| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 45.0, | |
| "eval_steps": 500, | |
| "global_step": 11250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.48472875356674194, | |
| "learning_rate": 0.0003, | |
| "loss": 2.3219, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.4583953619003296, | |
| "learning_rate": 0.0003, | |
| "loss": 2.2487, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5168253968253969, | |
| "eval_loss": 2.311781167984009, | |
| "eval_runtime": 5.4425, | |
| "eval_samples_per_second": 91.87, | |
| "eval_steps_per_second": 11.576, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 18.6, | |
| "eval_f1": 26.999121042356347, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.6610094904899597, | |
| "learning_rate": 0.0003, | |
| "loss": 2.083, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.7361629605293274, | |
| "learning_rate": 0.0003, | |
| "loss": 1.9274, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6793610453605652, | |
| "learning_rate": 0.0003, | |
| "loss": 1.9113, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5153650793650794, | |
| "eval_loss": 2.3749232292175293, | |
| "eval_runtime": 5.2167, | |
| "eval_samples_per_second": 95.846, | |
| "eval_steps_per_second": 12.077, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 18.2, | |
| "eval_f1": 26.209704184704197, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.8843489289283752, | |
| "learning_rate": 0.0003, | |
| "loss": 1.4602, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.8380613923072815, | |
| "learning_rate": 0.0003, | |
| "loss": 1.4924, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5098730158730159, | |
| "eval_loss": 2.5289740562438965, | |
| "eval_runtime": 5.3631, | |
| "eval_samples_per_second": 93.23, | |
| "eval_steps_per_second": 11.747, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 16.6, | |
| "eval_f1": 24.526153291153296, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.068247675895691, | |
| "learning_rate": 0.0003, | |
| "loss": 1.2869, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 1.041741967201233, | |
| "learning_rate": 0.0003, | |
| "loss": 1.0749, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.267968773841858, | |
| "learning_rate": 0.0003, | |
| "loss": 1.1157, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5038412698412699, | |
| "eval_loss": 2.794942855834961, | |
| "eval_runtime": 5.0088, | |
| "eval_samples_per_second": 99.825, | |
| "eval_steps_per_second": 12.578, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 16.4, | |
| "eval_f1": 23.978561993561993, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.1971005201339722, | |
| "learning_rate": 0.0003, | |
| "loss": 0.7367, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 1.2383127212524414, | |
| "learning_rate": 0.0003, | |
| "loss": 0.7814, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.49853968253968256, | |
| "eval_loss": 3.0137181282043457, | |
| "eval_runtime": 5.1368, | |
| "eval_samples_per_second": 97.337, | |
| "eval_steps_per_second": 12.264, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 14.8, | |
| "eval_f1": 21.880294705294705, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.9558005928993225, | |
| "learning_rate": 0.0003, | |
| "loss": 0.6786, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 1.1341361999511719, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5741, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.1375033855438232, | |
| "learning_rate": 0.0003, | |
| "loss": 0.605, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.49657142857142855, | |
| "eval_loss": 3.220507860183716, | |
| "eval_runtime": 5.2708, | |
| "eval_samples_per_second": 94.863, | |
| "eval_steps_per_second": 11.953, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 15.4, | |
| "eval_f1": 22.281807987334307, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 1.118528962135315, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4684, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.9311291575431824, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4968, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.4946984126984127, | |
| "eval_loss": 3.3760199546813965, | |
| "eval_runtime": 5.0734, | |
| "eval_samples_per_second": 98.554, | |
| "eval_steps_per_second": 12.418, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 14.8, | |
| "eval_f1": 21.79480811586075, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 1.0299904346466064, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4631, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.7518107295036316, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4411, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.1140515804290771, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4597, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.49234920634920637, | |
| "eval_loss": 3.4634218215942383, | |
| "eval_runtime": 5.8652, | |
| "eval_samples_per_second": 85.249, | |
| "eval_steps_per_second": 10.741, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 21.617115196631318, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 1.2363754510879517, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4026, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.8174892067909241, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4242, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.4955873015873016, | |
| "eval_loss": 3.4379212856292725, | |
| "eval_runtime": 5.214, | |
| "eval_samples_per_second": 95.896, | |
| "eval_steps_per_second": 12.083, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 22.36426184926185, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.551261842250824, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4135, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.6361530423164368, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3999, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.9359388947486877, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4223, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.494952380952381, | |
| "eval_loss": 3.600064277648926, | |
| "eval_runtime": 5.8832, | |
| "eval_samples_per_second": 84.988, | |
| "eval_steps_per_second": 10.709, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 14.8, | |
| "eval_f1": 21.784345099345103, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.7834428548812866, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3855, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 1.0201836824417114, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.4955555555555556, | |
| "eval_loss": 3.571756362915039, | |
| "eval_runtime": 5.6646, | |
| "eval_samples_per_second": 88.267, | |
| "eval_steps_per_second": 11.122, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 22.02530358530358, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.738467812538147, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3977, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.535826563835144, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3904, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.8586340546607971, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4007, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.49323809523809525, | |
| "eval_loss": 3.568419933319092, | |
| "eval_runtime": 5.5173, | |
| "eval_samples_per_second": 90.624, | |
| "eval_steps_per_second": 11.419, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_exact_match": 15.2, | |
| "eval_f1": 21.94431457431457, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.6024760007858276, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3721, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.7764168381690979, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3929, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.49314285714285716, | |
| "eval_loss": 3.602919816970825, | |
| "eval_runtime": 5.1504, | |
| "eval_samples_per_second": 97.08, | |
| "eval_steps_per_second": 12.232, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_exact_match": 13.8, | |
| "eval_f1": 20.855486735486743, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "grad_norm": 0.668953537940979, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3886, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "grad_norm": 0.7968106865882874, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3845, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.7598584294319153, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4003, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.49206349206349204, | |
| "eval_loss": 3.584076404571533, | |
| "eval_runtime": 5.5138, | |
| "eval_samples_per_second": 90.682, | |
| "eval_steps_per_second": 11.426, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_exact_match": 15.6, | |
| "eval_f1": 22.729776889776886, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 0.7232977151870728, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3713, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "grad_norm": 0.6823480725288391, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3834, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.4925079365079365, | |
| "eval_loss": 3.6552798748016357, | |
| "eval_runtime": 5.0332, | |
| "eval_samples_per_second": 99.341, | |
| "eval_steps_per_second": 12.517, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_exact_match": 13.8, | |
| "eval_f1": 20.994360084360082, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "grad_norm": 0.5730023384094238, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3795, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "grad_norm": 0.943156898021698, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3786, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.8560875654220581, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3955, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.49196825396825394, | |
| "eval_loss": 3.6384711265563965, | |
| "eval_runtime": 5.73, | |
| "eval_samples_per_second": 87.26, | |
| "eval_steps_per_second": 10.995, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 21.078827926459503, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "grad_norm": 0.5993083715438843, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3651, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "grad_norm": 0.8744499683380127, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3843, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.49082539682539683, | |
| "eval_loss": 3.658356189727783, | |
| "eval_runtime": 5.2853, | |
| "eval_samples_per_second": 94.602, | |
| "eval_steps_per_second": 11.92, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 20.76140470640471, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "grad_norm": 0.7331182360649109, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3791, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "grad_norm": 0.9493301510810852, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3737, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.7808287143707275, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3916, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.49247619047619046, | |
| "eval_loss": 3.659234046936035, | |
| "eval_runtime": 5.7356, | |
| "eval_samples_per_second": 87.176, | |
| "eval_steps_per_second": 10.984, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_exact_match": 14.8, | |
| "eval_f1": 21.322143412143415, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "grad_norm": 0.7472721934318542, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3692, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "grad_norm": 0.5398291349411011, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3825, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.4913015873015873, | |
| "eval_loss": 3.660428762435913, | |
| "eval_runtime": 5.6417, | |
| "eval_samples_per_second": 88.626, | |
| "eval_steps_per_second": 11.167, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_exact_match": 15.8, | |
| "eval_f1": 23.026702741702746, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "grad_norm": 0.5717331171035767, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3746, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "grad_norm": 0.8405028581619263, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3756, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.7853202223777771, | |
| "learning_rate": 0.0003, | |
| "loss": 0.387, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.49044444444444446, | |
| "eval_loss": 3.7399282455444336, | |
| "eval_runtime": 5.2464, | |
| "eval_samples_per_second": 95.303, | |
| "eval_steps_per_second": 12.008, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_exact_match": 13.8, | |
| "eval_f1": 21.514708624708625, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.4, | |
| "grad_norm": 0.8046431541442871, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3643, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "grad_norm": 0.6186045408248901, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3832, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.4913968253968254, | |
| "eval_loss": 3.6563804149627686, | |
| "eval_runtime": 5.7336, | |
| "eval_samples_per_second": 87.206, | |
| "eval_steps_per_second": 10.988, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_exact_match": 13.8, | |
| "eval_f1": 20.836753246753247, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 21.2, | |
| "grad_norm": 0.4458799660205841, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3733, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 21.6, | |
| "grad_norm": 0.6755269765853882, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3675, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.8081554174423218, | |
| "learning_rate": 0.0003, | |
| "loss": 0.384, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.4899047619047619, | |
| "eval_loss": 3.686178207397461, | |
| "eval_runtime": 5.0623, | |
| "eval_samples_per_second": 98.77, | |
| "eval_steps_per_second": 12.445, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_exact_match": 15.6, | |
| "eval_f1": 22.18729215229216, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "grad_norm": 0.5928174257278442, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3561, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 22.8, | |
| "grad_norm": 0.5147334933280945, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3753, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.49063492063492065, | |
| "eval_loss": 3.6690633296966553, | |
| "eval_runtime": 5.6639, | |
| "eval_samples_per_second": 88.278, | |
| "eval_steps_per_second": 11.123, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_exact_match": 14.2, | |
| "eval_f1": 21.33359729642339, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 23.2, | |
| "grad_norm": 0.8577277660369873, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3655, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 23.6, | |
| "grad_norm": 0.6065458655357361, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3609, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 1.1300039291381836, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3816, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.4908571428571429, | |
| "eval_loss": 3.7181262969970703, | |
| "eval_runtime": 5.7672, | |
| "eval_samples_per_second": 86.698, | |
| "eval_steps_per_second": 10.924, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_exact_match": 13.6, | |
| "eval_f1": 20.043032849503437, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 24.4, | |
| "grad_norm": 0.722024142742157, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3555, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 24.8, | |
| "grad_norm": 0.773781955242157, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3711, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.4895555555555556, | |
| "eval_loss": 3.715923309326172, | |
| "eval_runtime": 5.35, | |
| "eval_samples_per_second": 93.457, | |
| "eval_steps_per_second": 11.776, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_exact_match": 14.2, | |
| "eval_f1": 21.817198315959928, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 25.2, | |
| "grad_norm": 0.5301027894020081, | |
| "learning_rate": 0.0003, | |
| "loss": 0.365, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 25.6, | |
| "grad_norm": 0.6550167202949524, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3632, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.9597300291061401, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3758, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.49073015873015874, | |
| "eval_loss": 3.7265737056732178, | |
| "eval_runtime": 5.7349, | |
| "eval_samples_per_second": 87.185, | |
| "eval_steps_per_second": 10.985, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_exact_match": 13.0, | |
| "eval_f1": 19.8548159030512, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 26.4, | |
| "grad_norm": 0.5059230327606201, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3536, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 26.8, | |
| "grad_norm": 0.5931850671768188, | |
| "learning_rate": 0.0003, | |
| "loss": 0.372, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.4915873015873016, | |
| "eval_loss": 3.7632603645324707, | |
| "eval_runtime": 4.9892, | |
| "eval_samples_per_second": 100.216, | |
| "eval_steps_per_second": 12.627, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_exact_match": 14.6, | |
| "eval_f1": 21.288471691706988, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 27.2, | |
| "grad_norm": 0.8211339712142944, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3667, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 27.6, | |
| "grad_norm": 0.7241945862770081, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3655, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.8376013040542603, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3813, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.48926984126984124, | |
| "eval_loss": 3.751114845275879, | |
| "eval_runtime": 5.7876, | |
| "eval_samples_per_second": 86.391, | |
| "eval_steps_per_second": 10.885, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 21.447649735885033, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 28.4, | |
| "grad_norm": 1.0523216724395752, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3588, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 28.8, | |
| "grad_norm": 1.0963407754898071, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3699, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.4903174603174603, | |
| "eval_loss": 3.7290866374969482, | |
| "eval_runtime": 5.3378, | |
| "eval_samples_per_second": 93.671, | |
| "eval_steps_per_second": 11.803, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 21.026401539636833, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 29.2, | |
| "grad_norm": 0.6481629610061646, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3655, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 29.6, | |
| "grad_norm": 0.6700736284255981, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3657, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.6046570539474487, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3772, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.48857142857142855, | |
| "eval_loss": 3.7827305793762207, | |
| "eval_runtime": 4.9957, | |
| "eval_samples_per_second": 100.086, | |
| "eval_steps_per_second": 12.611, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_exact_match": 14.2, | |
| "eval_f1": 20.93110366757426, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 30.4, | |
| "grad_norm": 0.5793018937110901, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3515, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 30.8, | |
| "grad_norm": 0.733734667301178, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3673, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.48917460317460315, | |
| "eval_loss": 3.8032257556915283, | |
| "eval_runtime": 5.7717, | |
| "eval_samples_per_second": 86.629, | |
| "eval_steps_per_second": 10.915, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_exact_match": 13.2, | |
| "eval_f1": 19.896597454832754, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 31.2, | |
| "grad_norm": 0.8759809732437134, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3642, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 31.6, | |
| "grad_norm": 0.7398084998130798, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3647, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.9378136992454529, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3708, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.4894603174603175, | |
| "eval_loss": 3.830268144607544, | |
| "eval_runtime": 5.6155, | |
| "eval_samples_per_second": 89.039, | |
| "eval_steps_per_second": 11.219, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_exact_match": 14.0, | |
| "eval_f1": 20.862043120278422, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 32.4, | |
| "grad_norm": 0.4455881118774414, | |
| "learning_rate": 0.0003, | |
| "loss": 0.346, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 32.8, | |
| "grad_norm": 0.6031106114387512, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3632, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.4886984126984127, | |
| "eval_loss": 3.821800947189331, | |
| "eval_runtime": 5.21, | |
| "eval_samples_per_second": 95.968, | |
| "eval_steps_per_second": 12.092, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_exact_match": 14.0, | |
| "eval_f1": 20.771498109733407, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 33.2, | |
| "grad_norm": 0.6439555287361145, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3553, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 33.6, | |
| "grad_norm": 0.6980018019676208, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3533, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.6426189541816711, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3692, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.49, | |
| "eval_loss": 3.7488105297088623, | |
| "eval_runtime": 4.9468, | |
| "eval_samples_per_second": 101.075, | |
| "eval_steps_per_second": 12.735, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_exact_match": 14.8, | |
| "eval_f1": 21.277384282384293, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 34.4, | |
| "grad_norm": 0.7607336044311523, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3475, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 34.8, | |
| "grad_norm": 0.590248167514801, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3678, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.4868571428571429, | |
| "eval_loss": 3.852421760559082, | |
| "eval_runtime": 5.7678, | |
| "eval_samples_per_second": 86.688, | |
| "eval_steps_per_second": 10.923, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_exact_match": 13.4, | |
| "eval_f1": 20.523333496568796, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 35.2, | |
| "grad_norm": 0.4356330931186676, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3592, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 35.6, | |
| "grad_norm": 0.8419196009635925, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3609, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 0.7648630142211914, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3762, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.4875238095238095, | |
| "eval_loss": 3.8220598697662354, | |
| "eval_runtime": 5.6351, | |
| "eval_samples_per_second": 88.729, | |
| "eval_steps_per_second": 11.18, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_exact_match": 13.4, | |
| "eval_f1": 20.17646758470288, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 36.4, | |
| "grad_norm": 0.7656174898147583, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3553, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 36.8, | |
| "grad_norm": 0.7704755067825317, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3702, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.4862222222222222, | |
| "eval_loss": 3.8083200454711914, | |
| "eval_runtime": 5.2407, | |
| "eval_samples_per_second": 95.407, | |
| "eval_steps_per_second": 12.021, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_exact_match": 13.4, | |
| "eval_f1": 20.7023073005426, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 37.2, | |
| "grad_norm": 1.0103473663330078, | |
| "learning_rate": 0.0003, | |
| "loss": 0.361, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 37.6, | |
| "grad_norm": 0.7305041551589966, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3583, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "grad_norm": 0.8439377546310425, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3745, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.48603174603174604, | |
| "eval_loss": 3.83286452293396, | |
| "eval_runtime": 5.1879, | |
| "eval_samples_per_second": 96.377, | |
| "eval_steps_per_second": 12.144, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_exact_match": 15.4, | |
| "eval_f1": 21.705345210345218, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 38.4, | |
| "grad_norm": 0.44817036390304565, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3493, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 38.8, | |
| "grad_norm": 0.8559367060661316, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3611, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.48784126984126985, | |
| "eval_loss": 3.8968539237976074, | |
| "eval_runtime": 4.9847, | |
| "eval_samples_per_second": 100.306, | |
| "eval_steps_per_second": 12.639, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 20.261010819246113, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 39.2, | |
| "grad_norm": 0.6502194404602051, | |
| "learning_rate": 0.0003, | |
| "loss": 0.358, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 39.6, | |
| "grad_norm": 1.0540180206298828, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3528, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.6658892631530762, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3648, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.4868571428571429, | |
| "eval_loss": 3.84969425201416, | |
| "eval_runtime": 5.7485, | |
| "eval_samples_per_second": 86.98, | |
| "eval_steps_per_second": 10.959, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_exact_match": 13.0, | |
| "eval_f1": 20.333614881850174, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 40.4, | |
| "grad_norm": 0.5723844766616821, | |
| "learning_rate": 0.0003, | |
| "loss": 0.347, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 40.8, | |
| "grad_norm": 0.5007913708686829, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3616, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_accuracy": 0.4865079365079365, | |
| "eval_loss": 3.8460936546325684, | |
| "eval_runtime": 5.3089, | |
| "eval_samples_per_second": 94.181, | |
| "eval_steps_per_second": 11.867, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_exact_match": 12.8, | |
| "eval_f1": 19.745857084092385, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 41.2, | |
| "grad_norm": 0.597031831741333, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3535, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 41.6, | |
| "grad_norm": 0.4940139651298523, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3511, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "grad_norm": 0.9585108757019043, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3659, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.48768253968253966, | |
| "eval_loss": 3.8721938133239746, | |
| "eval_runtime": 4.8871, | |
| "eval_samples_per_second": 102.31, | |
| "eval_steps_per_second": 12.891, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_exact_match": 13.4, | |
| "eval_f1": 20.243903482138784, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 42.4, | |
| "grad_norm": 0.5968766212463379, | |
| "learning_rate": 0.0003, | |
| "loss": 0.345, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 42.8, | |
| "grad_norm": 0.5495741367340088, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3585, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_accuracy": 0.48742857142857143, | |
| "eval_loss": 3.876279830932617, | |
| "eval_runtime": 5.7458, | |
| "eval_samples_per_second": 87.021, | |
| "eval_steps_per_second": 10.965, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_exact_match": 13.2, | |
| "eval_f1": 20.778424516659825, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 43.2, | |
| "grad_norm": 0.6515761613845825, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3534, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 43.6, | |
| "grad_norm": 0.6179555654525757, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3506, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "grad_norm": 0.5593932271003723, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3628, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.48768253968253966, | |
| "eval_loss": 3.850708484649658, | |
| "eval_runtime": 5.1171, | |
| "eval_samples_per_second": 97.711, | |
| "eval_steps_per_second": 12.312, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_exact_match": 14.2, | |
| "eval_f1": 22.1020778808395, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 44.4, | |
| "grad_norm": 0.6734414100646973, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3459, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 44.8, | |
| "grad_norm": 0.840034008026123, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3616, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_accuracy": 0.4875555555555556, | |
| "eval_loss": 3.878772020339966, | |
| "eval_runtime": 4.9819, | |
| "eval_samples_per_second": 100.364, | |
| "eval_steps_per_second": 12.646, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_exact_match": 13.6, | |
| "eval_f1": 20.89890237213767, | |
| "step": 11250 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 12500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.910461842256036e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |