{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 16.0,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.8,
"grad_norm": 1.3812153339385986,
"learning_rate": 5.7005e-05,
"loss": 0.3305,
"step": 100
},
{
"epoch": 1.0,
"eval_accuracy_no_text": 0.958563454164822,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.958563454164822,
"eval_iou_text": 0.0,
"eval_loss": 0.1845872551202774,
"eval_mean_accuracy": 0.958563454164822,
"eval_mean_iou": 0.479281727082411,
"eval_overall_accuracy": 0.958563454164822,
"eval_runtime": 9.574,
"eval_samples_per_second": 26.53,
"eval_steps_per_second": 3.342,
"step": 125
},
{
"epoch": 1.6,
"grad_norm": 2.7767746448516846,
"learning_rate": 5.401e-05,
"loss": 0.2037,
"step": 200
},
{
"epoch": 2.0,
"eval_accuracy_no_text": 0.9706028320744683,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9706028320744683,
"eval_iou_text": 0.0,
"eval_loss": 0.13224510848522186,
"eval_mean_accuracy": 0.9706028320744683,
"eval_mean_iou": 0.48530141603723415,
"eval_overall_accuracy": 0.9706028320744683,
"eval_runtime": 10.1041,
"eval_samples_per_second": 25.138,
"eval_steps_per_second": 3.167,
"step": 250
},
{
"epoch": 2.4,
"grad_norm": 1.592877984046936,
"learning_rate": 5.1015e-05,
"loss": 0.1534,
"step": 300
},
{
"epoch": 3.0,
"eval_accuracy_no_text": 0.9784367089170211,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9784367089170211,
"eval_iou_text": 0.0,
"eval_loss": 0.10737518221139908,
"eval_mean_accuracy": 0.9784367089170211,
"eval_mean_iou": 0.48921835445851053,
"eval_overall_accuracy": 0.9784367089170211,
"eval_runtime": 9.6164,
"eval_samples_per_second": 26.413,
"eval_steps_per_second": 3.328,
"step": 375
},
{
"epoch": 3.2,
"grad_norm": 0.9847292304039001,
"learning_rate": 4.8020000000000004e-05,
"loss": 0.1451,
"step": 400
},
{
"epoch": 4.0,
"grad_norm": 0.3838343918323517,
"learning_rate": 4.5025000000000003e-05,
"loss": 0.1313,
"step": 500
},
{
"epoch": 4.0,
"eval_accuracy_no_text": 0.9839048439461286,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9839048439461286,
"eval_iou_text": 0.0,
"eval_loss": 0.09760396927595139,
"eval_mean_accuracy": 0.9839048439461286,
"eval_mean_iou": 0.4919524219730643,
"eval_overall_accuracy": 0.9839048439461286,
"eval_runtime": 10.3548,
"eval_samples_per_second": 24.53,
"eval_steps_per_second": 3.09,
"step": 500
},
{
"epoch": 4.8,
"grad_norm": 1.1984161138534546,
"learning_rate": 4.203e-05,
"loss": 0.1156,
"step": 600
},
{
"epoch": 5.0,
"eval_accuracy_no_text": 0.9799307811869516,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9799307811869516,
"eval_iou_text": 0.0,
"eval_loss": 0.1000521183013916,
"eval_mean_accuracy": 0.9799307811869516,
"eval_mean_iou": 0.4899653905934758,
"eval_overall_accuracy": 0.9799307811869516,
"eval_runtime": 9.868,
"eval_samples_per_second": 25.74,
"eval_steps_per_second": 3.243,
"step": 625
},
{
"epoch": 5.6,
"grad_norm": 2.07499361038208,
"learning_rate": 3.9035e-05,
"loss": 0.1123,
"step": 700
},
{
"epoch": 6.0,
"eval_accuracy_no_text": 0.986626541167033,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.986626541167033,
"eval_iou_text": 0.0,
"eval_loss": 0.09204956144094467,
"eval_mean_accuracy": 0.986626541167033,
"eval_mean_iou": 0.4933132705835165,
"eval_overall_accuracy": 0.986626541167033,
"eval_runtime": 10.6441,
"eval_samples_per_second": 23.863,
"eval_steps_per_second": 3.006,
"step": 750
},
{
"epoch": 6.4,
"grad_norm": 1.8263524770736694,
"learning_rate": 3.604e-05,
"loss": 0.108,
"step": 800
},
{
"epoch": 7.0,
"eval_accuracy_no_text": 0.9815042758929395,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9815042758929395,
"eval_iou_text": 0.0,
"eval_loss": 0.09463128447532654,
"eval_mean_accuracy": 0.9815042758929395,
"eval_mean_iou": 0.49075213794646977,
"eval_overall_accuracy": 0.9815042758929395,
"eval_runtime": 9.8625,
"eval_samples_per_second": 25.754,
"eval_steps_per_second": 3.245,
"step": 875
},
{
"epoch": 7.2,
"grad_norm": 0.35387736558914185,
"learning_rate": 3.3045000000000006e-05,
"loss": 0.1015,
"step": 900
},
{
"epoch": 8.0,
"grad_norm": 0.5748061537742615,
"learning_rate": 3.0050000000000002e-05,
"loss": 0.1017,
"step": 1000
},
{
"epoch": 8.0,
"eval_accuracy_no_text": 0.980515017943572,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.980515017943572,
"eval_iou_text": 0.0,
"eval_loss": 0.09426650404930115,
"eval_mean_accuracy": 0.980515017943572,
"eval_mean_iou": 0.490257508971786,
"eval_overall_accuracy": 0.980515017943572,
"eval_runtime": 9.8323,
"eval_samples_per_second": 25.833,
"eval_steps_per_second": 3.255,
"step": 1000
},
{
"epoch": 8.8,
"grad_norm": 3.100071668624878,
"learning_rate": 2.7054999999999998e-05,
"loss": 0.0994,
"step": 1100
},
{
"epoch": 9.0,
"eval_accuracy_no_text": 0.9807879619513361,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9807879619513361,
"eval_iou_text": 0.0,
"eval_loss": 0.0927177369594574,
"eval_mean_accuracy": 0.9807879619513361,
"eval_mean_iou": 0.49039398097566805,
"eval_overall_accuracy": 0.9807879619513361,
"eval_runtime": 9.5796,
"eval_samples_per_second": 26.515,
"eval_steps_per_second": 3.34,
"step": 1125
},
{
"epoch": 9.6,
"grad_norm": 0.8802123665809631,
"learning_rate": 2.406e-05,
"loss": 0.0926,
"step": 1200
},
{
"epoch": 10.0,
"eval_accuracy_no_text": 0.9782629100297493,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9782629100297493,
"eval_iou_text": 0.0,
"eval_loss": 0.0956711545586586,
"eval_mean_accuracy": 0.9782629100297493,
"eval_mean_iou": 0.48913145501487465,
"eval_overall_accuracy": 0.9782629100297493,
"eval_runtime": 10.2441,
"eval_samples_per_second": 24.795,
"eval_steps_per_second": 3.124,
"step": 1250
},
{
"epoch": 10.4,
"grad_norm": 0.4630286693572998,
"learning_rate": 2.1065e-05,
"loss": 0.0907,
"step": 1300
},
{
"epoch": 11.0,
"eval_accuracy_no_text": 0.9829907049995675,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9829907049995675,
"eval_iou_text": 0.0,
"eval_loss": 0.09134823828935623,
"eval_mean_accuracy": 0.9829907049995675,
"eval_mean_iou": 0.49149535249978377,
"eval_overall_accuracy": 0.9829907049995675,
"eval_runtime": 9.7246,
"eval_samples_per_second": 26.119,
"eval_steps_per_second": 3.291,
"step": 1375
},
{
"epoch": 11.2,
"grad_norm": 57.36943435668945,
"learning_rate": 1.807e-05,
"loss": 0.0964,
"step": 1400
},
{
"epoch": 12.0,
"grad_norm": 1.3222465515136719,
"learning_rate": 1.5075000000000002e-05,
"loss": 0.0893,
"step": 1500
},
{
"epoch": 12.0,
"eval_accuracy_no_text": 0.9838469054451806,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9838469054451806,
"eval_iou_text": 0.0,
"eval_loss": 0.08928046375513077,
"eval_mean_accuracy": 0.9838469054451806,
"eval_mean_iou": 0.4919234527225903,
"eval_overall_accuracy": 0.9838469054451806,
"eval_runtime": 9.6536,
"eval_samples_per_second": 26.311,
"eval_steps_per_second": 3.315,
"step": 1500
},
{
"epoch": 12.8,
"grad_norm": 1.2742468118667603,
"learning_rate": 1.2079999999999998e-05,
"loss": 0.0853,
"step": 1600
},
{
"epoch": 13.0,
"eval_accuracy_no_text": 0.9803520856406479,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9803520856406479,
"eval_iou_text": 0.0,
"eval_loss": 0.09133532643318176,
"eval_mean_accuracy": 0.9803520856406479,
"eval_mean_iou": 0.49017604282032395,
"eval_overall_accuracy": 0.9803520856406479,
"eval_runtime": 10.3368,
"eval_samples_per_second": 24.572,
"eval_steps_per_second": 3.096,
"step": 1625
},
{
"epoch": 13.6,
"grad_norm": 5.432977199554443,
"learning_rate": 9.085000000000002e-06,
"loss": 0.0834,
"step": 1700
},
{
"epoch": 14.0,
"eval_accuracy_no_text": 0.98199298972385,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.98199298972385,
"eval_iou_text": 0.0,
"eval_loss": 0.08987393975257874,
"eval_mean_accuracy": 0.98199298972385,
"eval_mean_iou": 0.490996494861925,
"eval_overall_accuracy": 0.98199298972385,
"eval_runtime": 9.5985,
"eval_samples_per_second": 26.463,
"eval_steps_per_second": 3.334,
"step": 1750
},
{
"epoch": 14.4,
"grad_norm": 1.3897353410720825,
"learning_rate": 6.0899999999999984e-06,
"loss": 0.0861,
"step": 1800
},
{
"epoch": 15.0,
"eval_accuracy_no_text": 0.9814577689058006,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9814577689058006,
"eval_iou_text": 0.0,
"eval_loss": 0.0901678055524826,
"eval_mean_accuracy": 0.9814577689058006,
"eval_mean_iou": 0.4907288844529003,
"eval_overall_accuracy": 0.9814577689058006,
"eval_runtime": 9.8323,
"eval_samples_per_second": 25.833,
"eval_steps_per_second": 3.255,
"step": 1875
},
{
"epoch": 15.2,
"grad_norm": 0.7514830231666565,
"learning_rate": 3.0950000000000026e-06,
"loss": 0.0882,
"step": 1900
},
{
"epoch": 16.0,
"grad_norm": 0.9411144852638245,
"learning_rate": 1e-07,
"loss": 0.0803,
"step": 2000
},
{
"epoch": 16.0,
"eval_accuracy_no_text": 0.9793073981417643,
"eval_accuracy_text": NaN,
"eval_iou_no_text": 0.9793073981417643,
"eval_iou_text": 0.0,
"eval_loss": 0.09290074557065964,
"eval_mean_accuracy": 0.9793073981417643,
"eval_mean_iou": 0.48965369907088213,
"eval_overall_accuracy": 0.9793073981417643,
"eval_runtime": 9.6102,
"eval_samples_per_second": 26.43,
"eval_steps_per_second": 3.33,
"step": 2000
},
{
"epoch": 16.0,
"step": 2000,
"total_flos": 1.032636453617664e+18,
"train_loss": 0.11973714399337769,
"train_runtime": 853.4119,
"train_samples_per_second": 18.748,
"train_steps_per_second": 2.344
}
],
"logging_steps": 100,
"max_steps": 2000,
"num_input_tokens_seen": 0,
"num_train_epochs": 16,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.032636453617664e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}