| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 27.0, | |
| "eval_steps": 500, | |
| "global_step": 14283, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1890359168241966, | |
| "grad_norm": 0.5558584332466125, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1103, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 0.4733140468597412, | |
| "learning_rate": 3e-05, | |
| "loss": 1.931, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 0.48580050468444824, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8948, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 0.40655234456062317, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8988, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.945179584120983, | |
| "grad_norm": 0.6303955316543579, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8745, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5701538461538461, | |
| "eval_loss": 1.6455246210098267, | |
| "eval_runtime": 5.0805, | |
| "eval_samples_per_second": 98.415, | |
| "eval_steps_per_second": 12.4, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 23.927619047619068, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1342155009451795, | |
| "grad_norm": 0.4278818666934967, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8798, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3232514177693762, | |
| "grad_norm": 0.449830025434494, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8613, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5122873345935728, | |
| "grad_norm": 0.5130725502967834, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8709, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7013232514177694, | |
| "grad_norm": 0.4889715611934662, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8448, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8903591682419658, | |
| "grad_norm": 11.186380386352539, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8764, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5723589743589743, | |
| "eval_loss": 1.6143561601638794, | |
| "eval_runtime": 5.0466, | |
| "eval_samples_per_second": 99.077, | |
| "eval_steps_per_second": 12.484, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 15.6, | |
| "eval_f1": 25.825238095238113, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0793950850661624, | |
| "grad_norm": 0.5414931774139404, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8398, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.268431001890359, | |
| "grad_norm": 7.441005706787109, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8348, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4574669187145557, | |
| "grad_norm": 24.312145233154297, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8421, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6465028355387523, | |
| "grad_norm": 0.7407662868499756, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8292, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.835538752362949, | |
| "grad_norm": 0.9653642177581787, | |
| "learning_rate": 3e-05, | |
| "loss": 1.828, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5734358974358974, | |
| "eval_loss": 1.617864727973938, | |
| "eval_runtime": 5.2153, | |
| "eval_samples_per_second": 95.872, | |
| "eval_steps_per_second": 12.08, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 16.2, | |
| "eval_f1": 25.811428571428586, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0245746691871456, | |
| "grad_norm": 450.6055603027344, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8204, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.213610586011342, | |
| "grad_norm": 2.1127028465270996, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8361, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.402646502835539, | |
| "grad_norm": 564.9320068359375, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8338, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.5916824196597354, | |
| "grad_norm": 52.5078010559082, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8075, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.780718336483932, | |
| "grad_norm": 27.87557601928711, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8175, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.9697542533081287, | |
| "grad_norm": 4.810656547546387, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8218, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5728205128205128, | |
| "eval_loss": 1.6224406957626343, | |
| "eval_runtime": 5.0588, | |
| "eval_samples_per_second": 98.838, | |
| "eval_steps_per_second": 12.454, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 17.4, | |
| "eval_f1": 27.194285714285726, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 4.158790170132325, | |
| "grad_norm": 13.004076957702637, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8089, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 2.23368501663208, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8292, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.536862003780718, | |
| "grad_norm": 0.8635586500167847, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7992, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.725897920604915, | |
| "grad_norm": 6.5326738357543945, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7955, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.914933837429111, | |
| "grad_norm": 463.4647521972656, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7953, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5740512820512821, | |
| "eval_loss": 1.6217337846755981, | |
| "eval_runtime": 5.1898, | |
| "eval_samples_per_second": 96.342, | |
| "eval_steps_per_second": 12.139, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 17.6, | |
| "eval_f1": 26.161428571428587, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 5.103969754253308, | |
| "grad_norm": 203.82815551757812, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7845, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.293005671077505, | |
| "grad_norm": 660.8076782226562, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7909, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.482041587901701, | |
| "grad_norm": 5727.51806640625, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7909, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.671077504725898, | |
| "grad_norm": 4880698.0, | |
| "learning_rate": 3e-05, | |
| "loss": 3.5781, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.8601134215500945, | |
| "grad_norm": 1192.6517333984375, | |
| "learning_rate": 3e-05, | |
| "loss": 3.8703, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5433846153846154, | |
| "eval_loss": 1.886108636856079, | |
| "eval_runtime": 5.5162, | |
| "eval_samples_per_second": 90.643, | |
| "eval_steps_per_second": 11.421, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 11.4, | |
| "eval_f1": 20.358571428571437, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 6.049149338374291, | |
| "grad_norm": 532.3778686523438, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1701, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.238185255198488, | |
| "grad_norm": 565.6697998046875, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8533, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.427221172022684, | |
| "grad_norm": 25206.482421875, | |
| "learning_rate": 3e-05, | |
| "loss": 2.623, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.616257088846881, | |
| "grad_norm": 71.88556671142578, | |
| "learning_rate": 3e-05, | |
| "loss": 3.1573, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.805293005671078, | |
| "grad_norm": 269563.15625, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2696, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.994328922495274, | |
| "grad_norm": 11768.7880859375, | |
| "learning_rate": 3e-05, | |
| "loss": 3.1506, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.2834871794871795, | |
| "eval_loss": 8.311681747436523, | |
| "eval_runtime": 5.3223, | |
| "eval_samples_per_second": 93.944, | |
| "eval_steps_per_second": 11.837, | |
| "step": 3703 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.027458851580988985, | |
| "step": 3703 | |
| }, | |
| { | |
| "epoch": 7.183364839319471, | |
| "grad_norm": 6220.75244140625, | |
| "learning_rate": 3e-05, | |
| "loss": 4.273, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.3724007561436675, | |
| "grad_norm": 3387.953857421875, | |
| "learning_rate": 3e-05, | |
| "loss": 7.7199, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.561436672967864, | |
| "grad_norm": 372.8681640625, | |
| "learning_rate": 3e-05, | |
| "loss": 7.7045, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.750472589792061, | |
| "grad_norm": 159.64283752441406, | |
| "learning_rate": 3e-05, | |
| "loss": 6.8901, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 7.939508506616257, | |
| "grad_norm": 751876.125, | |
| "learning_rate": 3e-05, | |
| "loss": 6.5238, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.22764102564102565, | |
| "eval_loss": 7.12347936630249, | |
| "eval_runtime": 5.3818, | |
| "eval_samples_per_second": 92.905, | |
| "eval_steps_per_second": 11.706, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.013333333333333336, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 8.128544423440454, | |
| "grad_norm": 1035664128.0, | |
| "learning_rate": 3e-05, | |
| "loss": 5.4157, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 8.31758034026465, | |
| "grad_norm": 11571242.0, | |
| "learning_rate": 3e-05, | |
| "loss": 5.8257, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.506616257088847, | |
| "grad_norm": 88236256.0, | |
| "learning_rate": 3e-05, | |
| "loss": 6.3609, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 105426080.0, | |
| "learning_rate": 3e-05, | |
| "loss": 6.1824, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.88468809073724, | |
| "grad_norm": 51438844.0, | |
| "learning_rate": 3e-05, | |
| "loss": 6.1818, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 11.187646865844727, | |
| "eval_runtime": 5.154, | |
| "eval_samples_per_second": 97.013, | |
| "eval_steps_per_second": 12.224, | |
| "step": 4761 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 4761 | |
| }, | |
| { | |
| "epoch": 9.073724007561436, | |
| "grad_norm": 8742571.0, | |
| "learning_rate": 3e-05, | |
| "loss": 7.5029, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.262759924385634, | |
| "grad_norm": 4885384192.0, | |
| "learning_rate": 3e-05, | |
| "loss": 7.805, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 9.45179584120983, | |
| "grad_norm": 919074304.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.0712, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.640831758034027, | |
| "grad_norm": 9706883072.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.4915, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 9.829867674858223, | |
| "grad_norm": 35723780096.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.3286, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 13.374640464782715, | |
| "eval_runtime": 5.2089, | |
| "eval_samples_per_second": 95.99, | |
| "eval_steps_per_second": 12.095, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 10.01890359168242, | |
| "grad_norm": 172018032.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.4403, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 10.207939508506616, | |
| "grad_norm": 329336992.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.3233, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.396975425330814, | |
| "grad_norm": 74210560.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.3673, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.58601134215501, | |
| "grad_norm": 497968480.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.8311, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 10.775047258979207, | |
| "grad_norm": 34138104.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.8161, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 10.964083175803403, | |
| "grad_norm": 308232992.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.2827, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 12.878591537475586, | |
| "eval_runtime": 5.1038, | |
| "eval_samples_per_second": 97.966, | |
| "eval_steps_per_second": 12.344, | |
| "step": 5819 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 5819 | |
| }, | |
| { | |
| "epoch": 11.1531190926276, | |
| "grad_norm": 255807824.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.8782, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 11.342155009451796, | |
| "grad_norm": 27976615936.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.879, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.531190926275993, | |
| "grad_norm": 13783190732800.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.8617, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 11.720226843100189, | |
| "grad_norm": 3467661017088.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.1028, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 11.909262759924385, | |
| "grad_norm": 333752.3125, | |
| "learning_rate": 3e-05, | |
| "loss": 9.4972, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 13.161654472351074, | |
| "eval_runtime": 5.0927, | |
| "eval_samples_per_second": 98.18, | |
| "eval_steps_per_second": 12.371, | |
| "step": 6348 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 6348 | |
| }, | |
| { | |
| "epoch": 12.098298676748582, | |
| "grad_norm": 6690696704.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.2939, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.287334593572778, | |
| "grad_norm": 148906835968.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.3499, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.476370510396976, | |
| "grad_norm": 15515338752.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.3496, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.665406427221171, | |
| "grad_norm": 76318760960.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.2728, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 12.854442344045369, | |
| "grad_norm": 85637283840.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.4453, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 12.748088836669922, | |
| "eval_runtime": 5.1332, | |
| "eval_samples_per_second": 97.405, | |
| "eval_steps_per_second": 12.273, | |
| "step": 6877 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 6877 | |
| }, | |
| { | |
| "epoch": 13.043478260869565, | |
| "grad_norm": 370487552.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.8639, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 13.232514177693762, | |
| "grad_norm": 243782.78125, | |
| "learning_rate": 3e-05, | |
| "loss": 9.2996, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.421550094517958, | |
| "grad_norm": 572714688.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.0366, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 13.610586011342155, | |
| "grad_norm": 807026112.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.9777, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 13.799621928166351, | |
| "grad_norm": 401406112.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.3094, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 13.988657844990549, | |
| "grad_norm": 20012445696.0, | |
| "learning_rate": 3e-05, | |
| "loss": 11.1388, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 11.845151901245117, | |
| "eval_runtime": 5.7887, | |
| "eval_samples_per_second": 86.376, | |
| "eval_steps_per_second": 10.883, | |
| "step": 7406 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 7406 | |
| }, | |
| { | |
| "epoch": 14.177693761814744, | |
| "grad_norm": 61066740.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.4144, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.366729678638942, | |
| "grad_norm": 58445844.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.1294, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.555765595463138, | |
| "grad_norm": 40712032.0, | |
| "learning_rate": 3e-05, | |
| "loss": 12.1512, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 14.744801512287335, | |
| "grad_norm": 118321208.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.8949, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 14.93383742911153, | |
| "grad_norm": 61603804.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.849, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 14.529706954956055, | |
| "eval_runtime": 5.7662, | |
| "eval_samples_per_second": 86.712, | |
| "eval_steps_per_second": 10.926, | |
| "step": 7935 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 7935 | |
| }, | |
| { | |
| "epoch": 15.122873345935728, | |
| "grad_norm": 1755613568.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.7302, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.311909262759924, | |
| "grad_norm": 1215193088.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.1428, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 15.500945179584122, | |
| "grad_norm": 40452497408.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.0763, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 15.689981096408317, | |
| "grad_norm": 10178840576.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.2204, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 15.879017013232515, | |
| "grad_norm": 10069305344.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.9088, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.22523076923076923, | |
| "eval_loss": 14.802911758422852, | |
| "eval_runtime": 5.0764, | |
| "eval_samples_per_second": 98.494, | |
| "eval_steps_per_second": 12.41, | |
| "step": 8464 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 8464 | |
| }, | |
| { | |
| "epoch": 16.068052930056712, | |
| "grad_norm": 9244783616.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.2574, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.257088846880908, | |
| "grad_norm": 556177664.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.6479, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.446124763705104, | |
| "grad_norm": 44021864.0, | |
| "learning_rate": 3e-05, | |
| "loss": 11.8164, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 16.6351606805293, | |
| "grad_norm": 17835092.0, | |
| "learning_rate": 3e-05, | |
| "loss": 12.391, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 16.8241965973535, | |
| "grad_norm": 39907684.0, | |
| "learning_rate": 3e-05, | |
| "loss": 12.3715, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 15.876049995422363, | |
| "eval_runtime": 5.1121, | |
| "eval_samples_per_second": 97.807, | |
| "eval_steps_per_second": 12.324, | |
| "step": 8993 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 8993 | |
| }, | |
| { | |
| "epoch": 17.013232514177695, | |
| "grad_norm": 5489768.0, | |
| "learning_rate": 3e-05, | |
| "loss": 12.9394, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.20226843100189, | |
| "grad_norm": 552121280.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.0591, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 4284707.5, | |
| "learning_rate": 3e-05, | |
| "loss": 14.7709, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.58034026465028, | |
| "grad_norm": 32860224.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.33, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 17.76937618147448, | |
| "grad_norm": 500232672.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.2314, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 17.958412098298677, | |
| "grad_norm": 138954320.0, | |
| "learning_rate": 3e-05, | |
| "loss": 14.0399, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 15.959660530090332, | |
| "eval_runtime": 5.4552, | |
| "eval_samples_per_second": 91.656, | |
| "eval_steps_per_second": 11.549, | |
| "step": 9522 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 9522 | |
| }, | |
| { | |
| "epoch": 18.147448015122873, | |
| "grad_norm": 17529152.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.3582, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 18.33648393194707, | |
| "grad_norm": 14452664.0, | |
| "learning_rate": 3e-05, | |
| "loss": 16.0595, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 18.525519848771268, | |
| "grad_norm": 9175341.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.2152, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 18.714555765595463, | |
| "grad_norm": 3109339.75, | |
| "learning_rate": 3e-05, | |
| "loss": 15.6083, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 18.90359168241966, | |
| "grad_norm": 1141542.125, | |
| "learning_rate": 3e-05, | |
| "loss": 15.6927, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 16.572154998779297, | |
| "eval_runtime": 5.784, | |
| "eval_samples_per_second": 86.445, | |
| "eval_steps_per_second": 10.892, | |
| "step": 10051 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 10051 | |
| }, | |
| { | |
| "epoch": 19.092627599243855, | |
| "grad_norm": 3051544.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.4752, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 19.281663516068054, | |
| "grad_norm": 32653528.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.3329, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.47069943289225, | |
| "grad_norm": 373167.34375, | |
| "learning_rate": 3e-05, | |
| "loss": 15.1992, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 19.659735349716446, | |
| "grad_norm": 761316507648.0, | |
| "learning_rate": 3e-05, | |
| "loss": 14.5326, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 19.84877126654064, | |
| "grad_norm": 15194965.0, | |
| "learning_rate": 3e-05, | |
| "loss": 14.6703, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.2026153846153846, | |
| "eval_loss": 12.01396656036377, | |
| "eval_runtime": 5.164, | |
| "eval_samples_per_second": 96.824, | |
| "eval_steps_per_second": 12.2, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 20.03780718336484, | |
| "grad_norm": 1443913.625, | |
| "learning_rate": 3e-05, | |
| "loss": 13.4546, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 20.226843100189036, | |
| "grad_norm": 38678328.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.1074, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 20.415879017013232, | |
| "grad_norm": 235378832.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.8794, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 20.604914933837428, | |
| "grad_norm": 7457460.5, | |
| "learning_rate": 3e-05, | |
| "loss": 15.9728, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 20.793950850661627, | |
| "grad_norm": 360401824.0, | |
| "learning_rate": 3e-05, | |
| "loss": 16.3397, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 20.982986767485823, | |
| "grad_norm": 5285015.0, | |
| "learning_rate": 3e-05, | |
| "loss": 16.1349, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 16.624380111694336, | |
| "eval_runtime": 5.0669, | |
| "eval_samples_per_second": 98.68, | |
| "eval_steps_per_second": 12.434, | |
| "step": 11109 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 11109 | |
| }, | |
| { | |
| "epoch": 21.17202268431002, | |
| "grad_norm": 3120185999360.0, | |
| "learning_rate": 3e-05, | |
| "loss": 15.2325, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 21.361058601134214, | |
| "grad_norm": 459053662208.0, | |
| "learning_rate": 3e-05, | |
| "loss": 14.9804, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 21.550094517958414, | |
| "grad_norm": 33530312704.0, | |
| "learning_rate": 3e-05, | |
| "loss": 13.4482, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 21.73913043478261, | |
| "grad_norm": 524117879029760.0, | |
| "learning_rate": 3e-05, | |
| "loss": 12.8065, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 21.928166351606805, | |
| "grad_norm": 73415352385536.0, | |
| "learning_rate": 3e-05, | |
| "loss": 13.1619, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 15.610301971435547, | |
| "eval_runtime": 5.8348, | |
| "eval_samples_per_second": 85.693, | |
| "eval_steps_per_second": 10.797, | |
| "step": 11638 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 11638 | |
| }, | |
| { | |
| "epoch": 22.117202268431, | |
| "grad_norm": 24663130374144.0, | |
| "learning_rate": 3e-05, | |
| "loss": 13.4567, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 22.3062381852552, | |
| "grad_norm": 1152798.75, | |
| "learning_rate": 3e-05, | |
| "loss": 13.3323, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 22.495274102079396, | |
| "grad_norm": 6407195.0, | |
| "learning_rate": 3e-05, | |
| "loss": 12.224, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 22.68431001890359, | |
| "grad_norm": 3026044.75, | |
| "learning_rate": 3e-05, | |
| "loss": 11.6534, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 22.873345935727787, | |
| "grad_norm": 433589.25, | |
| "learning_rate": 3e-05, | |
| "loss": 11.176, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.20005128205128206, | |
| "eval_loss": 9.37710952758789, | |
| "eval_runtime": 5.4109, | |
| "eval_samples_per_second": 92.407, | |
| "eval_steps_per_second": 11.643, | |
| "step": 12167 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 12167 | |
| }, | |
| { | |
| "epoch": 23.062381852551987, | |
| "grad_norm": 221080977408.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.2085, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 23.251417769376182, | |
| "grad_norm": 819255836672.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.3606, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 23.440453686200378, | |
| "grad_norm": 1077065216.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.3544, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 23.629489603024574, | |
| "grad_norm": 147283214336.0, | |
| "learning_rate": 3e-05, | |
| "loss": 10.6161, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 23.81852551984877, | |
| "grad_norm": 479103936.0, | |
| "learning_rate": 3e-05, | |
| "loss": 11.0412, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.2362051282051282, | |
| "eval_loss": 8.926570892333984, | |
| "eval_runtime": 5.7624, | |
| "eval_samples_per_second": 86.769, | |
| "eval_steps_per_second": 10.933, | |
| "step": 12696 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 12696 | |
| }, | |
| { | |
| "epoch": 24.00756143667297, | |
| "grad_norm": 18736.490234375, | |
| "learning_rate": 3e-05, | |
| "loss": 10.5723, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 24.196597353497165, | |
| "grad_norm": 10122.208984375, | |
| "learning_rate": 3e-05, | |
| "loss": 9.1505, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 24.38563327032136, | |
| "grad_norm": 884941.875, | |
| "learning_rate": 3e-05, | |
| "loss": 8.6122, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 24.574669187145556, | |
| "grad_norm": 51996.421875, | |
| "learning_rate": 3e-05, | |
| "loss": 8.5405, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 24.763705103969755, | |
| "grad_norm": 11833.41796875, | |
| "learning_rate": 3e-05, | |
| "loss": 8.4026, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 24.95274102079395, | |
| "grad_norm": 645689536.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.8444, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.20235897435897435, | |
| "eval_loss": 10.400785446166992, | |
| "eval_runtime": 5.8159, | |
| "eval_samples_per_second": 85.971, | |
| "eval_steps_per_second": 10.832, | |
| "step": 13225 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 13225 | |
| }, | |
| { | |
| "epoch": 25.141776937618147, | |
| "grad_norm": 33455.7734375, | |
| "learning_rate": 3e-05, | |
| "loss": 8.9314, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 25.330812854442343, | |
| "grad_norm": 665939.625, | |
| "learning_rate": 3e-05, | |
| "loss": 8.9419, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 25.519848771266542, | |
| "grad_norm": 366503.25, | |
| "learning_rate": 3e-05, | |
| "loss": 9.1085, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 25.708884688090738, | |
| "grad_norm": 1146194755584.0, | |
| "learning_rate": 3e-05, | |
| "loss": 9.1301, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 25.897920604914933, | |
| "grad_norm": 5115191230464.0, | |
| "learning_rate": 3e-05, | |
| "loss": 8.9435, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.1995897435897436, | |
| "eval_loss": 12.846183776855469, | |
| "eval_runtime": 5.1674, | |
| "eval_samples_per_second": 96.76, | |
| "eval_steps_per_second": 12.192, | |
| "step": 13754 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 13754 | |
| }, | |
| { | |
| "epoch": 26.08695652173913, | |
| "grad_norm": 119.87647247314453, | |
| "learning_rate": 3e-05, | |
| "loss": 10.1216, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 26.27599243856333, | |
| "grad_norm": 5832.49462890625, | |
| "learning_rate": 3e-05, | |
| "loss": 8.5386, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 26.465028355387524, | |
| "grad_norm": 7308.30908203125, | |
| "learning_rate": 3e-05, | |
| "loss": 8.1093, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 26.65406427221172, | |
| "grad_norm": 21071.32421875, | |
| "learning_rate": 3e-05, | |
| "loss": 8.0342, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 26.843100189035916, | |
| "grad_norm": 69589.0625, | |
| "learning_rate": 3e-05, | |
| "loss": 8.0313, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.23364102564102565, | |
| "eval_loss": 7.521541595458984, | |
| "eval_runtime": 5.0339, | |
| "eval_samples_per_second": 99.326, | |
| "eval_steps_per_second": 12.515, | |
| "step": 14283 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "step": 14283 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 26450, | |
| "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": 1.1626185967072707e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |