| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1533.9529579439338, | |
| "learning_rate": 1.4285714285714284e-09, | |
| "logits/chosen": -4.185730934143066, | |
| "logits/rejected": -4.509836196899414, | |
| "logps/chosen": -274.000732421875, | |
| "logps/rejected": -205.8054962158203, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1798.989663851012, | |
| "learning_rate": 1.4285714285714284e-08, | |
| "logits/chosen": -4.211680889129639, | |
| "logits/rejected": -4.4850640296936035, | |
| "logps/chosen": -318.5819091796875, | |
| "logps/rejected": -257.2215270996094, | |
| "loss": 0.7173, | |
| "rewards/accuracies": 0.3888888955116272, | |
| "rewards/chosen": -0.03539733216166496, | |
| "rewards/margins": -0.0653969869017601, | |
| "rewards/rejected": 0.02999965287744999, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1668.4002920396833, | |
| "learning_rate": 2.857142857142857e-08, | |
| "logits/chosen": -4.237612724304199, | |
| "logits/rejected": -4.492175102233887, | |
| "logps/chosen": -305.21356201171875, | |
| "logps/rejected": -244.23391723632812, | |
| "loss": 0.7172, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.011527794413268566, | |
| "rewards/margins": 0.015434956178069115, | |
| "rewards/rejected": -0.003907163627445698, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1590.2675307754077, | |
| "learning_rate": 4.285714285714285e-08, | |
| "logits/chosen": -4.289905548095703, | |
| "logits/rejected": -4.550080299377441, | |
| "logps/chosen": -306.20733642578125, | |
| "logps/rejected": -259.1973571777344, | |
| "loss": 0.6742, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.15369151532649994, | |
| "rewards/margins": 0.08493399620056152, | |
| "rewards/rejected": 0.06875751912593842, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1234.2347053418307, | |
| "learning_rate": 4.996892303047306e-08, | |
| "logits/chosen": -4.294736862182617, | |
| "logits/rejected": -4.560345649719238, | |
| "logps/chosen": -294.7554931640625, | |
| "logps/rejected": -234.93032836914062, | |
| "loss": 0.5901, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.3894590735435486, | |
| "rewards/margins": 0.2519429624080658, | |
| "rewards/rejected": 0.1375161111354828, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1024.357051064495, | |
| "learning_rate": 4.972077065562821e-08, | |
| "logits/chosen": -4.300290584564209, | |
| "logits/rejected": -4.545838832855225, | |
| "logps/chosen": -300.6907043457031, | |
| "logps/rejected": -252.9388427734375, | |
| "loss": 0.4845, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 1.0174586772918701, | |
| "rewards/margins": 0.6066709160804749, | |
| "rewards/rejected": 0.41078776121139526, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 993.3434930666515, | |
| "learning_rate": 4.922693215572695e-08, | |
| "logits/chosen": -4.309796333312988, | |
| "logits/rejected": -4.538843154907227, | |
| "logps/chosen": -308.38818359375, | |
| "logps/rejected": -261.2379150390625, | |
| "loss": 0.4421, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": 1.4816973209381104, | |
| "rewards/margins": 0.7946016192436218, | |
| "rewards/rejected": 0.6870955228805542, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1067.0305947544678, | |
| "learning_rate": 4.849231551964771e-08, | |
| "logits/chosen": -4.422641277313232, | |
| "logits/rejected": -4.553921699523926, | |
| "logps/chosen": -285.1670227050781, | |
| "logps/rejected": -236.08352661132812, | |
| "loss": 0.4036, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 1.8685623407363892, | |
| "rewards/margins": 1.2992055416107178, | |
| "rewards/rejected": 0.5693566203117371, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1032.775358630634, | |
| "learning_rate": 4.7524221697560474e-08, | |
| "logits/chosen": -4.270499229431152, | |
| "logits/rejected": -4.5226149559021, | |
| "logps/chosen": -291.38165283203125, | |
| "logps/rejected": -239.2751922607422, | |
| "loss": 0.379, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 2.1530368328094482, | |
| "rewards/margins": 1.5649009943008423, | |
| "rewards/rejected": 0.5881360769271851, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 872.5516189664596, | |
| "learning_rate": 4.633227204080389e-08, | |
| "logits/chosen": -4.196683406829834, | |
| "logits/rejected": -4.429889678955078, | |
| "logps/chosen": -302.6337890625, | |
| "logps/rejected": -257.6473388671875, | |
| "loss": 0.3672, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 2.4647421836853027, | |
| "rewards/margins": 1.6267496347427368, | |
| "rewards/rejected": 0.8379926681518555, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1367.4563539236096, | |
| "learning_rate": 4.4928312680573064e-08, | |
| "logits/chosen": -4.303212642669678, | |
| "logits/rejected": -4.523682594299316, | |
| "logps/chosen": -290.59271240234375, | |
| "logps/rejected": -240.2918701171875, | |
| "loss": 0.3536, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 2.6176071166992188, | |
| "rewards/margins": 1.6629337072372437, | |
| "rewards/rejected": 0.9546731114387512, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_logits/chosen": -3.2369751930236816, | |
| "eval_logits/rejected": -3.2369751930236816, | |
| "eval_logps/chosen": -158.35089111328125, | |
| "eval_logps/rejected": -158.35089111328125, | |
| "eval_loss": 0.6931471824645996, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": -1.3869400024414062, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": -1.3869400024414062, | |
| "eval_runtime": 1.4807, | |
| "eval_samples_per_second": 0.675, | |
| "eval_steps_per_second": 0.675, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 934.0407259676991, | |
| "learning_rate": 4.3326296795745654e-08, | |
| "logits/chosen": -4.324947357177734, | |
| "logits/rejected": -4.566166400909424, | |
| "logps/chosen": -297.800537109375, | |
| "logps/rejected": -249.71450805664062, | |
| "loss": 0.3471, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": 2.8152554035186768, | |
| "rewards/margins": 2.0201711654663086, | |
| "rewards/rejected": 0.7950841784477234, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1096.418461846408, | |
| "learning_rate": 4.1542145939921484e-08, | |
| "logits/chosen": -4.368184566497803, | |
| "logits/rejected": -4.613701820373535, | |
| "logps/chosen": -287.2909851074219, | |
| "logps/rejected": -237.5634002685547, | |
| "loss": 0.3344, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 2.821599245071411, | |
| "rewards/margins": 2.0445477962493896, | |
| "rewards/rejected": 0.7770514488220215, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 722.1338343165834, | |
| "learning_rate": 3.959359180586975e-08, | |
| "logits/chosen": -4.369457244873047, | |
| "logits/rejected": -4.6166582107543945, | |
| "logps/chosen": -277.1622314453125, | |
| "logps/rejected": -218.28323364257812, | |
| "loss": 0.3351, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 2.7689156532287598, | |
| "rewards/margins": 1.9403083324432373, | |
| "rewards/rejected": 0.8286076784133911, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 854.554160074848, | |
| "learning_rate": 3.75e-08, | |
| "logits/chosen": -4.2873382568359375, | |
| "logits/rejected": -4.604461669921875, | |
| "logps/chosen": -294.3111267089844, | |
| "logps/rejected": -239.65628051757812, | |
| "loss": 0.3544, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 2.786259174346924, | |
| "rewards/margins": 2.21262526512146, | |
| "rewards/rejected": 0.5736337900161743, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1024.227354138342, | |
| "learning_rate": 3.5282177578265295e-08, | |
| "logits/chosen": -4.251392364501953, | |
| "logits/rejected": -4.515843391418457, | |
| "logps/chosen": -293.588623046875, | |
| "logps/rejected": -250.3025360107422, | |
| "loss": 0.3157, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 2.910778045654297, | |
| "rewards/margins": 2.2114222049713135, | |
| "rewards/rejected": 0.6993557810783386, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 958.7773617526785, | |
| "learning_rate": 3.296216625629211e-08, | |
| "logits/chosen": -4.2807087898254395, | |
| "logits/rejected": -4.571825981140137, | |
| "logps/chosen": -294.7529296875, | |
| "logps/rejected": -235.1978759765625, | |
| "loss": 0.3382, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 3.0874907970428467, | |
| "rewards/margins": 2.296449899673462, | |
| "rewards/rejected": 0.7910411953926086, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 844.9738579712869, | |
| "learning_rate": 3.056302334890786e-08, | |
| "logits/chosen": -4.230380058288574, | |
| "logits/rejected": -4.466236114501953, | |
| "logps/chosen": -294.3287658691406, | |
| "logps/rejected": -241.07357788085938, | |
| "loss": 0.3284, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 3.165980815887451, | |
| "rewards/margins": 2.4775989055633545, | |
| "rewards/rejected": 0.6883817911148071, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 878.602697816289, | |
| "learning_rate": 2.8108592616187133e-08, | |
| "logits/chosen": -4.364335060119629, | |
| "logits/rejected": -4.596997261047363, | |
| "logps/chosen": -283.4359130859375, | |
| "logps/rejected": -240.91796875, | |
| "loss": 0.3163, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 3.220620632171631, | |
| "rewards/margins": 2.4390523433685303, | |
| "rewards/rejected": 0.7815683484077454, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 831.8981359537083, | |
| "learning_rate": 2.562326729345182e-08, | |
| "logits/chosen": -4.325879096984863, | |
| "logits/rejected": -4.494770526885986, | |
| "logps/chosen": -306.01171875, | |
| "logps/rejected": -257.24298095703125, | |
| "loss": 0.2864, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 3.1413779258728027, | |
| "rewards/margins": 2.4198684692382812, | |
| "rewards/rejected": 0.7215089797973633, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 752.1347907356251, | |
| "learning_rate": 2.3131747660339392e-08, | |
| "logits/chosen": -4.358603000640869, | |
| "logits/rejected": -4.599839687347412, | |
| "logps/chosen": -290.7633972167969, | |
| "logps/rejected": -249.0517120361328, | |
| "loss": 0.3091, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 3.1841001510620117, | |
| "rewards/margins": 2.3024184703826904, | |
| "rewards/rejected": 0.8816817998886108, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_logits/chosen": -3.230353832244873, | |
| "eval_logits/rejected": -3.230353832244873, | |
| "eval_logps/chosen": -159.3397674560547, | |
| "eval_logps/rejected": -159.3397674560547, | |
| "eval_loss": 0.6931471824645996, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": -1.8813800811767578, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": -1.8813800811767578, | |
| "eval_runtime": 1.4678, | |
| "eval_samples_per_second": 0.681, | |
| "eval_steps_per_second": 0.681, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 873.0977005391466, | |
| "learning_rate": 2.065879555832674e-08, | |
| "logits/chosen": -4.293581962585449, | |
| "logits/rejected": -4.587343692779541, | |
| "logps/chosen": -304.852294921875, | |
| "logps/rejected": -254.5345916748047, | |
| "loss": 0.3125, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 3.1841378211975098, | |
| "rewards/margins": 2.6728169918060303, | |
| "rewards/rejected": 0.5113206505775452, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 896.6576085052702, | |
| "learning_rate": 1.8228988296424874e-08, | |
| "logits/chosen": -4.245147705078125, | |
| "logits/rejected": -4.591282367706299, | |
| "logps/chosen": -295.0483703613281, | |
| "logps/rejected": -224.8583984375, | |
| "loss": 0.3117, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 3.232748031616211, | |
| "rewards/margins": 2.8655002117156982, | |
| "rewards/rejected": 0.367247611284256, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 896.8048647139675, | |
| "learning_rate": 1.5866474390840123e-08, | |
| "logits/chosen": -4.371578216552734, | |
| "logits/rejected": -4.561336994171143, | |
| "logps/chosen": -279.5164489746094, | |
| "logps/rejected": -225.96630859375, | |
| "loss": 0.2801, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 2.720884323120117, | |
| "rewards/margins": 2.2243082523345947, | |
| "rewards/rejected": 0.4965757727622986, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 794.1572537064102, | |
| "learning_rate": 1.3594733566170923e-08, | |
| "logits/chosen": -4.393925666809082, | |
| "logits/rejected": -4.711283206939697, | |
| "logps/chosen": -256.6236877441406, | |
| "logps/rejected": -215.0515594482422, | |
| "loss": 0.3001, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 2.945354700088501, | |
| "rewards/margins": 2.4272878170013428, | |
| "rewards/rejected": 0.5180668830871582, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1028.7241893624246, | |
| "learning_rate": 1.1436343403356017e-08, | |
| "logits/chosen": -4.338345527648926, | |
| "logits/rejected": -4.6125383377075195, | |
| "logps/chosen": -276.8526611328125, | |
| "logps/rejected": -232.35494995117188, | |
| "loss": 0.296, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 2.86142897605896, | |
| "rewards/margins": 2.335439443588257, | |
| "rewards/rejected": 0.5259896516799927, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 955.6521893892625, | |
| "learning_rate": 9.412754953531663e-09, | |
| "logits/chosen": -4.384321212768555, | |
| "logits/rejected": -4.630730628967285, | |
| "logps/chosen": -269.96063232421875, | |
| "logps/rejected": -240.08322143554688, | |
| "loss": 0.2927, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 2.748866558074951, | |
| "rewards/margins": 2.221247434616089, | |
| "rewards/rejected": 0.5276187658309937, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 861.1823956058561, | |
| "learning_rate": 7.544079547848182e-09, | |
| "logits/chosen": -4.517904281616211, | |
| "logits/rejected": -4.6855998039245605, | |
| "logps/chosen": -271.70074462890625, | |
| "logps/rejected": -227.885009765625, | |
| "loss": 0.3225, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 2.6885054111480713, | |
| "rewards/margins": 2.017667293548584, | |
| "rewards/rejected": 0.670837938785553, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 732.1581791035536, | |
| "learning_rate": 5.8488889220255525e-09, | |
| "logits/chosen": -4.298913955688477, | |
| "logits/rejected": -4.587487697601318, | |
| "logps/chosen": -272.83935546875, | |
| "logps/rejected": -229.5067596435547, | |
| "loss": 0.2873, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 3.1228713989257812, | |
| "rewards/margins": 2.495274066925049, | |
| "rewards/rejected": 0.6275972723960876, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1125.5972329270946, | |
| "learning_rate": 4.344030642100133e-09, | |
| "logits/chosen": -4.320634365081787, | |
| "logits/rejected": -4.5083842277526855, | |
| "logps/chosen": -280.2475280761719, | |
| "logps/rejected": -236.018798828125, | |
| "loss": 0.3136, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 3.1591415405273438, | |
| "rewards/margins": 2.4951727390289307, | |
| "rewards/rejected": 0.6639689207077026, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 788.6539907080908, | |
| "learning_rate": 3.0444606657442835e-09, | |
| "logits/chosen": -4.298077583312988, | |
| "logits/rejected": -4.541121959686279, | |
| "logps/chosen": -296.12176513671875, | |
| "logps/rejected": -239.7339324951172, | |
| "loss": 0.3363, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 3.3548855781555176, | |
| "rewards/margins": 2.5390803813934326, | |
| "rewards/rejected": 0.8158050775527954, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_logits/chosen": -3.2283477783203125, | |
| "eval_logits/rejected": -3.2283477783203125, | |
| "eval_logps/chosen": -159.417236328125, | |
| "eval_logps/rejected": -159.417236328125, | |
| "eval_loss": 0.6931471824645996, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": -1.9201126098632812, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": -1.9201126098632812, | |
| "eval_runtime": 1.4723, | |
| "eval_samples_per_second": 0.679, | |
| "eval_steps_per_second": 0.679, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1232.483796740804, | |
| "learning_rate": 1.9630947032398067e-09, | |
| "logits/chosen": -4.439385414123535, | |
| "logits/rejected": -4.579739093780518, | |
| "logps/chosen": -260.4383850097656, | |
| "logps/rejected": -226.1852264404297, | |
| "loss": 0.2944, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 2.838869571685791, | |
| "rewards/margins": 2.284980297088623, | |
| "rewards/rejected": 0.5538893342018127, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 734.9918892690486, | |
| "learning_rate": 1.1106798553464803e-09, | |
| "logits/chosen": -4.274772644042969, | |
| "logits/rejected": -4.528196334838867, | |
| "logps/chosen": -290.4882507324219, | |
| "logps/rejected": -243.47128295898438, | |
| "loss": 0.3081, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.197766065597534, | |
| "rewards/margins": 2.4614479541778564, | |
| "rewards/rejected": 0.7363181114196777, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1013.5360319244875, | |
| "learning_rate": 4.956878037864043e-10, | |
| "logits/chosen": -4.329029083251953, | |
| "logits/rejected": -4.53527307510376, | |
| "logps/chosen": -303.6526794433594, | |
| "logps/rejected": -251.791259765625, | |
| "loss": 0.3026, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 2.9665093421936035, | |
| "rewards/margins": 2.3622777462005615, | |
| "rewards/rejected": 0.6042317152023315, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1029.0631247014248, | |
| "learning_rate": 1.2423061586496474e-10, | |
| "logits/chosen": -4.317473888397217, | |
| "logits/rejected": -4.609925270080566, | |
| "logps/chosen": -283.091552734375, | |
| "logps/rejected": -227.1355438232422, | |
| "loss": 0.3064, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 2.9268617630004883, | |
| "rewards/margins": 2.242302417755127, | |
| "rewards/rejected": 0.6845596432685852, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 657.4804286362323, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -4.288041114807129, | |
| "logits/rejected": -4.490693092346191, | |
| "logps/chosen": -289.8371887207031, | |
| "logps/rejected": -243.97164916992188, | |
| "loss": 0.2999, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 3.091984987258911, | |
| "rewards/margins": 2.4618594646453857, | |
| "rewards/rejected": 0.6301255226135254, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 350, | |
| "total_flos": 0.0, | |
| "train_loss": 0.37015721593584333, | |
| "train_runtime": 5302.4484, | |
| "train_samples_per_second": 8.44, | |
| "train_steps_per_second": 0.066 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 350, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |