoch": 4.543514974104931,
      "grad_norm": 0.41129541397094727,
      "learning_rate": 0.001,
      "loss": 2.1709,
      "step": 80711
    },
    {
      "epoch": 4.54593560009007,
      "grad_norm": 0.42347556352615356,
      "learning_rate": 0.001,
      "loss": 2.1762,
      "step": 80754
    },
    {
      "epoch": 4.548356226075208,
      "grad_norm": 0.4304637908935547,
      "learning_rate": 0.001,
      "loss": 2.1681,
      "step": 80797
    },
    {
      "epoch": 4.550776852060347,
      "grad_norm": 0.4054338335990906,
      "learning_rate": 0.001,
      "loss": 2.1705,
      "step": 80840
    },
    {
      "epoch": 4.553197478045485,
      "grad_norm": 0.4024193584918976,
      "learning_rate": 0.001,
      "loss": 2.1687,
      "step": 80883
    },
    {
      "epoch": 4.555618104030624,
      "grad_norm": 0.3873487710952759,
      "learning_rate": 0.001,
      "loss": 2.1711,
      "step": 80926
    },
    {
      "epoch": 4.558038730015762,
      "grad_norm": 0.49862486124038696,
      "learning_rate": 0.001,
      "loss": 2.1743,
      "step": 80969
    },
    {
      "epoch": 4.560459356000901,
      "grad_norm": 0.4342634379863739,
      "learning_rate": 0.001,
      "loss": 2.1671,
      "step": 81012
    },
    {
      "epoch": 4.562823688358478,
      "eval_ag_news_bleu_score": 4.84310583941333,
      "eval_ag_news_bleu_score_sem": 0.18517648992607447,
      "eval_ag_news_emb_cos_sim": 0.9925985932350159,
      "eval_ag_news_emb_cos_sim_sem": 0.0002690161412937138,
      "eval_ag_news_emb_top1_equal": 0.6600000262260437,
      "eval_ag_news_emb_top1_equal_sem": 0.0273952875069568,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.4761600494384766,
      "eval_ag_news_n_ngrams_match_1": 5.532,
      "eval_ag_news_n_ngrams_match_2": 1.148,
      "eval_ag_news_n_ngrams_match_3": 0.318,
      "eval_ag_news_num_pred_words": 22.53,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.31096883593394264,
      "eval_ag_news_runtime": 15.6938,
      "eval_ag_news_samples_per_second": 31.86,
      "eval_ag_news_steps_per_second": 0.127,
      "eval_ag_news_token_set_f1": 0.25724727528790997,
      "eval_ag_news_token_set_f1_sem": 0.005173843920959195,
      "eval_ag_news_token_set_precision": 0.25041011303963173,
      "eval_ag_news_token_set_recall": 0.27373945210022527,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 81054
    },
    {
      "epoch": 4.562823688358478,
      "eval_anthropic_toxic_prompts_bleu_score": 6.629546235785943,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.283355645824598,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9921368360519409,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002782946542052335,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6333333253860474,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02786867456387452,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.766996145248413,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.62,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.928,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.866,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.416,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3324903659072721,
      "eval_anthropic_toxic_prompts_runtime": 6.8528,
      "eval_anthropic_toxic_prompts_samples_per_second": 72.963,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.292,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3673403196145623,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006532126586863277,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.44335446413781865,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.33991012670164383,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 81054
    },
    {
      "epoch": 4.562823688358478,
      "eval_arxiv_bleu_score": 4.188328852564937,
      "eval_arxiv_bleu_score_sem": 0.14512687130100432,
      "eval_arxiv_emb_cos_sim": 0.9923128485679626,
      "eval_arxiv_emb_cos_sim_sem": 0.00022155934876157298,
      "eval_arxiv_emb_top1_equal": 0.5899999737739563,
      "eval_arxiv_emb_top1_equal_sem": 0.028443455370097265,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.498317241668701,
      "eval_arxiv_n_ngrams_match_1": 5.33,
      "eval_arxiv_n_ngrams_match_2": 0.916,
      "eval_arxiv_n_ngrams_match_3": 0.192,
      "eval_arxiv_num_pred_words": 18.146,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2892251257408493,
      "eval_arxiv_runtime": 6.6061,
      "eval_arxiv_samples_per_second": 75.688,
      "eval_arxiv_steps_per_second": 0.303,
      "eval_arxiv_token_set_f1": 0.26534321399748545,
      "eval_arxiv_token_set_f1_sem": 0.0053012469809167176,
      "eval_arxiv_token_set_precision": 0.23248359364370952,
      "eval_arxiv_token_set_recall": 0.3242091564894033,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 81054
    },
    {
      "epoch": 4.562823688358478,
      "eval_python_code_alpaca_bleu_score": 7.01366094358916,
      "eval_python_code_alpaca_bleu_score_sem": 0.2662734590334812,
      "eval_python_code_alpaca_emb_cos_sim": 0.9905908703804016,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0003215445274352633,
      "eval_python_code_alpaca_emb_top1_equal": 0.6299999952316284,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027921293391044915,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.7659196853637695,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.454,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.77,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.708,
      "eval_python_code_alpaca_num_pred_words": 20.472,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35553070459417324,
      "eval_python_code_alpaca_runtime": 6.6257,
      "eval_python_code_alpaca_samples_per_second": 75.463,
      "eval_python_code_alpaca_steps_per_second": 0.302,
      "eval_python_code_alpaca_token_set_f1": 0.3826391714879347,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006254579047580335,
      "eval_python_code_alpaca_token_set_precision": 0.3882205145780027,
      "eval_python_code_alpaca_token_set_recall": 0.4057336962277941,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 81054
    },
    {
      "epoch": 4.562823688358478,
      "eval_wikibio_bleu_score": 4.368645671512637,
      "eval_wikibio_bleu_score_sem": 0.14675738198184096,
      "eval_wikibio_emb_cos_sim": 0.9920841455459595,
      "eval_wikibio_emb_cos_sim_sem": 0.0002478789920766983,
      "eval_wikibio_emb_top1_equal": 0.6499999761581421,
      "eval_wikibio_emb_top1_equal_sem": 0.027583864257272155,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.875680685043335,
      "eval_wikibio_n_ngrams_match_1": 3.52,
      "eval_wikibio_n_ngrams_match_2": 1.024,
      "eval_wikibio_n_ngrams_match_3": 0.25,
      "eval_wikibio_num_pred_words": 20.128,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3286577097068035,
      "eval_wikibio_runtime": 6.5561,
      "eval_wikibio_samples_per_second": 76.265,
      "eval_wikibio_steps_per_second": 0.305,
      "eval_wikibio_token_set_f1": 0.2300209300732391,
      "eval_wikibio_token_set_f1_sem": 0.005511673911363225,
      "eval_wikibio_token_set_precision": 0.26906774293705604,
      "eval_wikibio_token_set_recall": 0.2096943222332327,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 81054
    },
    {
      "epoch": 4.562823688358478,
      "eval_nq_5round_bleu_score": 16.163037764253964,
      "eval_nq_5round_bleu_score_sem": 0.6998174695591868,
      "eval_nq_5round_emb_cos_sim": 0.9948708415031433,
      "eval_nq_5round_emb_cos_sim_sem": 0.0002166017742970324,
      "eval_nq_5round_emb_top1_equal": 0.6466666460037231,
      "eval_nq_5round_emb_top1_equal_sem": 0.027643749338232177,
      "eval_nq_5round_exact_match": 0.0,
      "eval_nq_5round_exact_match_sem": 0.0,
      "eval_nq_5round_n_ngrams_match_1": 11.844,
      "eval_nq_5round_n_ngrams_match_2": 5.112,
      "eval_nq_5round_n_ngrams_match_3": 2.736,
      "eval_nq_5round_num_pred_words": 24.0,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.48935509919302733,
      "eval_nq_5round_token_set_f1": 0.5079546055946796,
      "eval_nq_5round_token_set_f1_sem": 0.006967009933727219,
      "eval_nq_5round_token_set_precision": 0.48906253141965383,
      "eval_nq_5round_token_set_recall": 0.5362528681068262,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.147215349713496,
      "eval_nq_bleu_score_sem": 0.5536356023375396,
      "eval_nq_emb_cos_sim": 0.9942432045936584,
      "eval_nq_emb_cos_sim_sem": 0.00023184017356873106,
      "eval_nq_emb_top1_equal": 0.6733333468437195,
      "eval_nq_emb_top1_equal_sem": 0.027122635227651973,
      "eval_nq_exact_match": 0.0,
      "eval_nq_exact_match_sem": 0.0,
      "eval_nq_loss": 1.8965650796890259,
      "eval_nq_n_ngrams_match_1": 10.834,
      "eval_nq_n_ngrams_match_2": 4.234,
      "eval_nq_n_ngrams_match_3": 2.102,
      "eval_nq_num_pred_words": 24.11,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4406794803832336,
      "eval_nq_runtime": 18.8413,
      "eval_nq_samples_per_second": 26.537,
      "eval_nq_steps_per_second": 0.106,
      "eval_nq_token_set_f1": 0.46361311308985914,
      "eval_nq_token_set_f1_sem": 0.006280387706175517,
      "eval_nq_token_set_precision": 0.44416781383842885,
      "eval_nq_token_set_recall": 0.49543798338836276,
      "eval_nq_true_num_tokens": 32.0,
      "step": 81054
    },
    {
      "epoch": 4.562879981986039,
      "grad_norm": 0.42799586057662964,
      "learning_rate": 0.001,
      "loss": 2.1816,
      "step": 81055
    },
    {
      "epoch": 4.565300607971178,
      "grad_norm": 0.4626941382884979,
      "learning_rate": 0.001,
      "loss": 2.1806,
      "step": 81098
    },
    {
      "epoch": 4.567721233956316,
      "grad_norm": 0.4514025151729584,
      "learning_rate": 0.001,
      "loss": 2.1778,
      "step": 81141
    },
    {
      "epoch": 4.5701418599414545,
      "grad_norm": 0.47585391998291016,
      "learning_rate": 0.001,
      "loss": 2.1733,
      "step": 81184
    },
    {
      "epoch": 4.572562485926593,
      "grad_norm": 0.4325316846370697,
      "learning_rate": 0.001,
      "loss": 2.1596,
      "step": 81227
    },
    {
      "epoch": 4.574983111911732,
      "grad_norm": 0.48716941475868225,
      "learning_rate": 0.001,
      "loss": 2.1665,
      "step": 81270
    },
    {
      "epoch": 4.57740373789687,
      "grad_norm": 0.393110990524292,
      "learning_rate": 0.001,
      "loss": 2.173,
      "step": 81313
    },
    {
      "epoch": 4.579824363882008,
      "grad_norm": 0.4489499032497406,
      "learning_rate": 0.001,
      "loss": 2.1654,
      "step": 81356
    },
    {
      "epoch": 4.582244989867147,
      "grad_norm": 0.4425242245197296,
      "learning_rate": 0.001,
      "loss": 2.1662,
      "step": 81399
    },
    {
      "epoch": 4.584665615852286,
      "grad_norm": 0.41805747151374817,
      "learning_rate": 0.001,
      "loss": 2.1701,
      "step": 81442
    },
    {
      "epoch": 4.587086241837424,
      "grad_norm": 0.481441468000412,
      "learning_rate": 0.001,
      "loss": 2.1632,
      "step": 81485
    },
    {
      "epoch": 4.589506867822562,
      "grad_norm": 0.41779929399490356,
      "learning_rate": 0.001,
      "loss": 2.1693,
      "step": 81528
    },
    {
      "epoch": 4.591927493807701,
      "grad_norm": 0.4759984612464905,
      "learning_rate": 0.001,
      "loss": 2.162,
      "step": 81571
    },
    {
      "epoch": 4.59434811979284,
      "grad_norm": 0.4244650602340698,
      "learning_rate": 0.001,
      "loss": 2.1681,
      "step": 81614
    },
    {
      "epoch": 4.596768745777978,
      "grad_norm": 0.47671157121658325,
      "learning_rate": 0.001,
      "loss": 2.1741,
      "step": 81657
    },
    {
      "epoch": 4.599189371763116,
      "grad_norm": 0.46471601724624634,
      "learning_rate": 0.001,
      "loss": 2.174,
      "step": 81700
    },
    {
      "epoch": 4.6016099977482545,
      "grad_norm": 0.411458283662796,
      "learning_rate": 0.001,
      "loss": 2.1775,
      "step": 81743
    },
    {
      "epoch": 4.604030623733394,
      "grad_norm": 0.49985894560813904,
      "learning_rate": 0.001,
      "loss": 2.1728,
      "step": 81786
    },
    {
      "epoch": 4.606451249718532,
      "grad_norm": 0.3663926124572754,
      "learning_rate": 0.001,
      "loss": 2.1783,
      "step": 81829
    },
    {
      "epoch": 4.60887187570367,
      "grad_norm": 0.43184593319892883,
      "learning_rate": 0.001,
      "loss": 2.1753,
      "step": 81872
    },
    {
      "epoch": 4.611292501688808,
      "grad_norm": 0.4047897160053253,
      "learning_rate": 0.001,
      "loss": 2.1666,
      "step": 81915
    },
    {
      "epoch": 4.613713127673948,
      "grad_norm": 0.4517394006252289,
      "learning_rate": 0.001,
      "loss": 2.1667,
      "step": 81958
    },
    {
      "epoch": 4.616133753659086,
      "grad_norm": 0.44069617986679077,
      "learning_rate": 0.001,
      "loss": 2.1623,
      "step": 82001
    },
    {
      "epoch": 4.618554379644224,
      "grad_norm": 0.3872191607952118,
      "learning_rate": 0.001,
      "loss": 2.1742,
      "step": 82044
    },
    {
      "epoch": 4.620975005629363,
      "grad_norm": 0.405265748500824,
      "learning_rate": 0.001,
      "loss": 2.1698,
      "step": 82087
    },
    {
      "epoch": 4.6233956316145015,
      "grad_norm": 0.39199554920196533,
      "learning_rate": 0.001,
      "loss": 2.1673,
      "step": 82130
    },
    {
      "epoch": 4.62581625759964,
      "grad_norm": 0.43144914507865906,
      "learning_rate": 0.001,
      "loss": 2.1674,
      "step": 82173
    },
    {
      "epoch": 4.628236883584778,
      "grad_norm": 0.44490689039230347,
      "learning_rate": 0.001,
      "loss": 2.1794,
      "step": 82216
    },
    {
      "epoch": 4.630657509569916,
      "grad_norm": 0.4078328311443329,
      "learning_rate": 0.001,
      "loss": 2.164,
      "step": 82259
    },
    {
      "epoch": 4.633078135555055,
      "grad_norm": 0.4068223237991333,
      "learning_rate": 0.001,
      "loss": 2.1737,
      "step": 82302
    },
    {
      "epoch": 4.635498761540194,
      "grad_norm": 0.4786173105239868,
      "learning_rate": 0.001,
      "loss": 2.1695,
      "step": 82345
    },
    {
      "epoch": 4.637919387525332,
      "grad_norm": 0.4057946801185608,
      "learning_rate": 0.001,
      "loss": 2.173,
      "step": 82388
    },
    {
      "epoch": 4.640340013510471,
      "grad_norm": 0.3676082193851471,
      "learning_rate": 0.001,
      "loss": 2.1657,
      "step": 82431
    },
    {
      "epoch": 4.642760639495609,
      "grad_norm": 0.37304627895355225,
      "learning_rate": 0.001,
      "loss": 2.1631,
      "step": 82474
    },
    {
      "epoch": 4.645181265480748,
      "grad_norm": 0.40869560837745667,
      "learning_rate": 0.001,
      "loss": 2.1709,
      "step": 82517
    },
    {
      "epoch": 4.647601891465886,
      "grad_norm": 0.4178003668785095,
      "learning_rate": 0.001,
      "loss": 2.1556,
      "step": 82560
    },
    {
      "epoch": 4.650022517451024,
      "grad_norm": 0.4179328978061676,
      "learning_rate": 0.001,
      "loss": 2.1673,
      "step": 82603
    },
    {
      "epoch": 4.652443143436163,
      "grad_norm": 0.42980238795280457,
      "learning_rate": 0.001,
      "loss": 2.174,
      "step": 82646
    },
    {
      "epoch": 4.6548637694213015,
      "grad_norm": 0.45535463094711304,
      "learning_rate": 0.001,
      "loss": 2.1745,
      "step": 82689
    },
    {
      "epoch": 4.65728439540644,
      "grad_norm": 0.44546443223953247,
      "learning_rate": 0.001,
      "loss": 2.1668,
      "step": 82732
    },
    {
      "epoch": 4.659705021391579,
      "grad_norm": 0.3890509307384491,
      "learning_rate": 0.001,
      "loss": 2.1712,
      "step": 82775
    },
    {
      "epoch": 4.662125647376717,
      "grad_norm": 0.43610167503356934,
      "learning_rate": 0.001,
      "loss": 2.167,
      "step": 82818
    },
    {
      "epoch": 4.6645462733618555,
      "grad_norm": 0.43030595779418945,
      "learning_rate": 0.001,
      "loss": 2.1763,
      "step": 82861
    },
    {
      "epoch": 4.666966899346994,
      "grad_norm": 0.38396334648132324,
      "learning_rate": 0.001,
      "loss": 2.1706,
      "step": 82904
    },
    {
      "epoch": 4.669387525332132,
      "grad_norm": 0.45648977160453796,
      "learning_rate": 0.001,
      "loss": 2.1651,
      "step": 82947
    },
    {
      "epoch": 4.671808151317271,
      "grad_norm": 0.4790378212928772,
      "learning_rate": 0.001,
      "loss": 2.1667,
      "step": 82990
    },
    {
      "epoch": 4.674228777302409,
      "grad_norm": 0.502964198589325,
      "learning_rate": 0.001,
      "loss": 2.1667,
      "step": 83033
    },
    {
      "epoch": 4.676649403287548,
      "grad_norm": 0.428313285112381,
      "learning_rate": 0.001,
      "loss": 2.1638,
      "step": 83076
    },
    {
      "epoch": 4.679070029272687,
      "grad_norm": 0.3879488706588745,
      "learning_rate": 0.001,
      "loss": 2.1629,
      "step": 83119
    },
    {
      "epoch": 4.681490655257825,
      "grad_norm": 0.39904749393463135,
      "learning_rate": 0.001,
      "loss": 2.1767,
      "step": 83162
    },
    {
      "epoch": 4.682897995946859,
      "eval_ag_news_bleu_score": 4.716912166617226,
      "eval_ag_news_bleu_score_sem": 0.18367274936057174,
      "eval_ag_news_emb_cos_sim": 0.9924965500831604,
      "eval_ag_news_emb_cos_sim_sem": 0.00031242447451899306,
      "eval_ag_news_emb_top1_equal": 0.6499999761581421,
      "eval_ag_news_emb_top1_equal_sem": 0.027583864257272155,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.47908353805542,
      "eval_ag_news_n_ngrams_match_1": 5.504,
      "eval_ag_news_n_ngrams_match_2": 1.124,
      "eval_ag_news_n_ngrams_match_3": 0.304,
      "eval_ag_news_num_pred_words": 22.46,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.30797003985785254,
      "eval_ag_news_runtime": 15.608,
      "eval_ag_news_samples_per_second": 32.035,
      "eval_ag_news_steps_per_second": 0.128,
      "eval_ag_news_token_set_f1": 0.2570767824070472,
      "eval_ag_news_token_set_f1_sem": 0.005408732644713141,
      "eval_ag_news_token_set_precision": 0.25044040122138667,
      "eval_ag_news_token_set_recall": 0.2728357673323074,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 83187
    },
    {
      "epoch": 4.682897995946859,
      "eval_anthropic_toxic_prompts_bleu_score": 6.775740894107273,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.28472352361528297,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9924960732460022,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00024870038158719395,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6566666960716248,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027459642805142834,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.7599523067474365,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.724,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 2.046,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.916,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.84,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.3347503096323678,
      "eval_anthropic_toxic_prompts_runtime": 6.6659,
      "eval_anthropic_toxic_prompts_samples_per_second": 75.009,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.3,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.36910465841352275,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006499530887395656,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.45187449469890695,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.34005983853784544,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 83187
    },
    {
      "epoch": 4.682897995946859,
      "eval_arxiv_bleu_score": 4.1491454489483734,
      "eval_arxiv_bleu_score_sem": 0.13161792673944278,
      "eval_arxiv_emb_cos_sim": 0.9921376705169678,
      "eval_arxiv_emb_cos_sim_sem": 0.00025325407663561047,
      "eval_arxiv_emb_top1_equal": 0.6133333444595337,
      "eval_arxiv_emb_top1_equal_sem": 0.028163139369651306,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.5004703998565674,
      "eval_arxiv_n_ngrams_match_1": 5.26,
      "eval_arxiv_n_ngrams_match_2": 0.892,
      "eval_arxiv_n_ngrams_match_3": 0.2,
      "eval_arxiv_num_pred_words": 18.336,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.28467855988220564,
      "eval_arxiv_runtime": 6.643,
      "eval_arxiv_samples_per_second": 75.267,
      "eval_arxiv_steps_per_second": 0.301,
      "eval_arxiv_token_set_f1": 0.25903046076091624,
      "eval_arxiv_token_set_f1_sem": 0.005100995694644901,
      "eval_arxiv_token_set_precision": 0.2285343499660527,
      "eval_arxiv_token_set_recall": 0.31332487765192474,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 83187
    },
    {
      "epoch": 4.682897995946859,
      "eval_python_code_alpaca_bleu_score": 7.000969480887834,
      "eval_python_code_alpaca_bleu_score_sem": 0.2698195344005243,
      "eval_python_code_alpaca_emb_cos_sim": 0.9901992082595825,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0004584695566116292,
      "eval_python_code_alpaca_emb_top1_equal": 0.6399999856948853,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.02775911810844162,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.7734780311584473,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.428,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.788,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.714,
      "eval_python_code_alpaca_num_pred_words": 21.066,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.35791057305697727,
      "eval_python_code_alpaca_runtime": 6.64,
      "eval_python_code_alpaca_samples_per_second": 75.302,
      "eval_python_code_alpaca_steps_per_second": 0.301,
      "eval_python_code_alpaca_token_set_f1": 0.3813868602250411,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006290758358051402,
      "eval_python_code_alpaca_token_set_precision": 0.38387790871305993,
      "eval_python_code_alpaca_token_set_recall": 0.41376226481636713,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 83187
    },
    {
      "epoch": 4.682897995946859,
      "eval_wikibio_bleu_score": 4.265269712073308,
      "eval_wikibio_bleu_score_sem": 0.1425854909796595,
      "eval_wikibio_emb_cos_sim": 0.9918727874755859,
      "eval_wikibio_emb_cos_sim_sem": 0.0003187475498978054,
      "eval_wikibio_emb_top1_equal": 0.6466666460037231,
      "eval_wikibio_emb_top1_equal_sem": 0.027643749338232177,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.87795352935791,
      "eval_wikibio_n_ngrams_match_1": 3.394,
      "eval_wikibio_n_ngrams_match_2": 0.994,
      "eval_wikibio_n_ngrams_match_3": 0.242,
      "eval_wikibio_num_pred_words": 19.486,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3213421156159437,
      "eval_wikibio_runtime": 6.6001,
      "eval_wikibio_samples_per_second": 75.757,
      "eval_wikibio_steps_per_second": 0.303,
      "eval_wikibio_token_set_f1": 0.22450435953281206,
      "eval_wikibio_token_set_f1_sem": 0.005689096846375953,
      "eval_wikibio_token_set_precision": 0.2598370231361497,
      "eval_wikibio_token_set_recall": 0.20715065774991684,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 83187
    },
    {
      "epoch": 4.682897995946859,
      "eval_nq_5round_bleu_score": 16.246274828591496,
      "eval_nq_5round_bleu_score_sem": 0.7196800917916749,
      "eval_nq_5round_emb_cos_sim": 0.9947715401649475,
      "eval_nq_5round_emb_cos_sim_sem": 0.00031203786869413944,
      "eval_nq_5round_emb_top1_equal": 0.6666666865348816,
      "eval_nq_5round_emb_top1_equal_sem": 0.027262027544015993,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 11.882,
      "eval_nq_5round_n_ngrams_match_2": 5.186,
      "eval_nq_5round_n_ngrams_match_3": 2.764,
      "eval_nq_5round_num_pred_words": 23.984,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4880339051853573,
      "eval_nq_5round_token_set_f1": 0.508456256531921,
      "eval_nq_5round_token_set_f1_sem": 0.007384766765247546,
      "eval_nq_5round_token_set_precision": 0.48947200919672845,
      "eval_nq_5round_token_set_recall": 0.5362961736850763,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.310658114104443,
      "eval_nq_bleu_score_sem": 0.6160023063073823,
      "eval_nq_emb_cos_sim": 0.9944655299186707,
      "eval_nq_emb_cos_sim_sem": 0.00031019100588897125,
      "eval_nq_emb_top1_equal": 0.653333306312561,
      "eval_nq_emb_top1_equal_sem": 0.027522495986455002,
      "eval_nq_exact_match": 0.002,
      "eval_nq_exact_match_sem": 0.002,
      "eval_nq_loss": 1.8979929685592651,
      "eval_nq_n_ngrams_match_1": 10.846,
      "eval_nq_n_ngrams_match_2": 4.23,
      "eval_nq_n_ngrams_match_3": 2.146,
      "eval_nq_num_pred_words": 23.936,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4457826506464539,
      "eval_nq_runtime": 18.7789,
      "eval_nq_samples_per_second": 26.626,
      "eval_nq_steps_per_second": 0.107,
      "eval_nq_token_set_f1": 0.4666977634057931,
      "eval_nq_token_set_f1_sem": 0.006727192203604155,
      "eval_nq_token_set_precision": 0.4476166883134128,
      "eval_nq_token_set_recall": 0.4968940910076295,
      "eval_nq_true_num_tokens": 32.0,
      "step": 83187
    },
    {
      "epoch": 4.683911281242963,
      "grad_norm": 0.39875921607017517,
      "learning_rate": 0.001,
      "loss": 2.1684,
      "step": 83205
    },
    {
      "epoch": 4.686331907228102,
      "grad_norm": 0.4202415347099304,
      "learning_rate": 0.001,
      "loss": 2.169,
      "step": 83248
    },
    {
      "epoch": 4.68875253321324,
      "grad_norm": 0.39961162209510803,
      "learning_rate": 0.001,
      "loss": 2.1705,
      "step": 83291
    },
    {
      "epoch": 4.691173159198379,
      "grad_norm": 0.37972477078437805,
      "learning_rate": 0.001,
      "loss": 2.1661,
      "step": 83334
    },
    {
      "epoch": 4.693593785183517,
      "grad_norm": 0.4472878575325012,
      "learning_rate": 0.001,
      "loss": 2.1703,
      "step": 83377
    },
    {
      "epoch": 4.6960144111686555,
      "grad_norm": 0.42802074551582336,
      "learning_rate": 0.001,
      "loss": 2.1649,
      "step": 83420
    },
    {
      "epoch": 4.698435037153795,
      "grad_norm": 0.39894959330558777,
      "learning_rate": 0.001,
      "loss": 2.1582,
      "step": 83463
    },
    {
      "epoch": 4.700855663138933,
      "grad_norm": 0.4253007173538208,
      "learning_rate": 0.001,
      "loss": 2.1672,
      "step": 83506
    },
    {
      "epoch": 4.703276289124071,
      "grad_norm": 0.425686776638031,
      "learning_rate": 0.001,
      "loss": 2.1713,
      "step": 83549
    },
    {
      "epoch": 4.705696915109209,
      "grad_norm": 0.49295321106910706,
      "learning_rate": 0.001,
      "loss": 2.1527,
      "step": 83592
    },
    {
      "epoch": 4.708117541094348,
      "grad_norm": 0.43196389079093933,
      "learning_rate": 0.001,
      "loss": 2.1624,
      "step": 83635
    },
    {
      "epoch": 4.710538167079487,
      "grad_norm": 0.3788060247898102,
      "learning_rate": 0.001,
      "loss": 2.1699,
      "step": 83678
    },
    {
      "epoch": 4.712958793064625,
      "grad_norm": 0.477478951215744,
      "learning_rate": 0.001,
      "loss": 2.1604,
      "step": 83721
    },
    {
      "epoch": 4.715379419049763,
      "grad_norm": 0.48170551657676697,
      "learning_rate": 0.001,
      "loss": 2.1656,
      "step": 83764
    },
    {
      "epoch": 4.7178000450349025,
      "grad_norm": 0.37953945994377136,
      "learning_rate": 0.001,
      "loss": 2.16,
      "step": 83807
    },
    {
      "epoch": 4.720220671020041,
      "grad_norm": 0.436394065618515,
      "learning_rate": 0.001,
      "loss": 2.1685,
      "step": 83850
    },
    {
      "epoch": 4.722641297005179,
      "grad_norm": 0.4227990508079529,
      "learning_rate": 0.001,
      "loss": 2.1585,
      "step": 83893
    },
    {
      "epoch": 4.725061922990317,
      "grad_norm": 0.4594305157661438,
      "learning_rate": 0.001,
      "loss": 2.1692,
      "step": 83936
    },
    {
      "epoch": 4.7274825489754555,
      "grad_norm": 0.3859253525733948,
      "learning_rate": 0.001,
      "loss": 2.1694,
      "step": 83979
    },
    {
      "epoch": 4.729903174960595,
      "grad_norm": 0.4457191228866577,
      "learning_rate": 0.001,
      "loss": 2.1559,
      "step": 84022
    },
    {
      "epoch": 4.732323800945733,
      "grad_norm": 0.3857077658176422,
      "learning_rate": 0.001,
      "loss": 2.1668,
      "step": 84065
    },
    {
      "epoch": 4.734744426930871,
      "grad_norm": 0.42101576924324036,
      "learning_rate": 0.001,
      "loss": 2.1632,
      "step": 84108
    },
    {
      "epoch": 4.73716505291601,
      "grad_norm": 0.38279804587364197,
      "learning_rate": 0.001,
      "loss": 2.1703,
      "step": 84151
    },
    {
      "epoch": 4.739585678901149,
      "grad_norm": 0.4573419690132141,
      "learning_rate": 0.001,
      "loss": 2.1724,
      "step": 84194
    },
    {
      "epoch": 4.742006304886287,
      "grad_norm": 0.44520696997642517,
      "learning_rate": 0.001,
      "loss": 2.1685,
      "step": 84237
    },
    {
      "epoch": 4.744426930871425,
      "grad_norm": 0.48556455969810486,
      "learning_rate": 0.001,
      "loss": 2.1763,
      "step": 84280
    },
    {
      "epoch": 4.746847556856563,
      "grad_norm": 0.43710505962371826,
      "learning_rate": 0.001,
      "loss": 2.1663,
      "step": 84323
    },
    {
      "epoch": 4.7492681828417025,
      "grad_norm": 0.4232434928417206,
      "learning_rate": 0.001,
      "loss": 2.1646,
      "step": 84366
    },
    {
      "epoch": 4.751688808826841,
      "grad_norm": 0.4203433096408844,
      "learning_rate": 0.001,
      "loss": 2.1641,
      "step": 84409
    },
    {
      "epoch": 4.754109434811979,
      "grad_norm": 0.4102879762649536,
      "learning_rate": 0.001,
      "loss": 2.1768,
      "step": 84452
    },
    {
      "epoch": 4.756530060797118,
      "grad_norm": 0.401623398065567,
      "learning_rate": 0.001,
      "loss": 2.1704,
      "step": 84495
    },
    {
      "epoch": 4.758950686782256,
      "grad_norm": 0.4311046004295349,
      "learning_rate": 0.001,
      "loss": 2.1703,
      "step": 84538
    },
    {
      "epoch": 4.761371312767395,
      "grad_norm": 0.42004165053367615,
      "learning_rate": 0.001,
      "loss": 2.1654,
      "step": 84581
    },
    {
      "epoch": 4.763791938752533,
      "grad_norm": 0.4006766378879547,
      "learning_rate": 0.001,
      "loss": 2.1602,
      "step": 84624
    },
    {
      "epoch": 4.766212564737672,
      "grad_norm": 0.4127703309059143,
      "learning_rate": 0.001,
      "loss": 2.1651,
      "step": 84667
    },
    {
      "epoch": 4.76863319072281,
      "grad_norm": 0.4062407314777374,
      "learning_rate": 0.001,
      "loss": 2.1676,
      "step": 84710
    },
    {
      "epoch": 4.771053816707949,
      "grad_norm": 0.45703238248825073,
      "learning_rate": 0.001,
      "loss": 2.1628,
      "step": 84753
    },
    {
      "epoch": 4.773474442693087,
      "grad_norm": 0.3892706036567688,
      "learning_rate": 0.001,
      "loss": 2.1618,
      "step": 84796
    },
    {
      "epoch": 4.775895068678226,
      "grad_norm": 0.4671939015388489,
      "learning_rate": 0.001,
      "loss": 2.1585,
      "step": 84839
    },
    {
      "epoch": 4.778315694663364,
      "grad_norm": 0.38782036304473877,
      "learning_rate": 0.001,
      "loss": 2.1592,
      "step": 84882
    },
    {
      "epoch": 4.7807363206485025,
      "grad_norm": 0.438733845949173,
      "learning_rate": 0.001,
      "loss": 2.1619,
      "step": 84925
    },
    {
      "epoch": 4.783156946633641,
      "grad_norm": 0.45775145292282104,
      "learning_rate": 0.001,
      "loss": 2.1584,
      "step": 84968
    },
    {
      "epoch": 4.78557757261878,
      "grad_norm": 0.4601728022098541,
      "learning_rate": 0.001,
      "loss": 2.1536,
      "step": 85011
    },
    {
      "epoch": 4.787998198603918,
      "grad_norm": 0.4032277464866638,
      "learning_rate": 0.001,
      "loss": 2.1787,
      "step": 85054
    },
    {
      "epoch": 4.790418824589056,
      "grad_norm": 0.41465839743614197,
      "learning_rate": 0.001,
      "loss": 2.1637,
      "step": 85097
    },
    {
      "epoch": 4.792839450574195,
      "grad_norm": 0.36137452721595764,
      "learning_rate": 0.001,
      "loss": 2.1757,
      "step": 85140
    },
    {
      "epoch": 4.795260076559334,
      "grad_norm": 0.3882429599761963,
      "learning_rate": 0.001,
      "loss": 2.1668,
      "step": 85183
    },
    {
      "epoch": 4.797680702544472,
      "grad_norm": 0.395302414894104,
      "learning_rate": 0.001,
      "loss": 2.1563,
      "step": 85226
    },
    {
      "epoch": 4.80010132852961,
      "grad_norm": 0.3965873122215271,
      "learning_rate": 0.001,
      "loss": 2.161,
      "step": 85269
    },
    {
      "epoch": 4.802521954514749,
      "grad_norm": 0.4338971972465515,
      "learning_rate": 0.001,
      "loss": 2.1696,
      "step": 85312
    },
    {
      "epoch": 4.80297230353524,
      "eval_ag_news_bleu_score": 5.030141780362625,
      "eval_ag_news_bleu_score_sem": 0.18692293381226216,
      "eval_ag_news_emb_cos_sim": 0.9927579760551453,
      "eval_ag_news_emb_cos_sim_sem": 0.00024671226516562926,
      "eval_ag_news_emb_top1_equal": 0.6833333373069763,
      "eval_ag_news_emb_top1_equal_sem": 0.026901834371020696,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.447333812713623,
      "eval_ag_news_n_ngrams_match_1": 5.554,
      "eval_ag_news_n_ngrams_match_2": 1.202,
      "eval_ag_news_n_ngrams_match_3": 0.332,
      "eval_ag_news_num_pred_words": 22.314,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.3129971126773298,
      "eval_ag_news_runtime": 15.2472,
      "eval_ag_news_samples_per_second": 32.793,
      "eval_ag_news_steps_per_second": 0.131,
      "eval_ag_news_token_set_f1": 0.25924553485840773,
      "eval_ag_news_token_set_f1_sem": 0.005396354197165158,
      "eval_ag_news_token_set_precision": 0.25292828335336176,
      "eval_ag_news_token_set_recall": 0.27496943235568266,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 85320
    },
    {
      "epoch": 4.80297230353524,
      "eval_anthropic_toxic_prompts_bleu_score": 6.826581452223687,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.28898595160153495,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9923011660575867,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.0002651221226795415,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6333333253860474,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.02786867456387452,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.8018245697021484,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.628,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 2.016,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.902,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.554,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.33144107109920756,
      "eval_anthropic_toxic_prompts_runtime": 6.6802,
      "eval_anthropic_toxic_prompts_samples_per_second": 74.848,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.299,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3607760212653973,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.0065310511556016275,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.44230410059053743,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.32969193830361676,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 85320
    },
    {
      "epoch": 4.80297230353524,
      "eval_arxiv_bleu_score": 4.171612515792814,
      "eval_arxiv_bleu_score_sem": 0.12969386830670285,
      "eval_arxiv_emb_cos_sim": 0.9920535683631897,
      "eval_arxiv_emb_cos_sim_sem": 0.00023911486923798277,
      "eval_arxiv_emb_top1_equal": 0.5799999833106995,
      "eval_arxiv_emb_top1_equal_sem": 0.02854322483723857,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.507985830307007,
      "eval_arxiv_n_ngrams_match_1": 5.346,
      "eval_arxiv_n_ngrams_match_2": 0.918,
      "eval_arxiv_n_ngrams_match_3": 0.182,
      "eval_arxiv_num_pred_words": 18.486,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.2900857348074477,
      "eval_arxiv_runtime": 6.5823,
      "eval_arxiv_samples_per_second": 75.961,
      "eval_arxiv_steps_per_second": 0.304,
      "eval_arxiv_token_set_f1": 0.2626732006943977,
      "eval_arxiv_token_set_f1_sem": 0.005091170724038379,
      "eval_arxiv_token_set_precision": 0.2341133680990495,
      "eval_arxiv_token_set_recall": 0.3113394281755098,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 85320
    },
    {
      "epoch": 4.80297230353524,
      "eval_python_code_alpaca_bleu_score": 7.085096079866166,
      "eval_python_code_alpaca_bleu_score_sem": 0.28868014920459706,
      "eval_python_code_alpaca_emb_cos_sim": 0.9910838007926941,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.00028423491453953326,
      "eval_python_code_alpaca_emb_top1_equal": 0.6266666650772095,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027972489250684164,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.8079309463500977,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.498,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.784,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.716,
      "eval_python_code_alpaca_num_pred_words": 20.78,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3646798955138413,
      "eval_python_code_alpaca_runtime": 6.7281,
      "eval_python_code_alpaca_samples_per_second": 74.315,
      "eval_python_code_alpaca_steps_per_second": 0.297,
      "eval_python_code_alpaca_token_set_f1": 0.38465030829417823,
      "eval_python_code_alpaca_token_set_f1_sem": 0.006089971676490153,
      "eval_python_code_alpaca_token_set_precision": 0.39274859321897404,
      "eval_python_code_alpaca_token_set_recall": 0.40281209232220966,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 85320
    },
    {
      "epoch": 4.80297230353524,
      "eval_wikibio_bleu_score": 4.540165387055943,
      "eval_wikibio_bleu_score_sem": 0.15770056394971477,
      "eval_wikibio_emb_cos_sim": 0.9922193884849548,
      "eval_wikibio_emb_cos_sim_sem": 0.0002590801565152403,
      "eval_wikibio_emb_top1_equal": 0.653333306312561,
      "eval_wikibio_emb_top1_equal_sem": 0.027522495986455002,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.85897159576416,
      "eval_wikibio_n_ngrams_match_1": 3.536,
      "eval_wikibio_n_ngrams_match_2": 1.054,
      "eval_wikibio_n_ngrams_match_3": 0.252,
      "eval_wikibio_num_pred_words": 19.446,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.3339419243521582,
      "eval_wikibio_runtime": 6.4965,
      "eval_wikibio_samples_per_second": 76.964,
      "eval_wikibio_steps_per_second": 0.308,
      "eval_wikibio_token_set_f1": 0.23577928628541017,
      "eval_wikibio_token_set_f1_sem": 0.005573130116238389,
      "eval_wikibio_token_set_precision": 0.2712988063834134,
      "eval_wikibio_token_set_recall": 0.21524945930208383,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 85320
    },
    {
      "epoch": 4.80297230353524,
      "eval_nq_5round_bleu_score": 15.60492848994167,
      "eval_nq_5round_bleu_score_sem": 0.688273623106652,
      "eval_nq_5round_emb_cos_sim": 0.9948824048042297,
      "eval_nq_5round_emb_cos_sim_sem": 0.0003071714477100694,
      "eval_nq_5round_emb_top1_equal": 0.6700000166893005,
      "eval_nq_5round_emb_top1_equal_sem": 0.027193103952244484,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 11.626,
      "eval_nq_5round_n_ngrams_match_2": 4.882,
      "eval_nq_5round_n_ngrams_match_3": 2.598,
      "eval_nq_5round_num_pred_words": 23.68,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4844367520797786,
      "eval_nq_5round_token_set_f1": 0.5010092346370434,
      "eval_nq_5round_token_set_f1_sem": 0.007105482930037595,
      "eval_nq_5round_token_set_precision": 0.4822610555092911,
      "eval_nq_5round_token_set_recall": 0.5285965063000444,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.297339923904248,
      "eval_nq_bleu_score_sem": 0.5921139247894901,
      "eval_nq_emb_cos_sim": 0.9946077466011047,
      "eval_nq_emb_cos_sim_sem": 0.00021010105649904625,
      "eval_nq_emb_top1_equal": 0.7066666483879089,
      "eval_nq_emb_top1_equal_sem": 0.026330095652579162,
      "eval_nq_exact_match": 0.002,
      "eval_nq_exact_match_sem": 0.002,
      "eval_nq_loss": 1.8932467699050903,
      "eval_nq_n_ngrams_match_1": 10.78,
      "eval_nq_n_ngrams_match_2": 4.26,
      "eval_nq_n_ngrams_match_3": 2.114,
      "eval_nq_num_pred_words": 23.752,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.44233804313217834,
      "eval_nq_runtime": 19.3567,
      "eval_nq_samples_per_second": 25.831,
      "eval_nq_steps_per_second": 0.103,
      "eval_nq_token_set_f1": 0.4637964115147553,
      "eval_nq_token_set_f1_sem": 0.006679598687925456,
      "eval_nq_token_set_precision": 0.4439520508503713,
      "eval_nq_token_set_recall": 0.4927574039233326,
      "eval_nq_true_num_tokens": 32.0,
      "step": 85320
    },
    {
      "epoch": 4.804942580499888,
      "grad_norm": 0.4357750415802002,
      "learning_rate": 0.001,
      "loss": 2.1641,
      "step": 85355
    },
    {
      "epoch": 4.807363206485026,
      "grad_norm": 0.40105295181274414,
      "learning_rate": 0.001,
      "loss": 2.1562,
      "step": 85398
    },
    {
      "epoch": 4.809783832470164,
      "grad_norm": 0.463548481464386,
      "learning_rate": 0.001,
      "loss": 2.1589,
      "step": 85441
    },
    {
      "epoch": 4.8122044584553025,
      "grad_norm": 0.4390972852706909,
      "learning_rate": 0.001,
      "loss": 2.1606,
      "step": 85484
    },
    {
      "epoch": 4.814625084440442,
      "grad_norm": 0.41914552450180054,
      "learning_rate": 0.001,
      "loss": 2.1633,
      "step": 85527
    },
    {
      "epoch": 4.81704571042558,
      "grad_norm": 0.4117737114429474,
      "learning_rate": 0.001,
      "loss": 2.1697,
      "step": 85570
    },
    {
      "epoch": 4.819466336410718,
      "grad_norm": 0.5017421841621399,
      "learning_rate": 0.001,
      "loss": 2.1631,
      "step": 85613
    },
    {
      "epoch": 4.8218869623958565,
      "grad_norm": 0.4555579721927643,
      "learning_rate": 0.001,
      "loss": 2.1597,
      "step": 85656
    },
    {
      "epoch": 4.824307588380996,
      "grad_norm": 0.3921297788619995,
      "learning_rate": 0.001,
      "loss": 2.1638,
      "step": 85699
    },
    {
      "epoch": 4.826728214366134,
      "grad_norm": 0.43109309673309326,
      "learning_rate": 0.001,
      "loss": 2.1618,
      "step": 85742
    },
    {
      "epoch": 4.829148840351272,
      "grad_norm": 0.4430307447910309,
      "learning_rate": 0.001,
      "loss": 2.1576,
      "step": 85785
    },
    {
      "epoch": 4.83156946633641,
      "grad_norm": 0.44259944558143616,
      "learning_rate": 0.001,
      "loss": 2.1668,
      "step": 85828
    },
    {
      "epoch": 4.8339900923215495,
      "grad_norm": 0.44098979234695435,
      "learning_rate": 0.001,
      "loss": 2.1641,
      "step": 85871
    },
    {
      "epoch": 4.836410718306688,
      "grad_norm": 0.4537988007068634,
      "learning_rate": 0.001,
      "loss": 2.1577,
      "step": 85914
    },
    {
      "epoch": 4.838831344291826,
      "grad_norm": 0.3977867066860199,
      "learning_rate": 0.001,
      "loss": 2.164,
      "step": 85957
    },
    {
      "epoch": 4.841251970276964,
      "grad_norm": 0.44861045479774475,
      "learning_rate": 0.001,
      "loss": 2.1678,
      "step": 86000
    },
    {
      "epoch": 4.8436725962621034,
      "grad_norm": 0.4315987825393677,
      "learning_rate": 0.001,
      "loss": 2.1578,
      "step": 86043
    },
    {
      "epoch": 4.846093222247242,
      "grad_norm": 0.3949969410896301,
      "learning_rate": 0.001,
      "loss": 2.1526,
      "step": 86086
    },
    {
      "epoch": 4.84851384823238,
      "grad_norm": 0.4183890223503113,
      "learning_rate": 0.001,
      "loss": 2.1625,
      "step": 86129
    },
    {
      "epoch": 4.850934474217518,
      "grad_norm": 0.47212621569633484,
      "learning_rate": 0.001,
      "loss": 2.1738,
      "step": 86172
    },
    {
      "epoch": 4.853355100202657,
      "grad_norm": 0.368083119392395,
      "learning_rate": 0.001,
      "loss": 2.1516,
      "step": 86215
    },
    {
      "epoch": 4.855775726187796,
      "grad_norm": 0.4396151900291443,
      "learning_rate": 0.001,
      "loss": 2.1512,
      "step": 86258
    },
    {
      "epoch": 4.858196352172934,
      "grad_norm": 0.39211124181747437,
      "learning_rate": 0.001,
      "loss": 2.1647,
      "step": 86301
    },
    {
      "epoch": 4.860616978158072,
      "grad_norm": 0.44352832436561584,
      "learning_rate": 0.001,
      "loss": 2.1555,
      "step": 86344
    },
    {
      "epoch": 4.863037604143211,
      "grad_norm": 0.43583452701568604,
      "learning_rate": 0.001,
      "loss": 2.1714,
      "step": 86387
    },
    {
      "epoch": 4.8654582301283495,
      "grad_norm": 0.40865105390548706,
      "learning_rate": 0.001,
      "loss": 2.1635,
      "step": 86430
    },
    {
      "epoch": 4.867878856113488,
      "grad_norm": 0.38110360503196716,
      "learning_rate": 0.001,
      "loss": 2.1659,
      "step": 86473
    },
    {
      "epoch": 4.870299482098626,
      "grad_norm": 0.42054539918899536,
      "learning_rate": 0.001,
      "loss": 2.1666,
      "step": 86516
    },
    {
      "epoch": 4.872720108083765,
      "grad_norm": 0.4169979691505432,
      "learning_rate": 0.001,
      "loss": 2.1509,
      "step": 86559
    },
    {
      "epoch": 4.8751407340689035,
      "grad_norm": 0.40312865376472473,
      "learning_rate": 0.001,
      "loss": 2.1564,
      "step": 86602
    },
    {
      "epoch": 4.877561360054042,
      "grad_norm": 0.42840850353240967,
      "learning_rate": 0.001,
      "loss": 2.1614,
      "step": 86645
    },
    {
      "epoch": 4.87998198603918,
      "grad_norm": 0.43374863266944885,
      "learning_rate": 0.001,
      "loss": 2.1523,
      "step": 86688
    },
    {
      "epoch": 4.882402612024319,
      "grad_norm": 0.4378853738307953,
      "learning_rate": 0.001,
      "loss": 2.1632,
      "step": 86731
    },
    {
      "epoch": 4.884823238009457,
      "grad_norm": 0.4291220009326935,
      "learning_rate": 0.001,
      "loss": 2.156,
      "step": 86774
    },
    {
      "epoch": 4.887243863994596,
      "grad_norm": 0.45905306935310364,
      "learning_rate": 0.001,
      "loss": 2.1539,
      "step": 86817
    },
    {
      "epoch": 4.889664489979734,
      "grad_norm": 0.38533565402030945,
      "learning_rate": 0.001,
      "loss": 2.1537,
      "step": 86860
    },
    {
      "epoch": 4.892085115964873,
      "grad_norm": 0.43840882182121277,
      "learning_rate": 0.001,
      "loss": 2.1509,
      "step": 86903
    },
    {
      "epoch": 4.894505741950011,
      "grad_norm": 0.37712955474853516,
      "learning_rate": 0.001,
      "loss": 2.1762,
      "step": 86946
    },
    {
      "epoch": 4.89692636793515,
      "grad_norm": 0.4128488302230835,
      "learning_rate": 0.001,
      "loss": 2.1633,
      "step": 86989
    },
    {
      "epoch": 4.899346993920288,
      "grad_norm": 0.40435758233070374,
      "learning_rate": 0.001,
      "loss": 2.1594,
      "step": 87032
    },
    {
      "epoch": 4.901767619905427,
      "grad_norm": 0.4197578430175781,
      "learning_rate": 0.001,
      "loss": 2.1581,
      "step": 87075
    },
    {
      "epoch": 4.904188245890565,
      "grad_norm": 0.4019104838371277,
      "learning_rate": 0.001,
      "loss": 2.1602,
      "step": 87118
    },
    {
      "epoch": 4.9066088718757035,
      "grad_norm": 0.40227821469306946,
      "learning_rate": 0.001,
      "loss": 2.1596,
      "step": 87161
    },
    {
      "epoch": 4.909029497860843,
      "grad_norm": 0.43067941069602966,
      "learning_rate": 0.001,
      "loss": 2.16,
      "step": 87204
    },
    {
      "epoch": 4.911450123845981,
      "grad_norm": 0.394469678401947,
      "learning_rate": 0.001,
      "loss": 2.1537,
      "step": 87247
    },
    {
      "epoch": 4.913870749831119,
      "grad_norm": 0.4136824607849121,
      "learning_rate": 0.001,
      "loss": 2.1544,
      "step": 87290
    },
    {
      "epoch": 4.916291375816257,
      "grad_norm": 0.45566579699516296,
      "learning_rate": 0.001,
      "loss": 2.1627,
      "step": 87333
    },
    {
      "epoch": 4.918712001801396,
      "grad_norm": 0.43493157625198364,
      "learning_rate": 0.001,
      "loss": 2.1732,
      "step": 87376
    },
    {
      "epoch": 4.921132627786535,
      "grad_norm": 0.41205358505249023,
      "learning_rate": 0.001,
      "loss": 2.1602,
      "step": 87419
    },
    {
      "epoch": 4.92304661112362,
      "eval_ag_news_bleu_score": 4.89929831643001,
      "eval_ag_news_bleu_score_sem": 0.1835120470058315,
      "eval_ag_news_emb_cos_sim": 0.9923436641693115,
      "eval_ag_news_emb_cos_sim_sem": 0.0002921573766026669,
      "eval_ag_news_emb_top1_equal": 0.6899999976158142,
      "eval_ag_news_emb_top1_equal_sem": 0.026746674129075776,
      "eval_ag_news_exact_match": 0.0,
      "eval_ag_news_exact_match_sem": 0.0,
      "eval_ag_news_loss": 3.475796937942505,
      "eval_ag_news_n_ngrams_match_1": 5.676,
      "eval_ag_news_n_ngrams_match_2": 1.222,
      "eval_ag_news_n_ngrams_match_3": 0.324,
      "eval_ag_news_num_pred_words": 22.582,
      "eval_ag_news_num_true_words": 21.938,
      "eval_ag_news_pred_num_tokens": 31.0,
      "eval_ag_news_rouge_score": 0.3185973871361046,
      "eval_ag_news_runtime": 15.2288,
      "eval_ag_news_samples_per_second": 32.833,
      "eval_ag_news_steps_per_second": 0.131,
      "eval_ag_news_token_set_f1": 0.2648648260043954,
      "eval_ag_news_token_set_f1_sem": 0.00533764671773529,
      "eval_ag_news_token_set_precision": 0.2572172790433889,
      "eval_ag_news_token_set_recall": 0.2821940631804298,
      "eval_ag_news_true_num_tokens": 31.7265625,
      "step": 87453
    },
    {
      "epoch": 4.92304661112362,
      "eval_anthropic_toxic_prompts_bleu_score": 6.653195208606232,
      "eval_anthropic_toxic_prompts_bleu_score_sem": 0.2793297674226269,
      "eval_anthropic_toxic_prompts_emb_cos_sim": 0.9922748804092407,
      "eval_anthropic_toxic_prompts_emb_cos_sim_sem": 0.00023759887972116456,
      "eval_anthropic_toxic_prompts_emb_top1_equal": 0.6600000262260437,
      "eval_anthropic_toxic_prompts_emb_top1_equal_sem": 0.027395285786318915,
      "eval_anthropic_toxic_prompts_exact_match": 0.0,
      "eval_anthropic_toxic_prompts_exact_match_sem": 0.0,
      "eval_anthropic_toxic_prompts_loss": 2.7825043201446533,
      "eval_anthropic_toxic_prompts_n_ngrams_match_1": 5.676,
      "eval_anthropic_toxic_prompts_n_ngrams_match_2": 1.98,
      "eval_anthropic_toxic_prompts_n_ngrams_match_3": 0.86,
      "eval_anthropic_toxic_prompts_num_pred_words": 24.484,
      "eval_anthropic_toxic_prompts_num_true_words": 13.422,
      "eval_anthropic_toxic_prompts_pred_num_tokens": 31.0,
      "eval_anthropic_toxic_prompts_rouge_score": 0.336345380767139,
      "eval_anthropic_toxic_prompts_runtime": 6.7818,
      "eval_anthropic_toxic_prompts_samples_per_second": 73.726,
      "eval_anthropic_toxic_prompts_steps_per_second": 0.295,
      "eval_anthropic_toxic_prompts_token_set_f1": 0.3688550436483949,
      "eval_anthropic_toxic_prompts_token_set_f1_sem": 0.006697991014684084,
      "eval_anthropic_toxic_prompts_token_set_precision": 0.4485522660076778,
      "eval_anthropic_toxic_prompts_token_set_recall": 0.3377044396429321,
      "eval_anthropic_toxic_prompts_true_num_tokens": 16.6015625,
      "step": 87453
    },
    {
      "epoch": 4.92304661112362,
      "eval_arxiv_bleu_score": 4.257714709008939,
      "eval_arxiv_bleu_score_sem": 0.14032271971437568,
      "eval_arxiv_emb_cos_sim": 0.9920233488082886,
      "eval_arxiv_emb_cos_sim_sem": 0.00023793461918853397,
      "eval_arxiv_emb_top1_equal": 0.5933333039283752,
      "eval_arxiv_emb_top1_equal_sem": 0.028407504362121784,
      "eval_arxiv_exact_match": 0.0,
      "eval_arxiv_exact_match_sem": 0.0,
      "eval_arxiv_loss": 3.483067274093628,
      "eval_arxiv_n_ngrams_match_1": 5.46,
      "eval_arxiv_n_ngrams_match_2": 0.938,
      "eval_arxiv_n_ngrams_match_3": 0.194,
      "eval_arxiv_num_pred_words": 18.712,
      "eval_arxiv_num_true_words": 21.532,
      "eval_arxiv_pred_num_tokens": 31.0,
      "eval_arxiv_rouge_score": 0.29388356148951056,
      "eval_arxiv_runtime": 6.6696,
      "eval_arxiv_samples_per_second": 74.967,
      "eval_arxiv_steps_per_second": 0.3,
      "eval_arxiv_token_set_f1": 0.26450933611545047,
      "eval_arxiv_token_set_f1_sem": 0.005163638612445233,
      "eval_arxiv_token_set_precision": 0.2369693647905716,
      "eval_arxiv_token_set_recall": 0.31130350774291154,
      "eval_arxiv_true_num_tokens": 32.0,
      "step": 87453
    },
    {
      "epoch": 4.92304661112362,
      "eval_python_code_alpaca_bleu_score": 6.975729457750516,
      "eval_python_code_alpaca_bleu_score_sem": 0.2715802593149306,
      "eval_python_code_alpaca_emb_cos_sim": 0.9912140369415283,
      "eval_python_code_alpaca_emb_cos_sim_sem": 0.0002988268110107969,
      "eval_python_code_alpaca_emb_top1_equal": 0.6666666865348816,
      "eval_python_code_alpaca_emb_top1_equal_sem": 0.027262027544015993,
      "eval_python_code_alpaca_exact_match": 0.0,
      "eval_python_code_alpaca_exact_match_sem": 0.0,
      "eval_python_code_alpaca_loss": 2.796138286590576,
      "eval_python_code_alpaca_n_ngrams_match_1": 6.672,
      "eval_python_code_alpaca_n_ngrams_match_2": 1.834,
      "eval_python_code_alpaca_n_ngrams_match_3": 0.73,
      "eval_python_code_alpaca_num_pred_words": 21.638,
      "eval_python_code_alpaca_num_true_words": 16.86,
      "eval_python_code_alpaca_pred_num_tokens": 31.0,
      "eval_python_code_alpaca_rouge_score": 0.3596364057408109,
      "eval_python_code_alpaca_runtime": 6.6127,
      "eval_python_code_alpaca_samples_per_second": 75.613,
      "eval_python_code_alpaca_steps_per_second": 0.302,
      "eval_python_code_alpaca_token_set_f1": 0.3952441257169808,
      "eval_python_code_alpaca_token_set_f1_sem": 0.005846399731803927,
      "eval_python_code_alpaca_token_set_precision": 0.3974753269784519,
      "eval_python_code_alpaca_token_set_recall": 0.4229110024388802,
      "eval_python_code_alpaca_true_num_tokens": 21.7890625,
      "step": 87453
    },
    {
      "epoch": 4.92304661112362,
      "eval_wikibio_bleu_score": 4.427234393962183,
      "eval_wikibio_bleu_score_sem": 0.1573734298893321,
      "eval_wikibio_emb_cos_sim": 0.9924005270004272,
      "eval_wikibio_emb_cos_sim_sem": 0.00024553567147168675,
      "eval_wikibio_emb_top1_equal": 0.6399999856948853,
      "eval_wikibio_emb_top1_equal_sem": 0.027759119829079505,
      "eval_wikibio_exact_match": 0.0,
      "eval_wikibio_exact_match_sem": 0.0,
      "eval_wikibio_loss": 3.8516602516174316,
      "eval_wikibio_n_ngrams_match_1": 3.488,
      "eval_wikibio_n_ngrams_match_2": 1.028,
      "eval_wikibio_n_ngrams_match_3": 0.246,
      "eval_wikibio_num_pred_words": 19.514,
      "eval_wikibio_num_true_words": 12.252,
      "eval_wikibio_pred_num_tokens": 31.0,
      "eval_wikibio_rouge_score": 0.33671899203952016,
      "eval_wikibio_runtime": 6.5593,
      "eval_wikibio_samples_per_second": 76.227,
      "eval_wikibio_steps_per_second": 0.305,
      "eval_wikibio_token_set_f1": 0.23144633921684957,
      "eval_wikibio_token_set_f1_sem": 0.005533243007134642,
      "eval_wikibio_token_set_precision": 0.26535447138232926,
      "eval_wikibio_token_set_recall": 0.21542469020136512,
      "eval_wikibio_true_num_tokens": 31.8828125,
      "step": 87453
    },
    {
      "epoch": 4.92304661112362,
      "eval_nq_5round_bleu_score": 16.65292511550645,
      "eval_nq_5round_bleu_score_sem": 0.7335726116587187,
      "eval_nq_5round_emb_cos_sim": 0.99491286277771,
      "eval_nq_5round_emb_cos_sim_sem": 0.00031896808728176047,
      "eval_nq_5round_emb_top1_equal": 0.699999988079071,
      "eval_nq_5round_emb_top1_equal_sem": 0.026501718957810644,
      "eval_nq_5round_exact_match": 0.002,
      "eval_nq_5round_exact_match_sem": 0.002,
      "eval_nq_5round_n_ngrams_match_1": 12.06,
      "eval_nq_5round_n_ngrams_match_2": 5.254,
      "eval_nq_5round_n_ngrams_match_3": 2.818,
      "eval_nq_5round_num_pred_words": 24.024,
      "eval_nq_5round_num_true_words": 24.308,
      "eval_nq_5round_pred_num_tokens": 31.0,
      "eval_nq_5round_rouge_score": 0.4990164048403001,
      "eval_nq_5round_token_set_f1": 0.5138971253356215,
      "eval_nq_5round_token_set_f1_sem": 0.007403853991176766,
      "eval_nq_5round_token_set_precision": 0.4945622481503048,
      "eval_nq_5round_token_set_recall": 0.5432572783931494,
      "eval_nq_5round_true_num_tokens": 32.0,
      "eval_nq_bleu_score": 13.825437718474792,
      "eval_nq_bleu_score_sem": 0.6375440562728307,
      "eval_nq_emb_cos_sim": 0.9942188858985901,
      "eval_nq_emb_cos_sim_sem": 0.00031065495976376255,
      "eval_nq_emb_top1_equal": 0.6499999761581421,
      "eval_nq_emb_top1_equal_sem": 0.027583864257272155,
      "eval_nq_exact_match": 0.002,
      "eval_nq_exact_match_sem": 0.002,
      "eval_nq_loss": 1.8881789445877075,
      "eval_nq_n_ngrams_match_1": 10.952,
      "eval_nq_n_ngrams_match_2": 4.414,
      "eval_nq_n_ngrams_match_3": 2.25,
      "eval_nq_num_pred_words": 23.974,
      "eval_nq_num_true_words": 24.308,
      "eval_nq_pred_num_tokens": 31.0,
      "eval_nq_rouge_score": 0.4477455560051185,
      "eval_nq_runtime": 22.7402,
      "eval_nq_samples_per_second": 21.988,
      "eval_nq_steps_per_second": 0.088,
      "eval_nq_token_set_f1": 0.4687454006791462,
      "eval_nq_token_set_f1_sem": 0.006641530454631015,
      "eval_nq_token_set_precision": 0.45011621322166157,
      "eval_nq_token_set_recall": 0.49775017344723604,
      "eval_nq_true_num_tokens": 32.0,
      "step": 87453
    },
    {
      "epoch": 4.923553253771673,
      "grad_norm": 0.43100929260253906,
      "learning_rate": 0.001,
      "loss": 2.1513,
      "step": 87462
    },
    {
      "epoch": 4.925973879756811,
      "grad_norm": 0.4029097557067871,
      "learning_rate": 0.001,
      "loss": 2.1646,
      "step": 87505
    },
    {
      "epoch": 4.9283945057419505,
      "grad_norm": 0.39648857712745667,
      "learning_rate": 0.001,
      "loss": 2.1562,
      "step": 87548
    },
    {
      "epoch": 4.930815131727089,
      "grad_norm": 0.436490535736084,
      "learning_rate": 0.001,
      "loss": 2.1482,
      "step": 87591
    },
    {
      "epoch": 4.933235757712227,
      "grad_norm": 0.4007357060909271,
      "learning_rate": 0.001,
      "loss": 2.1561,
      "step": 87634
    },
    {
      "epoch": 4.935656383697365,
      "grad_norm": 0.40461575984954834,
      "learning_rate": 0.001,
      "loss": 2.1581,
      "step": 87677
    },
    {
      "epoch": 4.9380770096825035,
      "grad_norm": 0.4889591336250305,
      "learning_rate": 0.001,
      "loss": 2.1584,
      "step": 87720
    },
    {
      "epoch": 4.940497635667643,
      "grad_norm": 0.4698961675167084,
      "learning_rate": 0.001,
      "loss": 2.1535,
      "step": 87763
    },
    {
      "epoch": 4.942918261652781,
      "grad_norm": 0.44181737303733826,
      "learning_rate": 0.001,
      "loss": 2.1578,
      "step": 87806
    },
    {
      "epoch": 4.945338887637919,
      "grad_norm": 0.3783149719238281,
      "learning_rate": 0.001,
      "loss": 2.1539,
      "step": 87849
    },
    {
      "epoch": 4.947759513623058,
      "grad_norm": 0.42672866582870483,
      "learning_rate": 0.001,
      "loss": 2.1528,
      "step": 87892
    },
    {
      "epoch": 4.950180139608197,
      "grad_norm": 0.42486095428466797,
      "learning_rate": 0.001,
      "loss": 2.1663,
      "step": 87935
    },
    {
      "epoch": 4.952600765593335,
      "grad_norm": 0.41971880197525024,
      "learning_rate": 0.001,
      "loss": 2.163,
      "step": 87978
    },
    {
      "epoch": 4.955021391578473,
      "grad_norm": 0.4574248194694519,
      "learning_rate": 0.001,
      "loss": 2.1609,
      "step": 88021
    },
    {
      "epoch": 4.957442017563611,
      "grad_norm": 0.4163096845149994,
      "learning_rate": 0.001,
      "loss": 2.1481,
      "step": 88064
    },
    {
      "epoch": 4.9598626435487505,
      "grad_norm": 0.4051838517189026,
      "learning_rate": 0.001,
      "loss": 2.1608,
      "step": 88107
    },
    {
      "epoch": 4.962283269533889,
      "grad_norm": 0.3967275619506836,
      "learning_rate": 0.001,
      "loss": 2.1666,
      "step": 88150
    },
    {
      "epoch": 4.964703895519027,
      "grad_norm": 0.4189273715019226,
      "learning_rate": 0.001,
      "loss": 2.1611,
      "step": 88193
    },
    {
      "epoch": 4.967124521504166,
      "grad_norm": 0.5195689797401428,
      "learning_rate": 0.001,
      "loss": 2.1657,
      "step": 88236
    },
    {
      "epoch": 4.969545147489304,
      "grad_norm": 0.3926062285900116,
      "learning_rate": 0.001,
      "loss": 2.1656,
      "step": 88279
    },
    {
      "epoch": 4.971965773474443,
      "grad_norm": 0.5204746723175049,
      "learning_rate": 0.001,
      "loss": 2.1597,
      "step": 88322
    },
    {
      "epoch": 4.974386399459581,
      "grad_norm": 0.3778771758079529,
      "learning_rate": 0.001,
      "loss": 2.1562,
      "step": 88365
    },
    {
      "epoch": 4.976807025444719,
      "grad_norm": 0.3654591143131256,
      "learning_rate": 0.001,
      "loss": 2.1537,
      "step": 88408
    },
    {
      "epoch": 4.979227651429858,
      "grad_norm": 0.3602307438850403,
      "learning_rate": 0.001,
      "loss": 2.1495,
      "step": 88451
    },
    {
      "epoch": 4.981648277414997,
      "grad_norm": 0.37964996695518494,
      "learning_rate": 0.001,
      "loss": 2.164,
      "step": 88494
    },
    {
      "epoch": 4.984068903400135,
      "grad_norm": 0.4003719091415405,
      "learning_rate": 0.001,
      "loss": 2.1637,
      "step": 88537
    },
    {
      "epoch": 4.986489529385274,
      "grad_norm": 0.4669276773929596,
      "learning_rate": 0.001,
      "loss": 2.1596,
      "step": 88580
    },
    {
      "epoch": 4.988910155370412,
      "grad_norm": 0.46892163157463074,
      "learning_rate": 0.001,
      "loss": 2.1497,
      "step": 88623
    },
    {
      "epoch": 4.9913307813555505,
      "grad_norm": 0.43049657344818115,
      "learning_rate": 0.001,
      "loss": 2.1507,
      "step": 88666
    },
    {
      "epoch": 4.993751407340689,
      "grad_norm": 0.4570152461528778,
      "learning_rate": 0.001,
      "loss": 2.1554,
      "step": 88709
    },
    {
      "epoch": 4.996172033325827,
      "grad_norm": 0.4078734219074249,
      "learning_rate": 0.001,
      "loss": 2.1624,
      "step": 88752
    },
    {
      "epoch": 4.998592659310966,
      "grad_norm": 0.4075621962547302,
      "learning_rate": 0.001,
      "loss": 2.1422,
      "step": 88795
    },
    {
      "epoch": 5.0,
      "step": 88820,
      "total_flos": 1.168198305338327e+18,
      "train_loss": 2.4634555480792106,
      "train_runtime": 145129.5789,
      "train_samples_per_second": 183.594,
      "train_steps_per_second": 0.612
    }
  ],
  "logging_steps": 43,
  "max_steps": 88820,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 213,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.168198305338327e+18,
  "train_batch_size": 300,
  "trial_name": null,
  "trial_params": null
}