| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.944444444444445, | |
| "eval_steps": 300, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11574074074074074, | |
| "grad_norm": 39.743831634521484, | |
| "learning_rate": 8.654497354497356e-06, | |
| "loss": 2.2644, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.23148148148148148, | |
| "grad_norm": 80.18631744384766, | |
| "learning_rate": 8.50899470899471e-06, | |
| "loss": 1.7544, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3472222222222222, | |
| "grad_norm": 56.661163330078125, | |
| "learning_rate": 8.363492063492064e-06, | |
| "loss": 1.8464, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.46296296296296297, | |
| "grad_norm": 33.45421600341797, | |
| "learning_rate": 8.217989417989417e-06, | |
| "loss": 1.493, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5787037037037037, | |
| "grad_norm": 88.00537872314453, | |
| "learning_rate": 8.072486772486773e-06, | |
| "loss": 1.6763, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 31.951480865478516, | |
| "learning_rate": 7.926984126984127e-06, | |
| "loss": 1.5285, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "eval_accuracy": 0.5601851940155029, | |
| "eval_loss": 1.482226848602295, | |
| "eval_runtime": 189.0589, | |
| "eval_samples_per_second": 2.285, | |
| "eval_steps_per_second": 1.143, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8101851851851852, | |
| "grad_norm": 27.660470962524414, | |
| "learning_rate": 7.781481481481482e-06, | |
| "loss": 1.5798, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 76.16017150878906, | |
| "learning_rate": 7.635978835978836e-06, | |
| "loss": 1.4862, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0416666666666667, | |
| "grad_norm": 33.353267669677734, | |
| "learning_rate": 7.490476190476191e-06, | |
| "loss": 1.332, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1574074074074074, | |
| "grad_norm": 50.27310562133789, | |
| "learning_rate": 7.3449735449735455e-06, | |
| "loss": 1.2905, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2731481481481481, | |
| "grad_norm": 29.26256561279297, | |
| "learning_rate": 7.1994708994709e-06, | |
| "loss": 1.0945, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 142.0453643798828, | |
| "learning_rate": 7.053968253968255e-06, | |
| "loss": 1.4222, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "eval_accuracy": 0.6504629850387573, | |
| "eval_loss": 1.210780382156372, | |
| "eval_runtime": 191.2465, | |
| "eval_samples_per_second": 2.259, | |
| "eval_steps_per_second": 1.129, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5046296296296298, | |
| "grad_norm": 32.452552795410156, | |
| "learning_rate": 6.908465608465609e-06, | |
| "loss": 1.0473, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.6203703703703702, | |
| "grad_norm": 9.073972702026367, | |
| "learning_rate": 6.762962962962963e-06, | |
| "loss": 1.1119, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7361111111111112, | |
| "grad_norm": 81.59954833984375, | |
| "learning_rate": 6.617460317460318e-06, | |
| "loss": 1.039, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 36.21083068847656, | |
| "learning_rate": 6.471957671957673e-06, | |
| "loss": 1.1544, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9675925925925926, | |
| "grad_norm": 20.776355743408203, | |
| "learning_rate": 6.326455026455027e-06, | |
| "loss": 1.2092, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.0833333333333335, | |
| "grad_norm": 50.247093200683594, | |
| "learning_rate": 6.18095238095238e-06, | |
| "loss": 0.8443, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.0833333333333335, | |
| "eval_accuracy": 0.6805555820465088, | |
| "eval_loss": 1.0854352712631226, | |
| "eval_runtime": 192.1877, | |
| "eval_samples_per_second": 2.248, | |
| "eval_steps_per_second": 1.124, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.199074074074074, | |
| "grad_norm": 202.2728729248047, | |
| "learning_rate": 6.035449735449736e-06, | |
| "loss": 0.9478, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.314814814814815, | |
| "grad_norm": 44.58985900878906, | |
| "learning_rate": 5.889947089947091e-06, | |
| "loss": 0.9051, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.4305555555555554, | |
| "grad_norm": 18.935836791992188, | |
| "learning_rate": 5.744444444444445e-06, | |
| "loss": 0.8553, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.5462962962962963, | |
| "grad_norm": 2.024139642715454, | |
| "learning_rate": 5.5989417989418e-06, | |
| "loss": 1.0215, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.662037037037037, | |
| "grad_norm": 49.676944732666016, | |
| "learning_rate": 5.453439153439153e-06, | |
| "loss": 0.793, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 40.503150939941406, | |
| "learning_rate": 5.307936507936508e-06, | |
| "loss": 0.8355, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "eval_accuracy": 0.7129629850387573, | |
| "eval_loss": 1.0518003702163696, | |
| "eval_runtime": 192.7676, | |
| "eval_samples_per_second": 2.241, | |
| "eval_steps_per_second": 1.121, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.8935185185185186, | |
| "grad_norm": 26.035078048706055, | |
| "learning_rate": 5.162433862433862e-06, | |
| "loss": 0.8288, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.009259259259259, | |
| "grad_norm": 168.5670928955078, | |
| "learning_rate": 5.016931216931217e-06, | |
| "loss": 0.7811, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 19.110559463500977, | |
| "learning_rate": 4.871428571428572e-06, | |
| "loss": 0.6065, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.240740740740741, | |
| "grad_norm": 4.887545585632324, | |
| "learning_rate": 4.725925925925926e-06, | |
| "loss": 0.6886, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.3564814814814814, | |
| "grad_norm": 28.545923233032227, | |
| "learning_rate": 4.58042328042328e-06, | |
| "loss": 0.6935, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.4722222222222223, | |
| "grad_norm": 41.39081954956055, | |
| "learning_rate": 4.434920634920635e-06, | |
| "loss": 0.6307, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.4722222222222223, | |
| "eval_accuracy": 0.7060185074806213, | |
| "eval_loss": 1.0891976356506348, | |
| "eval_runtime": 184.4145, | |
| "eval_samples_per_second": 2.343, | |
| "eval_steps_per_second": 1.171, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.587962962962963, | |
| "grad_norm": 15.346806526184082, | |
| "learning_rate": 4.2894179894179895e-06, | |
| "loss": 0.6561, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.7037037037037037, | |
| "grad_norm": 12.401691436767578, | |
| "learning_rate": 4.143915343915344e-06, | |
| "loss": 0.7564, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.8194444444444446, | |
| "grad_norm": 0.9189475774765015, | |
| "learning_rate": 3.998412698412699e-06, | |
| "loss": 0.7843, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.935185185185185, | |
| "grad_norm": 2.660712242126465, | |
| "learning_rate": 3.8529100529100535e-06, | |
| "loss": 0.6449, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.050925925925926, | |
| "grad_norm": 2.3331284523010254, | |
| "learning_rate": 3.7074074074074073e-06, | |
| "loss": 0.5658, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 88.60041046142578, | |
| "learning_rate": 3.561904761904762e-06, | |
| "loss": 0.5227, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "eval_accuracy": 0.7407407164573669, | |
| "eval_loss": 0.9819123148918152, | |
| "eval_runtime": 191.5065, | |
| "eval_samples_per_second": 2.256, | |
| "eval_steps_per_second": 1.128, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.282407407407407, | |
| "grad_norm": 43.31499481201172, | |
| "learning_rate": 3.4164021164021167e-06, | |
| "loss": 0.4657, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.398148148148148, | |
| "grad_norm": 4.856300354003906, | |
| "learning_rate": 3.270899470899471e-06, | |
| "loss": 0.5274, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.513888888888889, | |
| "grad_norm": 21.986839294433594, | |
| "learning_rate": 3.1253968253968256e-06, | |
| "loss": 0.3385, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.62962962962963, | |
| "grad_norm": 3.5640218257904053, | |
| "learning_rate": 2.97989417989418e-06, | |
| "loss": 0.5006, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.74537037037037, | |
| "grad_norm": 35.41746520996094, | |
| "learning_rate": 2.8343915343915345e-06, | |
| "loss": 0.3829, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.861111111111111, | |
| "grad_norm": 110.1950912475586, | |
| "learning_rate": 2.6888888888888892e-06, | |
| "loss": 0.4838, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.861111111111111, | |
| "eval_accuracy": 0.7453703880310059, | |
| "eval_loss": 1.0233818292617798, | |
| "eval_runtime": 187.5989, | |
| "eval_samples_per_second": 2.303, | |
| "eval_steps_per_second": 1.151, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.976851851851852, | |
| "grad_norm": 9.007341384887695, | |
| "learning_rate": 2.5433862433862435e-06, | |
| "loss": 0.5366, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.092592592592593, | |
| "grad_norm": 23.511497497558594, | |
| "learning_rate": 2.397883597883598e-06, | |
| "loss": 0.4084, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.208333333333333, | |
| "grad_norm": 46.781288146972656, | |
| "learning_rate": 2.2523809523809524e-06, | |
| "loss": 0.3158, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.324074074074074, | |
| "grad_norm": 11.613898277282715, | |
| "learning_rate": 2.106878306878307e-06, | |
| "loss": 0.3805, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.439814814814815, | |
| "grad_norm": 108.98674011230469, | |
| "learning_rate": 1.9613756613756613e-06, | |
| "loss": 0.3677, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 67.47931671142578, | |
| "learning_rate": 1.8158730158730158e-06, | |
| "loss": 0.4089, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "eval_accuracy": 0.7708333134651184, | |
| "eval_loss": 0.9597391486167908, | |
| "eval_runtime": 193.0387, | |
| "eval_samples_per_second": 2.238, | |
| "eval_steps_per_second": 1.119, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.671296296296296, | |
| "grad_norm": 47.788421630859375, | |
| "learning_rate": 1.6703703703703705e-06, | |
| "loss": 0.3808, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 5.787037037037037, | |
| "grad_norm": 9.64560317993164, | |
| "learning_rate": 1.524867724867725e-06, | |
| "loss": 0.2779, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.902777777777778, | |
| "grad_norm": 59.49520492553711, | |
| "learning_rate": 1.3793650793650794e-06, | |
| "loss": 0.3443, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 6.018518518518518, | |
| "grad_norm": 20.228626251220703, | |
| "learning_rate": 1.2338624338624339e-06, | |
| "loss": 0.3234, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.1342592592592595, | |
| "grad_norm": 27.677474975585938, | |
| "learning_rate": 1.0883597883597883e-06, | |
| "loss": 0.2092, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.7405154705047607, | |
| "learning_rate": 9.428571428571429e-07, | |
| "loss": 0.3254, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "eval_accuracy": 0.7708333134651184, | |
| "eval_loss": 0.9336540102958679, | |
| "eval_runtime": 193.4468, | |
| "eval_samples_per_second": 2.233, | |
| "eval_steps_per_second": 1.117, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.3657407407407405, | |
| "grad_norm": 9.925797462463379, | |
| "learning_rate": 7.973544973544975e-07, | |
| "loss": 0.2506, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 6.481481481481482, | |
| "grad_norm": 0.22488392889499664, | |
| "learning_rate": 6.518518518518518e-07, | |
| "loss": 0.2574, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.597222222222222, | |
| "grad_norm": 52.721458435058594, | |
| "learning_rate": 5.063492063492064e-07, | |
| "loss": 0.1824, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 6.712962962962963, | |
| "grad_norm": 42.07158660888672, | |
| "learning_rate": 3.6084656084656085e-07, | |
| "loss": 0.3214, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 6.828703703703704, | |
| "grad_norm": 18.518312454223633, | |
| "learning_rate": 2.1534391534391534e-07, | |
| "loss": 0.2811, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 6.944444444444445, | |
| "grad_norm": 0.2595561742782593, | |
| "learning_rate": 6.984126984126985e-08, | |
| "loss": 0.2413, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.944444444444445, | |
| "eval_accuracy": 0.7800925970077515, | |
| "eval_loss": 0.9138823747634888, | |
| "eval_runtime": 193.0312, | |
| "eval_samples_per_second": 2.238, | |
| "eval_steps_per_second": 1.119, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 3024, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 600, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.28237560576e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |