breeze-listen-dsw-base-id / trainer_state.json
hanasim's picture
End of training
71d6b7e verified
{
"best_metric": 32.82730701996504,
"best_model_checkpoint": "/cosmos/home/sp-operator/ai/training/models/huggingface/scripts/../breeze-listen-dsw-base-id/checkpoint-600",
"epoch": 12.064,
"eval_steps": 200,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.898977360288234e-06,
"loss": 1.4272,
"step": 25
},
{
"epoch": 0.03,
"learning_rate": 6.160712527409633e-06,
"loss": 1.0004,
"step": 50
},
{
"epoch": 0.04,
"learning_rate": 6.85912902234906e-06,
"loss": 0.8231,
"step": 75
},
{
"epoch": 0.05,
"learning_rate": 7.344547104469332e-06,
"loss": 0.7054,
"step": 100
},
{
"epoch": 0.06,
"learning_rate": 7.716963756434345e-06,
"loss": 0.531,
"step": 125
},
{
"epoch": 0.07,
"learning_rate": 8.019180844200955e-06,
"loss": 0.4107,
"step": 150
},
{
"epoch": 1.01,
"learning_rate": 8.27351214279797e-06,
"loss": 0.546,
"step": 175
},
{
"epoch": 1.02,
"learning_rate": 8.49307723936858e-06,
"loss": 0.5452,
"step": 200
},
{
"epoch": 1.02,
"eval_loss": 0.54638671875,
"eval_runtime": 2392.0752,
"eval_samples_per_second": 1.523,
"eval_steps_per_second": 0.095,
"eval_wer": 35.168828779096515,
"step": 200
},
{
"epoch": 1.03,
"learning_rate": 8.686247975778677e-06,
"loss": 0.5942,
"step": 225
},
{
"epoch": 1.05,
"learning_rate": 8.858694625217149e-06,
"loss": 0.4591,
"step": 250
},
{
"epoch": 1.06,
"learning_rate": 9.014436199608479e-06,
"loss": 0.3818,
"step": 275
},
{
"epoch": 1.07,
"learning_rate": 9.156425255148058e-06,
"loss": 0.2725,
"step": 300
},
{
"epoch": 2.01,
"learning_rate": 9.28689473531776e-06,
"loss": 0.3299,
"step": 325
},
{
"epoch": 2.02,
"learning_rate": 9.407574351377137e-06,
"loss": 0.3841,
"step": 350
},
{
"epoch": 2.03,
"learning_rate": 9.519831289296397e-06,
"loss": 0.4134,
"step": 375
},
{
"epoch": 2.04,
"learning_rate": 9.624764935335318e-06,
"loss": 0.3445,
"step": 400
},
{
"epoch": 2.04,
"eval_loss": 0.54052734375,
"eval_runtime": 2421.7938,
"eval_samples_per_second": 1.504,
"eval_steps_per_second": 0.094,
"eval_wer": 34.06937160732358,
"step": 400
},
{
"epoch": 2.06,
"learning_rate": 9.723272550712454e-06,
"loss": 0.2894,
"step": 425
},
{
"epoch": 2.07,
"learning_rate": 9.816095971633122e-06,
"loss": 0.199,
"step": 450
},
{
"epoch": 3.0,
"learning_rate": 9.90385555539545e-06,
"loss": 0.1939,
"step": 475
},
{
"epoch": 3.02,
"learning_rate": 9.987075336738768e-06,
"loss": 0.2661,
"step": 500
},
{
"epoch": 3.03,
"learning_rate": 9.866666666666668e-06,
"loss": 0.2885,
"step": 525
},
{
"epoch": 3.04,
"learning_rate": 9.7e-06,
"loss": 0.2474,
"step": 550
},
{
"epoch": 3.05,
"learning_rate": 9.533333333333334e-06,
"loss": 0.195,
"step": 575
},
{
"epoch": 3.07,
"learning_rate": 9.366666666666668e-06,
"loss": 0.1397,
"step": 600
},
{
"epoch": 3.07,
"eval_loss": 0.53466796875,
"eval_runtime": 2319.7619,
"eval_samples_per_second": 1.57,
"eval_steps_per_second": 0.098,
"eval_wer": 32.82730701996504,
"step": 600
},
{
"epoch": 4.0,
"learning_rate": 9.200000000000002e-06,
"loss": 0.1158,
"step": 625
},
{
"epoch": 4.01,
"learning_rate": 9.033333333333334e-06,
"loss": 0.1686,
"step": 650
},
{
"epoch": 4.03,
"learning_rate": 8.866666666666668e-06,
"loss": 0.183,
"step": 675
},
{
"epoch": 4.04,
"learning_rate": 8.700000000000001e-06,
"loss": 0.1702,
"step": 700
},
{
"epoch": 4.05,
"learning_rate": 8.533333333333335e-06,
"loss": 0.138,
"step": 725
},
{
"epoch": 4.06,
"learning_rate": 8.366666666666667e-06,
"loss": 0.1002,
"step": 750
},
{
"epoch": 4.08,
"learning_rate": 8.2e-06,
"loss": 0.0697,
"step": 775
},
{
"epoch": 5.01,
"learning_rate": 8.033333333333335e-06,
"loss": 0.0988,
"step": 800
},
{
"epoch": 5.01,
"eval_loss": 0.5654296875,
"eval_runtime": 2396.345,
"eval_samples_per_second": 1.52,
"eval_steps_per_second": 0.095,
"eval_wer": 35.67485509246481,
"step": 800
},
{
"epoch": 5.02,
"learning_rate": 7.866666666666667e-06,
"loss": 0.1161,
"step": 825
},
{
"epoch": 5.04,
"learning_rate": 7.7e-06,
"loss": 0.1152,
"step": 850
},
{
"epoch": 5.05,
"learning_rate": 7.533333333333334e-06,
"loss": 0.0922,
"step": 875
},
{
"epoch": 5.06,
"learning_rate": 7.3666666666666676e-06,
"loss": 0.073,
"step": 900
},
{
"epoch": 5.07,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.0481,
"step": 925
},
{
"epoch": 6.01,
"learning_rate": 7.033333333333334e-06,
"loss": 0.056,
"step": 950
},
{
"epoch": 6.02,
"learning_rate": 6.866666666666667e-06,
"loss": 0.0705,
"step": 975
},
{
"epoch": 6.03,
"learning_rate": 6.700000000000001e-06,
"loss": 0.077,
"step": 1000
},
{
"epoch": 6.03,
"eval_loss": 0.57861328125,
"eval_runtime": 2311.9939,
"eval_samples_per_second": 1.575,
"eval_steps_per_second": 0.099,
"eval_wer": 33.945165148587726,
"step": 1000
},
{
"epoch": 6.04,
"learning_rate": 6.533333333333334e-06,
"loss": 0.0629,
"step": 1025
},
{
"epoch": 6.06,
"learning_rate": 6.366666666666668e-06,
"loss": 0.0496,
"step": 1050
},
{
"epoch": 6.07,
"learning_rate": 6.200000000000001e-06,
"loss": 0.0328,
"step": 1075
},
{
"epoch": 7.0,
"learning_rate": 6.033333333333335e-06,
"loss": 0.0331,
"step": 1100
},
{
"epoch": 7.02,
"learning_rate": 5.8666666666666675e-06,
"loss": 0.0432,
"step": 1125
},
{
"epoch": 7.03,
"learning_rate": 5.7e-06,
"loss": 0.0475,
"step": 1150
},
{
"epoch": 7.04,
"learning_rate": 5.533333333333334e-06,
"loss": 0.0446,
"step": 1175
},
{
"epoch": 7.05,
"learning_rate": 5.366666666666666e-06,
"loss": 0.0338,
"step": 1200
},
{
"epoch": 7.05,
"eval_loss": 0.60498046875,
"eval_runtime": 2281.4213,
"eval_samples_per_second": 1.596,
"eval_steps_per_second": 0.1,
"eval_wer": 33.98196706228724,
"step": 1200
},
{
"epoch": 7.07,
"learning_rate": 5.2e-06,
"loss": 0.0242,
"step": 1225
},
{
"epoch": 8.0,
"learning_rate": 5.033333333333333e-06,
"loss": 0.0197,
"step": 1250
},
{
"epoch": 8.01,
"learning_rate": 4.866666666666667e-06,
"loss": 0.0269,
"step": 1275
},
{
"epoch": 8.03,
"learning_rate": 4.7e-06,
"loss": 0.0319,
"step": 1300
},
{
"epoch": 8.04,
"learning_rate": 4.533333333333334e-06,
"loss": 0.0293,
"step": 1325
},
{
"epoch": 8.05,
"learning_rate": 4.366666666666667e-06,
"loss": 0.023,
"step": 1350
},
{
"epoch": 8.06,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.0189,
"step": 1375
},
{
"epoch": 8.08,
"learning_rate": 4.033333333333333e-06,
"loss": 0.0137,
"step": 1400
},
{
"epoch": 8.08,
"eval_loss": 0.6220703125,
"eval_runtime": 2300.7263,
"eval_samples_per_second": 1.583,
"eval_steps_per_second": 0.099,
"eval_wer": 34.10157328181065,
"step": 1400
},
{
"epoch": 9.01,
"learning_rate": 3.866666666666667e-06,
"loss": 0.0189,
"step": 1425
},
{
"epoch": 9.02,
"learning_rate": 3.7e-06,
"loss": 0.02,
"step": 1450
},
{
"epoch": 9.04,
"learning_rate": 3.5333333333333335e-06,
"loss": 0.0213,
"step": 1475
},
{
"epoch": 9.05,
"learning_rate": 3.366666666666667e-06,
"loss": 0.0179,
"step": 1500
},
{
"epoch": 9.06,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.0139,
"step": 1525
},
{
"epoch": 9.07,
"learning_rate": 3.0333333333333337e-06,
"loss": 0.0111,
"step": 1550
},
{
"epoch": 10.01,
"learning_rate": 2.866666666666667e-06,
"loss": 0.0129,
"step": 1575
},
{
"epoch": 10.02,
"learning_rate": 2.7000000000000004e-06,
"loss": 0.0153,
"step": 1600
},
{
"epoch": 10.02,
"eval_loss": 0.64306640625,
"eval_runtime": 2254.9316,
"eval_samples_per_second": 1.615,
"eval_steps_per_second": 0.101,
"eval_wer": 33.90376299567578,
"step": 1600
},
{
"epoch": 10.03,
"learning_rate": 2.5333333333333338e-06,
"loss": 0.0166,
"step": 1625
},
{
"epoch": 10.04,
"learning_rate": 2.3666666666666667e-06,
"loss": 0.0142,
"step": 1650
},
{
"epoch": 10.06,
"learning_rate": 2.2e-06,
"loss": 0.0119,
"step": 1675
},
{
"epoch": 10.07,
"learning_rate": 2.0333333333333335e-06,
"loss": 0.0101,
"step": 1700
},
{
"epoch": 11.0,
"learning_rate": 1.8666666666666669e-06,
"loss": 0.0101,
"step": 1725
},
{
"epoch": 11.02,
"learning_rate": 1.7000000000000002e-06,
"loss": 0.0125,
"step": 1750
},
{
"epoch": 11.03,
"learning_rate": 1.5333333333333334e-06,
"loss": 0.0141,
"step": 1775
},
{
"epoch": 11.04,
"learning_rate": 1.3666666666666668e-06,
"loss": 0.0125,
"step": 1800
},
{
"epoch": 11.04,
"eval_loss": 0.6513671875,
"eval_runtime": 2241.116,
"eval_samples_per_second": 1.625,
"eval_steps_per_second": 0.102,
"eval_wer": 33.75195510166529,
"step": 1800
},
{
"epoch": 11.05,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.0114,
"step": 1825
},
{
"epoch": 11.07,
"learning_rate": 1.0333333333333333e-06,
"loss": 0.0093,
"step": 1850
},
{
"epoch": 12.0,
"learning_rate": 8.666666666666668e-07,
"loss": 0.0087,
"step": 1875
},
{
"epoch": 12.01,
"learning_rate": 7.000000000000001e-07,
"loss": 0.0111,
"step": 1900
},
{
"epoch": 12.03,
"learning_rate": 5.333333333333335e-07,
"loss": 0.0129,
"step": 1925
},
{
"epoch": 12.04,
"learning_rate": 3.666666666666667e-07,
"loss": 0.0118,
"step": 1950
},
{
"epoch": 12.05,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.011,
"step": 1975
},
{
"epoch": 12.06,
"learning_rate": 3.333333333333334e-08,
"loss": 0.0092,
"step": 2000
},
{
"epoch": 12.06,
"eval_loss": 0.65283203125,
"eval_runtime": 2242.3057,
"eval_samples_per_second": 1.624,
"eval_steps_per_second": 0.102,
"eval_wer": 33.82555892906431,
"step": 2000
},
{
"epoch": 12.06,
"step": 2000,
"total_flos": 4.1331384945643356e+18,
"train_loss": 0.16917743301391602,
"train_runtime": 75059.9366,
"train_samples_per_second": 0.853,
"train_steps_per_second": 0.027
}
],
"logging_steps": 25,
"max_steps": 2000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 200,
"total_flos": 4.1331384945643356e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}