exper7_mesum5 / trainer_state.json
sudo-s's picture
🍻 cheers
cbf2277
{
"best_metric": 0.5888754725456238,
"best_model_checkpoint": "./exper7_mesum5/checkpoint-3300",
"epoch": 10.0,
"global_step": 4300,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 9.976744186046511e-05,
"loss": 4.9752,
"step": 10
},
{
"epoch": 0.05,
"learning_rate": 9.953488372093024e-05,
"loss": 4.9072,
"step": 20
},
{
"epoch": 0.07,
"learning_rate": 9.930232558139535e-05,
"loss": 4.7762,
"step": 30
},
{
"epoch": 0.09,
"learning_rate": 9.906976744186048e-05,
"loss": 4.6977,
"step": 40
},
{
"epoch": 0.12,
"learning_rate": 9.883720930232558e-05,
"loss": 4.6261,
"step": 50
},
{
"epoch": 0.14,
"learning_rate": 9.86046511627907e-05,
"loss": 4.4953,
"step": 60
},
{
"epoch": 0.16,
"learning_rate": 9.837209302325582e-05,
"loss": 4.4588,
"step": 70
},
{
"epoch": 0.19,
"learning_rate": 9.813953488372092e-05,
"loss": 4.3419,
"step": 80
},
{
"epoch": 0.21,
"learning_rate": 9.790697674418605e-05,
"loss": 4.2548,
"step": 90
},
{
"epoch": 0.23,
"learning_rate": 9.767441860465116e-05,
"loss": 4.2072,
"step": 100
},
{
"epoch": 0.23,
"eval_accuracy": 0.19230769230769232,
"eval_loss": 4.153234958648682,
"eval_runtime": 19.6138,
"eval_samples_per_second": 86.164,
"eval_steps_per_second": 10.809,
"step": 100
},
{
"epoch": 0.26,
"learning_rate": 9.744186046511629e-05,
"loss": 4.1426,
"step": 110
},
{
"epoch": 0.28,
"learning_rate": 9.72093023255814e-05,
"loss": 4.034,
"step": 120
},
{
"epoch": 0.3,
"learning_rate": 9.697674418604651e-05,
"loss": 3.9142,
"step": 130
},
{
"epoch": 0.33,
"learning_rate": 9.674418604651163e-05,
"loss": 3.9224,
"step": 140
},
{
"epoch": 0.35,
"learning_rate": 9.651162790697675e-05,
"loss": 3.8816,
"step": 150
},
{
"epoch": 0.37,
"learning_rate": 9.627906976744187e-05,
"loss": 3.7969,
"step": 160
},
{
"epoch": 0.4,
"learning_rate": 9.604651162790697e-05,
"loss": 3.8209,
"step": 170
},
{
"epoch": 0.42,
"learning_rate": 9.58139534883721e-05,
"loss": 3.7072,
"step": 180
},
{
"epoch": 0.44,
"learning_rate": 9.558139534883721e-05,
"loss": 3.722,
"step": 190
},
{
"epoch": 0.47,
"learning_rate": 9.534883720930233e-05,
"loss": 3.5433,
"step": 200
},
{
"epoch": 0.47,
"eval_accuracy": 0.28875739644970416,
"eval_loss": 3.568031072616577,
"eval_runtime": 21.594,
"eval_samples_per_second": 78.262,
"eval_steps_per_second": 9.818,
"step": 200
},
{
"epoch": 0.49,
"learning_rate": 9.511627906976745e-05,
"loss": 3.4805,
"step": 210
},
{
"epoch": 0.51,
"learning_rate": 9.488372093023256e-05,
"loss": 3.5134,
"step": 220
},
{
"epoch": 0.53,
"learning_rate": 9.465116279069768e-05,
"loss": 3.4267,
"step": 230
},
{
"epoch": 0.56,
"learning_rate": 9.441860465116279e-05,
"loss": 3.4925,
"step": 240
},
{
"epoch": 0.58,
"learning_rate": 9.418604651162792e-05,
"loss": 3.4093,
"step": 250
},
{
"epoch": 0.6,
"learning_rate": 9.395348837209302e-05,
"loss": 3.2541,
"step": 260
},
{
"epoch": 0.63,
"learning_rate": 9.372093023255814e-05,
"loss": 3.1907,
"step": 270
},
{
"epoch": 0.65,
"learning_rate": 9.348837209302326e-05,
"loss": 3.3283,
"step": 280
},
{
"epoch": 0.67,
"learning_rate": 9.325581395348838e-05,
"loss": 3.1848,
"step": 290
},
{
"epoch": 0.7,
"learning_rate": 9.30232558139535e-05,
"loss": 3.1388,
"step": 300
},
{
"epoch": 0.7,
"eval_accuracy": 0.39112426035502956,
"eval_loss": 3.120227336883545,
"eval_runtime": 21.1809,
"eval_samples_per_second": 79.789,
"eval_steps_per_second": 10.009,
"step": 300
},
{
"epoch": 0.72,
"learning_rate": 9.279069767441861e-05,
"loss": 3.1098,
"step": 310
},
{
"epoch": 0.74,
"learning_rate": 9.255813953488372e-05,
"loss": 3.0793,
"step": 320
},
{
"epoch": 0.77,
"learning_rate": 9.232558139534884e-05,
"loss": 3.0589,
"step": 330
},
{
"epoch": 0.79,
"learning_rate": 9.209302325581396e-05,
"loss": 2.999,
"step": 340
},
{
"epoch": 0.81,
"learning_rate": 9.186046511627907e-05,
"loss": 2.8881,
"step": 350
},
{
"epoch": 0.84,
"learning_rate": 9.162790697674419e-05,
"loss": 2.8647,
"step": 360
},
{
"epoch": 0.86,
"learning_rate": 9.139534883720931e-05,
"loss": 2.9089,
"step": 370
},
{
"epoch": 0.88,
"learning_rate": 9.116279069767443e-05,
"loss": 2.8886,
"step": 380
},
{
"epoch": 0.91,
"learning_rate": 9.093023255813953e-05,
"loss": 2.6773,
"step": 390
},
{
"epoch": 0.93,
"learning_rate": 9.069767441860465e-05,
"loss": 2.7924,
"step": 400
},
{
"epoch": 0.93,
"eval_accuracy": 0.47869822485207103,
"eval_loss": 2.7434420585632324,
"eval_runtime": 21.8697,
"eval_samples_per_second": 77.276,
"eval_steps_per_second": 9.694,
"step": 400
},
{
"epoch": 0.95,
"learning_rate": 9.046511627906977e-05,
"loss": 2.7934,
"step": 410
},
{
"epoch": 0.98,
"learning_rate": 9.023255813953489e-05,
"loss": 2.6406,
"step": 420
},
{
"epoch": 1.0,
"learning_rate": 9e-05,
"loss": 2.6162,
"step": 430
},
{
"epoch": 1.02,
"learning_rate": 8.976744186046512e-05,
"loss": 2.4643,
"step": 440
},
{
"epoch": 1.05,
"learning_rate": 8.953488372093024e-05,
"loss": 2.3165,
"step": 450
},
{
"epoch": 1.07,
"learning_rate": 8.930232558139535e-05,
"loss": 2.2989,
"step": 460
},
{
"epoch": 1.09,
"learning_rate": 8.906976744186048e-05,
"loss": 2.4173,
"step": 470
},
{
"epoch": 1.12,
"learning_rate": 8.883720930232558e-05,
"loss": 2.3148,
"step": 480
},
{
"epoch": 1.14,
"learning_rate": 8.86046511627907e-05,
"loss": 2.2514,
"step": 490
},
{
"epoch": 1.16,
"learning_rate": 8.837209302325582e-05,
"loss": 2.1269,
"step": 500
},
{
"epoch": 1.16,
"eval_accuracy": 0.5781065088757397,
"eval_loss": 2.326211929321289,
"eval_runtime": 22.2923,
"eval_samples_per_second": 75.811,
"eval_steps_per_second": 9.51,
"step": 500
},
{
"epoch": 1.19,
"learning_rate": 8.813953488372094e-05,
"loss": 2.2364,
"step": 510
},
{
"epoch": 1.21,
"learning_rate": 8.790697674418606e-05,
"loss": 2.1101,
"step": 520
},
{
"epoch": 1.23,
"learning_rate": 8.767441860465116e-05,
"loss": 2.1031,
"step": 530
},
{
"epoch": 1.26,
"learning_rate": 8.744186046511629e-05,
"loss": 1.9173,
"step": 540
},
{
"epoch": 1.28,
"learning_rate": 8.72093023255814e-05,
"loss": 2.0489,
"step": 550
},
{
"epoch": 1.3,
"learning_rate": 8.697674418604652e-05,
"loss": 1.9887,
"step": 560
},
{
"epoch": 1.33,
"learning_rate": 8.674418604651163e-05,
"loss": 1.8983,
"step": 570
},
{
"epoch": 1.35,
"learning_rate": 8.651162790697674e-05,
"loss": 1.9514,
"step": 580
},
{
"epoch": 1.37,
"learning_rate": 8.627906976744187e-05,
"loss": 1.8825,
"step": 590
},
{
"epoch": 1.4,
"learning_rate": 8.604651162790697e-05,
"loss": 1.8589,
"step": 600
},
{
"epoch": 1.4,
"eval_accuracy": 0.6272189349112426,
"eval_loss": 1.975429654121399,
"eval_runtime": 22.5072,
"eval_samples_per_second": 75.087,
"eval_steps_per_second": 9.419,
"step": 600
},
{
"epoch": 1.42,
"learning_rate": 8.58139534883721e-05,
"loss": 1.8157,
"step": 610
},
{
"epoch": 1.44,
"learning_rate": 8.558139534883721e-05,
"loss": 1.8396,
"step": 620
},
{
"epoch": 1.47,
"learning_rate": 8.534883720930233e-05,
"loss": 1.74,
"step": 630
},
{
"epoch": 1.49,
"learning_rate": 8.511627906976745e-05,
"loss": 1.8692,
"step": 640
},
{
"epoch": 1.51,
"learning_rate": 8.488372093023255e-05,
"loss": 1.7739,
"step": 650
},
{
"epoch": 1.53,
"learning_rate": 8.465116279069768e-05,
"loss": 1.7221,
"step": 660
},
{
"epoch": 1.56,
"learning_rate": 8.441860465116279e-05,
"loss": 1.6987,
"step": 670
},
{
"epoch": 1.58,
"learning_rate": 8.418604651162792e-05,
"loss": 1.6821,
"step": 680
},
{
"epoch": 1.6,
"learning_rate": 8.395348837209302e-05,
"loss": 1.6133,
"step": 690
},
{
"epoch": 1.63,
"learning_rate": 8.372093023255814e-05,
"loss": 1.7155,
"step": 700
},
{
"epoch": 1.63,
"eval_accuracy": 0.6840236686390533,
"eval_loss": 1.7626979351043701,
"eval_runtime": 19.4721,
"eval_samples_per_second": 86.791,
"eval_steps_per_second": 10.887,
"step": 700
},
{
"epoch": 1.65,
"learning_rate": 8.348837209302326e-05,
"loss": 1.6136,
"step": 710
},
{
"epoch": 1.67,
"learning_rate": 8.325581395348837e-05,
"loss": 1.6941,
"step": 720
},
{
"epoch": 1.7,
"learning_rate": 8.30232558139535e-05,
"loss": 1.6759,
"step": 730
},
{
"epoch": 1.72,
"learning_rate": 8.27906976744186e-05,
"loss": 1.6455,
"step": 740
},
{
"epoch": 1.74,
"learning_rate": 8.255813953488373e-05,
"loss": 1.5533,
"step": 750
},
{
"epoch": 1.77,
"learning_rate": 8.232558139534884e-05,
"loss": 1.7024,
"step": 760
},
{
"epoch": 1.79,
"learning_rate": 8.209302325581396e-05,
"loss": 1.5073,
"step": 770
},
{
"epoch": 1.81,
"learning_rate": 8.186046511627907e-05,
"loss": 1.5603,
"step": 780
},
{
"epoch": 1.84,
"learning_rate": 8.162790697674419e-05,
"loss": 1.4995,
"step": 790
},
{
"epoch": 1.86,
"learning_rate": 8.139534883720931e-05,
"loss": 1.4689,
"step": 800
},
{
"epoch": 1.86,
"eval_accuracy": 0.6994082840236686,
"eval_loss": 1.5936682224273682,
"eval_runtime": 19.4583,
"eval_samples_per_second": 86.852,
"eval_steps_per_second": 10.895,
"step": 800
},
{
"epoch": 1.88,
"learning_rate": 8.116279069767442e-05,
"loss": 1.3279,
"step": 810
},
{
"epoch": 1.91,
"learning_rate": 8.093023255813953e-05,
"loss": 1.4141,
"step": 820
},
{
"epoch": 1.93,
"learning_rate": 8.069767441860465e-05,
"loss": 1.3242,
"step": 830
},
{
"epoch": 1.95,
"learning_rate": 8.046511627906977e-05,
"loss": 1.3847,
"step": 840
},
{
"epoch": 1.98,
"learning_rate": 8.023255813953489e-05,
"loss": 1.3513,
"step": 850
},
{
"epoch": 2.0,
"learning_rate": 8e-05,
"loss": 1.2652,
"step": 860
},
{
"epoch": 2.02,
"learning_rate": 7.976744186046512e-05,
"loss": 1.0047,
"step": 870
},
{
"epoch": 2.05,
"learning_rate": 7.953488372093023e-05,
"loss": 1.0588,
"step": 880
},
{
"epoch": 2.07,
"learning_rate": 7.930232558139535e-05,
"loss": 1.0214,
"step": 890
},
{
"epoch": 2.09,
"learning_rate": 7.906976744186047e-05,
"loss": 1.0149,
"step": 900
},
{
"epoch": 2.09,
"eval_accuracy": 0.7497041420118343,
"eval_loss": 1.3167972564697266,
"eval_runtime": 20.0324,
"eval_samples_per_second": 84.363,
"eval_steps_per_second": 10.583,
"step": 900
},
{
"epoch": 2.12,
"learning_rate": 7.883720930232558e-05,
"loss": 1.0151,
"step": 910
},
{
"epoch": 2.14,
"learning_rate": 7.86046511627907e-05,
"loss": 0.9265,
"step": 920
},
{
"epoch": 2.16,
"learning_rate": 7.837209302325582e-05,
"loss": 0.8344,
"step": 930
},
{
"epoch": 2.19,
"learning_rate": 7.813953488372094e-05,
"loss": 0.842,
"step": 940
},
{
"epoch": 2.21,
"learning_rate": 7.790697674418606e-05,
"loss": 0.8307,
"step": 950
},
{
"epoch": 2.23,
"learning_rate": 7.767441860465116e-05,
"loss": 0.8094,
"step": 960
},
{
"epoch": 2.26,
"learning_rate": 7.744186046511628e-05,
"loss": 0.8446,
"step": 970
},
{
"epoch": 2.28,
"learning_rate": 7.723255813953488e-05,
"loss": 0.8674,
"step": 980
},
{
"epoch": 2.3,
"learning_rate": 7.7e-05,
"loss": 0.8391,
"step": 990
},
{
"epoch": 2.33,
"learning_rate": 7.676744186046512e-05,
"loss": 0.8148,
"step": 1000
},
{
"epoch": 2.33,
"eval_accuracy": 0.7615384615384615,
"eval_loss": 1.1630123853683472,
"eval_runtime": 19.1764,
"eval_samples_per_second": 88.129,
"eval_steps_per_second": 11.055,
"step": 1000
},
{
"epoch": 2.35,
"learning_rate": 7.653488372093024e-05,
"loss": 0.8528,
"step": 1010
},
{
"epoch": 2.37,
"learning_rate": 7.630232558139536e-05,
"loss": 0.862,
"step": 1020
},
{
"epoch": 2.4,
"learning_rate": 7.606976744186047e-05,
"loss": 0.7922,
"step": 1030
},
{
"epoch": 2.42,
"learning_rate": 7.583720930232558e-05,
"loss": 0.8446,
"step": 1040
},
{
"epoch": 2.44,
"learning_rate": 7.56046511627907e-05,
"loss": 0.835,
"step": 1050
},
{
"epoch": 2.47,
"learning_rate": 7.537209302325582e-05,
"loss": 0.7732,
"step": 1060
},
{
"epoch": 2.49,
"learning_rate": 7.513953488372093e-05,
"loss": 0.6997,
"step": 1070
},
{
"epoch": 2.51,
"learning_rate": 7.490697674418605e-05,
"loss": 0.5918,
"step": 1080
},
{
"epoch": 2.53,
"learning_rate": 7.467441860465117e-05,
"loss": 0.7436,
"step": 1090
},
{
"epoch": 2.56,
"learning_rate": 7.444186046511629e-05,
"loss": 0.7159,
"step": 1100
},
{
"epoch": 2.56,
"eval_accuracy": 0.7674556213017751,
"eval_loss": 1.0869174003601074,
"eval_runtime": 19.2855,
"eval_samples_per_second": 87.631,
"eval_steps_per_second": 10.993,
"step": 1100
},
{
"epoch": 2.58,
"learning_rate": 7.420930232558139e-05,
"loss": 0.6264,
"step": 1110
},
{
"epoch": 2.6,
"learning_rate": 7.397674418604652e-05,
"loss": 0.8402,
"step": 1120
},
{
"epoch": 2.63,
"learning_rate": 7.374418604651163e-05,
"loss": 0.7405,
"step": 1130
},
{
"epoch": 2.65,
"learning_rate": 7.351162790697675e-05,
"loss": 0.6389,
"step": 1140
},
{
"epoch": 2.67,
"learning_rate": 7.327906976744187e-05,
"loss": 0.6977,
"step": 1150
},
{
"epoch": 2.7,
"learning_rate": 7.304651162790698e-05,
"loss": 0.6999,
"step": 1160
},
{
"epoch": 2.72,
"learning_rate": 7.28139534883721e-05,
"loss": 0.5781,
"step": 1170
},
{
"epoch": 2.74,
"learning_rate": 7.25813953488372e-05,
"loss": 0.6413,
"step": 1180
},
{
"epoch": 2.77,
"learning_rate": 7.234883720930234e-05,
"loss": 0.5828,
"step": 1190
},
{
"epoch": 2.79,
"learning_rate": 7.211627906976744e-05,
"loss": 0.7257,
"step": 1200
},
{
"epoch": 2.79,
"eval_accuracy": 0.7893491124260354,
"eval_loss": 0.9606990218162537,
"eval_runtime": 19.5985,
"eval_samples_per_second": 86.231,
"eval_steps_per_second": 10.817,
"step": 1200
},
{
"epoch": 2.81,
"learning_rate": 7.188372093023256e-05,
"loss": 0.6517,
"step": 1210
},
{
"epoch": 2.84,
"learning_rate": 7.165116279069768e-05,
"loss": 0.603,
"step": 1220
},
{
"epoch": 2.86,
"learning_rate": 7.141860465116278e-05,
"loss": 0.6976,
"step": 1230
},
{
"epoch": 2.88,
"learning_rate": 7.118604651162792e-05,
"loss": 0.5142,
"step": 1240
},
{
"epoch": 2.91,
"learning_rate": 7.095348837209302e-05,
"loss": 0.5571,
"step": 1250
},
{
"epoch": 2.93,
"learning_rate": 7.072093023255815e-05,
"loss": 0.6057,
"step": 1260
},
{
"epoch": 2.95,
"learning_rate": 7.048837209302326e-05,
"loss": 0.5934,
"step": 1270
},
{
"epoch": 2.98,
"learning_rate": 7.025581395348837e-05,
"loss": 0.577,
"step": 1280
},
{
"epoch": 3.0,
"learning_rate": 7.002325581395349e-05,
"loss": 0.5954,
"step": 1290
},
{
"epoch": 3.02,
"learning_rate": 6.97906976744186e-05,
"loss": 0.4171,
"step": 1300
},
{
"epoch": 3.02,
"eval_accuracy": 0.7934911242603551,
"eval_loss": 0.8835369944572449,
"eval_runtime": 19.3276,
"eval_samples_per_second": 87.44,
"eval_steps_per_second": 10.969,
"step": 1300
},
{
"epoch": 3.05,
"learning_rate": 6.955813953488373e-05,
"loss": 0.4346,
"step": 1310
},
{
"epoch": 3.07,
"learning_rate": 6.932558139534883e-05,
"loss": 0.3768,
"step": 1320
},
{
"epoch": 3.09,
"learning_rate": 6.909302325581397e-05,
"loss": 0.3899,
"step": 1330
},
{
"epoch": 3.12,
"learning_rate": 6.886046511627907e-05,
"loss": 0.3805,
"step": 1340
},
{
"epoch": 3.14,
"learning_rate": 6.862790697674419e-05,
"loss": 0.3283,
"step": 1350
},
{
"epoch": 3.16,
"learning_rate": 6.83953488372093e-05,
"loss": 0.3698,
"step": 1360
},
{
"epoch": 3.19,
"learning_rate": 6.816279069767441e-05,
"loss": 0.376,
"step": 1370
},
{
"epoch": 3.21,
"learning_rate": 6.793023255813954e-05,
"loss": 0.3654,
"step": 1380
},
{
"epoch": 3.23,
"learning_rate": 6.769767441860465e-05,
"loss": 0.3163,
"step": 1390
},
{
"epoch": 3.26,
"learning_rate": 6.746511627906978e-05,
"loss": 0.2969,
"step": 1400
},
{
"epoch": 3.26,
"eval_accuracy": 0.8130177514792899,
"eval_loss": 0.8258921504020691,
"eval_runtime": 19.2047,
"eval_samples_per_second": 87.999,
"eval_steps_per_second": 11.039,
"step": 1400
},
{
"epoch": 3.28,
"learning_rate": 6.723255813953488e-05,
"loss": 0.3251,
"step": 1410
},
{
"epoch": 3.3,
"learning_rate": 6.7e-05,
"loss": 0.2872,
"step": 1420
},
{
"epoch": 3.33,
"learning_rate": 6.676744186046512e-05,
"loss": 0.3069,
"step": 1430
},
{
"epoch": 3.35,
"learning_rate": 6.653488372093024e-05,
"loss": 0.3083,
"step": 1440
},
{
"epoch": 3.37,
"learning_rate": 6.630232558139536e-05,
"loss": 0.3003,
"step": 1450
},
{
"epoch": 3.4,
"learning_rate": 6.606976744186046e-05,
"loss": 0.3506,
"step": 1460
},
{
"epoch": 3.42,
"learning_rate": 6.583720930232559e-05,
"loss": 0.2641,
"step": 1470
},
{
"epoch": 3.44,
"learning_rate": 6.56046511627907e-05,
"loss": 0.267,
"step": 1480
},
{
"epoch": 3.47,
"learning_rate": 6.537209302325582e-05,
"loss": 0.2905,
"step": 1490
},
{
"epoch": 3.49,
"learning_rate": 6.513953488372093e-05,
"loss": 0.2405,
"step": 1500
},
{
"epoch": 3.49,
"eval_accuracy": 0.8142011834319527,
"eval_loss": 0.7710559964179993,
"eval_runtime": 19.3379,
"eval_samples_per_second": 87.393,
"eval_steps_per_second": 10.963,
"step": 1500
},
{
"epoch": 3.51,
"learning_rate": 6.490697674418605e-05,
"loss": 0.3025,
"step": 1510
},
{
"epoch": 3.53,
"learning_rate": 6.467441860465117e-05,
"loss": 0.3211,
"step": 1520
},
{
"epoch": 3.56,
"learning_rate": 6.444186046511628e-05,
"loss": 0.3577,
"step": 1530
},
{
"epoch": 3.58,
"learning_rate": 6.42093023255814e-05,
"loss": 0.2338,
"step": 1540
},
{
"epoch": 3.6,
"learning_rate": 6.397674418604651e-05,
"loss": 0.298,
"step": 1550
},
{
"epoch": 3.63,
"learning_rate": 6.374418604651163e-05,
"loss": 0.262,
"step": 1560
},
{
"epoch": 3.65,
"learning_rate": 6.351162790697675e-05,
"loss": 0.2085,
"step": 1570
},
{
"epoch": 3.67,
"learning_rate": 6.327906976744187e-05,
"loss": 0.3226,
"step": 1580
},
{
"epoch": 3.7,
"learning_rate": 6.304651162790698e-05,
"loss": 0.2557,
"step": 1590
},
{
"epoch": 3.72,
"learning_rate": 6.28139534883721e-05,
"loss": 0.2948,
"step": 1600
},
{
"epoch": 3.72,
"eval_accuracy": 0.8112426035502959,
"eval_loss": 0.762892484664917,
"eval_runtime": 19.6106,
"eval_samples_per_second": 86.178,
"eval_steps_per_second": 10.81,
"step": 1600
},
{
"epoch": 3.74,
"learning_rate": 6.258139534883721e-05,
"loss": 0.218,
"step": 1610
},
{
"epoch": 3.77,
"learning_rate": 6.234883720930233e-05,
"loss": 0.2052,
"step": 1620
},
{
"epoch": 3.79,
"learning_rate": 6.211627906976744e-05,
"loss": 0.2579,
"step": 1630
},
{
"epoch": 3.81,
"learning_rate": 6.188372093023256e-05,
"loss": 0.3177,
"step": 1640
},
{
"epoch": 3.84,
"learning_rate": 6.165116279069768e-05,
"loss": 0.2408,
"step": 1650
},
{
"epoch": 3.86,
"learning_rate": 6.14186046511628e-05,
"loss": 0.2035,
"step": 1660
},
{
"epoch": 3.88,
"learning_rate": 6.118604651162792e-05,
"loss": 0.2023,
"step": 1670
},
{
"epoch": 3.91,
"learning_rate": 6.095348837209303e-05,
"loss": 0.1827,
"step": 1680
},
{
"epoch": 3.93,
"learning_rate": 6.072093023255814e-05,
"loss": 0.2693,
"step": 1690
},
{
"epoch": 3.95,
"learning_rate": 6.048837209302326e-05,
"loss": 0.1765,
"step": 1700
},
{
"epoch": 3.95,
"eval_accuracy": 0.8124260355029586,
"eval_loss": 0.7116672992706299,
"eval_runtime": 19.3562,
"eval_samples_per_second": 87.31,
"eval_steps_per_second": 10.953,
"step": 1700
},
{
"epoch": 3.98,
"learning_rate": 6.025581395348837e-05,
"loss": 0.2691,
"step": 1710
},
{
"epoch": 4.0,
"learning_rate": 6.0023255813953494e-05,
"loss": 0.1757,
"step": 1720
},
{
"epoch": 4.02,
"learning_rate": 5.9790697674418605e-05,
"loss": 0.1378,
"step": 1730
},
{
"epoch": 4.05,
"learning_rate": 5.955813953488373e-05,
"loss": 0.1462,
"step": 1740
},
{
"epoch": 4.07,
"learning_rate": 5.932558139534884e-05,
"loss": 0.1216,
"step": 1750
},
{
"epoch": 4.09,
"learning_rate": 5.909302325581396e-05,
"loss": 0.1812,
"step": 1760
},
{
"epoch": 4.12,
"learning_rate": 5.886046511627907e-05,
"loss": 0.152,
"step": 1770
},
{
"epoch": 4.14,
"learning_rate": 5.862790697674418e-05,
"loss": 0.1885,
"step": 1780
},
{
"epoch": 4.16,
"learning_rate": 5.839534883720931e-05,
"loss": 0.1154,
"step": 1790
},
{
"epoch": 4.19,
"learning_rate": 5.816279069767442e-05,
"loss": 0.1603,
"step": 1800
},
{
"epoch": 4.19,
"eval_accuracy": 0.8236686390532545,
"eval_loss": 0.6946221590042114,
"eval_runtime": 19.2754,
"eval_samples_per_second": 87.676,
"eval_steps_per_second": 10.998,
"step": 1800
},
{
"epoch": 4.21,
"learning_rate": 5.7930232558139544e-05,
"loss": 0.0944,
"step": 1810
},
{
"epoch": 4.23,
"learning_rate": 5.7697674418604655e-05,
"loss": 0.1028,
"step": 1820
},
{
"epoch": 4.26,
"learning_rate": 5.7465116279069774e-05,
"loss": 0.1365,
"step": 1830
},
{
"epoch": 4.28,
"learning_rate": 5.7232558139534885e-05,
"loss": 0.0906,
"step": 1840
},
{
"epoch": 4.3,
"learning_rate": 5.6999999999999996e-05,
"loss": 0.108,
"step": 1850
},
{
"epoch": 4.33,
"learning_rate": 5.676744186046512e-05,
"loss": 0.1046,
"step": 1860
},
{
"epoch": 4.35,
"learning_rate": 5.653488372093023e-05,
"loss": 0.1544,
"step": 1870
},
{
"epoch": 4.37,
"learning_rate": 5.630232558139535e-05,
"loss": 0.1031,
"step": 1880
},
{
"epoch": 4.4,
"learning_rate": 5.606976744186046e-05,
"loss": 0.1168,
"step": 1890
},
{
"epoch": 4.42,
"learning_rate": 5.583720930232559e-05,
"loss": 0.0955,
"step": 1900
},
{
"epoch": 4.42,
"eval_accuracy": 0.8349112426035503,
"eval_loss": 0.6597385406494141,
"eval_runtime": 19.4735,
"eval_samples_per_second": 86.784,
"eval_steps_per_second": 10.887,
"step": 1900
},
{
"epoch": 4.44,
"learning_rate": 5.56046511627907e-05,
"loss": 0.1175,
"step": 1910
},
{
"epoch": 4.47,
"learning_rate": 5.5372093023255824e-05,
"loss": 0.1221,
"step": 1920
},
{
"epoch": 4.49,
"learning_rate": 5.5139534883720935e-05,
"loss": 0.0875,
"step": 1930
},
{
"epoch": 4.51,
"learning_rate": 5.4906976744186047e-05,
"loss": 0.1444,
"step": 1940
},
{
"epoch": 4.53,
"learning_rate": 5.4674418604651165e-05,
"loss": 0.0731,
"step": 1950
},
{
"epoch": 4.56,
"learning_rate": 5.4441860465116276e-05,
"loss": 0.1046,
"step": 1960
},
{
"epoch": 4.58,
"learning_rate": 5.42093023255814e-05,
"loss": 0.1043,
"step": 1970
},
{
"epoch": 4.6,
"learning_rate": 5.397674418604651e-05,
"loss": 0.0724,
"step": 1980
},
{
"epoch": 4.63,
"learning_rate": 5.374418604651164e-05,
"loss": 0.1057,
"step": 1990
},
{
"epoch": 4.65,
"learning_rate": 5.351162790697675e-05,
"loss": 0.0769,
"step": 2000
},
{
"epoch": 4.65,
"eval_accuracy": 0.8266272189349112,
"eval_loss": 0.6531164646148682,
"eval_runtime": 19.7979,
"eval_samples_per_second": 85.363,
"eval_steps_per_second": 10.708,
"step": 2000
},
{
"epoch": 4.67,
"learning_rate": 5.327906976744186e-05,
"loss": 0.0888,
"step": 2010
},
{
"epoch": 4.7,
"learning_rate": 5.304651162790698e-05,
"loss": 0.1066,
"step": 2020
},
{
"epoch": 4.72,
"learning_rate": 5.281395348837209e-05,
"loss": 0.1076,
"step": 2030
},
{
"epoch": 4.74,
"learning_rate": 5.2581395348837215e-05,
"loss": 0.0877,
"step": 2040
},
{
"epoch": 4.77,
"learning_rate": 5.2348837209302326e-05,
"loss": 0.0887,
"step": 2050
},
{
"epoch": 4.79,
"learning_rate": 5.211627906976745e-05,
"loss": 0.0901,
"step": 2060
},
{
"epoch": 4.81,
"learning_rate": 5.188372093023256e-05,
"loss": 0.1371,
"step": 2070
},
{
"epoch": 4.84,
"learning_rate": 5.1651162790697674e-05,
"loss": 0.0769,
"step": 2080
},
{
"epoch": 4.86,
"learning_rate": 5.141860465116279e-05,
"loss": 0.1464,
"step": 2090
},
{
"epoch": 4.88,
"learning_rate": 5.1186046511627904e-05,
"loss": 0.0816,
"step": 2100
},
{
"epoch": 4.88,
"eval_accuracy": 0.8337278106508875,
"eval_loss": 0.6334948539733887,
"eval_runtime": 19.2914,
"eval_samples_per_second": 87.604,
"eval_steps_per_second": 10.989,
"step": 2100
},
{
"epoch": 4.91,
"learning_rate": 5.095348837209303e-05,
"loss": 0.0987,
"step": 2110
},
{
"epoch": 4.93,
"learning_rate": 5.072093023255814e-05,
"loss": 0.0898,
"step": 2120
},
{
"epoch": 4.95,
"learning_rate": 5.0488372093023265e-05,
"loss": 0.0964,
"step": 2130
},
{
"epoch": 4.98,
"learning_rate": 5.0255813953488376e-05,
"loss": 0.0683,
"step": 2140
},
{
"epoch": 5.0,
"learning_rate": 5.0023255813953495e-05,
"loss": 0.0871,
"step": 2150
},
{
"epoch": 5.02,
"learning_rate": 4.9790697674418606e-05,
"loss": 0.0679,
"step": 2160
},
{
"epoch": 5.05,
"learning_rate": 4.9558139534883724e-05,
"loss": 0.0463,
"step": 2170
},
{
"epoch": 5.07,
"learning_rate": 4.932558139534884e-05,
"loss": 0.0457,
"step": 2180
},
{
"epoch": 5.09,
"learning_rate": 4.909302325581396e-05,
"loss": 0.034,
"step": 2190
},
{
"epoch": 5.12,
"learning_rate": 4.886046511627907e-05,
"loss": 0.0315,
"step": 2200
},
{
"epoch": 5.12,
"eval_accuracy": 0.8402366863905325,
"eval_loss": 0.608655571937561,
"eval_runtime": 19.634,
"eval_samples_per_second": 86.075,
"eval_steps_per_second": 10.798,
"step": 2200
},
{
"epoch": 5.14,
"learning_rate": 4.8627906976744183e-05,
"loss": 0.0463,
"step": 2210
},
{
"epoch": 5.16,
"learning_rate": 4.83953488372093e-05,
"loss": 0.0547,
"step": 2220
},
{
"epoch": 5.19,
"learning_rate": 4.816279069767442e-05,
"loss": 0.0451,
"step": 2230
},
{
"epoch": 5.21,
"learning_rate": 4.793023255813954e-05,
"loss": 0.0399,
"step": 2240
},
{
"epoch": 5.23,
"learning_rate": 4.7697674418604656e-05,
"loss": 0.0486,
"step": 2250
},
{
"epoch": 5.26,
"learning_rate": 4.7465116279069774e-05,
"loss": 0.0439,
"step": 2260
},
{
"epoch": 5.28,
"learning_rate": 4.7232558139534886e-05,
"loss": 0.034,
"step": 2270
},
{
"epoch": 5.3,
"learning_rate": 4.7e-05,
"loss": 0.0616,
"step": 2280
},
{
"epoch": 5.33,
"learning_rate": 4.6767441860465115e-05,
"loss": 0.0595,
"step": 2290
},
{
"epoch": 5.35,
"learning_rate": 4.6534883720930234e-05,
"loss": 0.0368,
"step": 2300
},
{
"epoch": 5.35,
"eval_accuracy": 0.8443786982248521,
"eval_loss": 0.602620005607605,
"eval_runtime": 19.5598,
"eval_samples_per_second": 86.402,
"eval_steps_per_second": 10.839,
"step": 2300
},
{
"epoch": 5.37,
"learning_rate": 4.630232558139535e-05,
"loss": 0.0597,
"step": 2310
},
{
"epoch": 5.4,
"learning_rate": 4.606976744186047e-05,
"loss": 0.0514,
"step": 2320
},
{
"epoch": 5.42,
"learning_rate": 4.583720930232558e-05,
"loss": 0.0757,
"step": 2330
},
{
"epoch": 5.44,
"learning_rate": 4.56046511627907e-05,
"loss": 0.0421,
"step": 2340
},
{
"epoch": 5.47,
"learning_rate": 4.537209302325582e-05,
"loss": 0.0364,
"step": 2350
},
{
"epoch": 5.49,
"learning_rate": 4.513953488372093e-05,
"loss": 0.0304,
"step": 2360
},
{
"epoch": 5.51,
"learning_rate": 4.490697674418605e-05,
"loss": 0.029,
"step": 2370
},
{
"epoch": 5.53,
"learning_rate": 4.4674418604651166e-05,
"loss": 0.038,
"step": 2380
},
{
"epoch": 5.56,
"learning_rate": 4.4441860465116284e-05,
"loss": 0.0856,
"step": 2390
},
{
"epoch": 5.58,
"learning_rate": 4.4209302325581395e-05,
"loss": 0.0377,
"step": 2400
},
{
"epoch": 5.58,
"eval_accuracy": 0.827810650887574,
"eval_loss": 0.6450419425964355,
"eval_runtime": 19.8289,
"eval_samples_per_second": 85.229,
"eval_steps_per_second": 10.691,
"step": 2400
},
{
"epoch": 5.6,
"learning_rate": 4.397674418604651e-05,
"loss": 0.0394,
"step": 2410
},
{
"epoch": 5.63,
"learning_rate": 4.374418604651163e-05,
"loss": 0.0254,
"step": 2420
},
{
"epoch": 5.65,
"learning_rate": 4.351162790697675e-05,
"loss": 0.0495,
"step": 2430
},
{
"epoch": 5.67,
"learning_rate": 4.327906976744186e-05,
"loss": 0.076,
"step": 2440
},
{
"epoch": 5.7,
"learning_rate": 4.304651162790698e-05,
"loss": 0.0806,
"step": 2450
},
{
"epoch": 5.72,
"learning_rate": 4.281395348837209e-05,
"loss": 0.028,
"step": 2460
},
{
"epoch": 5.74,
"learning_rate": 4.258139534883721e-05,
"loss": 0.0399,
"step": 2470
},
{
"epoch": 5.77,
"learning_rate": 4.234883720930233e-05,
"loss": 0.0455,
"step": 2480
},
{
"epoch": 5.79,
"learning_rate": 4.2116279069767445e-05,
"loss": 0.0437,
"step": 2490
},
{
"epoch": 5.81,
"learning_rate": 4.1883720930232564e-05,
"loss": 0.0603,
"step": 2500
},
{
"epoch": 5.81,
"eval_accuracy": 0.834319526627219,
"eval_loss": 0.6563973426818848,
"eval_runtime": 19.385,
"eval_samples_per_second": 87.181,
"eval_steps_per_second": 10.936,
"step": 2500
},
{
"epoch": 5.84,
"learning_rate": 4.165116279069768e-05,
"loss": 0.0287,
"step": 2510
},
{
"epoch": 5.86,
"learning_rate": 4.141860465116279e-05,
"loss": 0.0249,
"step": 2520
},
{
"epoch": 5.88,
"learning_rate": 4.1186046511627905e-05,
"loss": 0.0349,
"step": 2530
},
{
"epoch": 5.91,
"learning_rate": 4.095348837209302e-05,
"loss": 0.0324,
"step": 2540
},
{
"epoch": 5.93,
"learning_rate": 4.072093023255814e-05,
"loss": 0.0269,
"step": 2550
},
{
"epoch": 5.95,
"learning_rate": 4.048837209302326e-05,
"loss": 0.0226,
"step": 2560
},
{
"epoch": 5.98,
"learning_rate": 4.025581395348838e-05,
"loss": 0.0478,
"step": 2570
},
{
"epoch": 6.0,
"learning_rate": 4.002325581395349e-05,
"loss": 0.0674,
"step": 2580
},
{
"epoch": 6.02,
"learning_rate": 3.979069767441861e-05,
"loss": 0.0285,
"step": 2590
},
{
"epoch": 6.05,
"learning_rate": 3.955813953488372e-05,
"loss": 0.0205,
"step": 2600
},
{
"epoch": 6.05,
"eval_accuracy": 0.8467455621301775,
"eval_loss": 0.61187744140625,
"eval_runtime": 19.5313,
"eval_samples_per_second": 86.528,
"eval_steps_per_second": 10.854,
"step": 2600
},
{
"epoch": 6.07,
"learning_rate": 3.9325581395348837e-05,
"loss": 0.0195,
"step": 2610
},
{
"epoch": 6.09,
"learning_rate": 3.9093023255813955e-05,
"loss": 0.0431,
"step": 2620
},
{
"epoch": 6.12,
"learning_rate": 3.886046511627907e-05,
"loss": 0.0198,
"step": 2630
},
{
"epoch": 6.14,
"learning_rate": 3.862790697674419e-05,
"loss": 0.0194,
"step": 2640
},
{
"epoch": 6.16,
"learning_rate": 3.83953488372093e-05,
"loss": 0.0191,
"step": 2650
},
{
"epoch": 6.19,
"learning_rate": 3.816279069767442e-05,
"loss": 0.0199,
"step": 2660
},
{
"epoch": 6.21,
"learning_rate": 3.793023255813954e-05,
"loss": 0.0185,
"step": 2670
},
{
"epoch": 6.23,
"learning_rate": 3.769767441860465e-05,
"loss": 0.0335,
"step": 2680
},
{
"epoch": 6.26,
"learning_rate": 3.746511627906977e-05,
"loss": 0.0181,
"step": 2690
},
{
"epoch": 6.28,
"learning_rate": 3.723255813953489e-05,
"loss": 0.019,
"step": 2700
},
{
"epoch": 6.28,
"eval_accuracy": 0.8479289940828403,
"eval_loss": 0.6070042252540588,
"eval_runtime": 19.3135,
"eval_samples_per_second": 87.503,
"eval_steps_per_second": 10.977,
"step": 2700
},
{
"epoch": 6.3,
"learning_rate": 3.7e-05,
"loss": 0.0238,
"step": 2710
},
{
"epoch": 6.33,
"learning_rate": 3.6767441860465116e-05,
"loss": 0.0212,
"step": 2720
},
{
"epoch": 6.35,
"learning_rate": 3.6534883720930234e-05,
"loss": 0.0192,
"step": 2730
},
{
"epoch": 6.37,
"learning_rate": 3.630232558139535e-05,
"loss": 0.0187,
"step": 2740
},
{
"epoch": 6.4,
"learning_rate": 3.606976744186047e-05,
"loss": 0.0197,
"step": 2750
},
{
"epoch": 6.42,
"learning_rate": 3.583720930232558e-05,
"loss": 0.0202,
"step": 2760
},
{
"epoch": 6.44,
"learning_rate": 3.56046511627907e-05,
"loss": 0.0168,
"step": 2770
},
{
"epoch": 6.47,
"learning_rate": 3.537209302325581e-05,
"loss": 0.0242,
"step": 2780
},
{
"epoch": 6.49,
"learning_rate": 3.513953488372093e-05,
"loss": 0.0152,
"step": 2790
},
{
"epoch": 6.51,
"learning_rate": 3.490697674418605e-05,
"loss": 0.0249,
"step": 2800
},
{
"epoch": 6.51,
"eval_accuracy": 0.8538461538461538,
"eval_loss": 0.600172758102417,
"eval_runtime": 19.2188,
"eval_samples_per_second": 87.935,
"eval_steps_per_second": 11.031,
"step": 2800
},
{
"epoch": 6.53,
"learning_rate": 3.4674418604651166e-05,
"loss": 0.0173,
"step": 2810
},
{
"epoch": 6.56,
"learning_rate": 3.4441860465116285e-05,
"loss": 0.017,
"step": 2820
},
{
"epoch": 6.58,
"learning_rate": 3.4209302325581396e-05,
"loss": 0.0164,
"step": 2830
},
{
"epoch": 6.6,
"learning_rate": 3.397674418604651e-05,
"loss": 0.0484,
"step": 2840
},
{
"epoch": 6.63,
"learning_rate": 3.3744186046511626e-05,
"loss": 0.017,
"step": 2850
},
{
"epoch": 6.65,
"learning_rate": 3.3511627906976744e-05,
"loss": 0.0158,
"step": 2860
},
{
"epoch": 6.67,
"learning_rate": 3.327906976744186e-05,
"loss": 0.0169,
"step": 2870
},
{
"epoch": 6.7,
"learning_rate": 3.304651162790698e-05,
"loss": 0.0151,
"step": 2880
},
{
"epoch": 6.72,
"learning_rate": 3.28139534883721e-05,
"loss": 0.0161,
"step": 2890
},
{
"epoch": 6.74,
"learning_rate": 3.258139534883721e-05,
"loss": 0.0145,
"step": 2900
},
{
"epoch": 6.74,
"eval_accuracy": 0.8497041420118343,
"eval_loss": 0.601220428943634,
"eval_runtime": 19.2252,
"eval_samples_per_second": 87.906,
"eval_steps_per_second": 11.027,
"step": 2900
},
{
"epoch": 6.77,
"learning_rate": 3.234883720930233e-05,
"loss": 0.0163,
"step": 2910
},
{
"epoch": 6.79,
"learning_rate": 3.211627906976744e-05,
"loss": 0.0308,
"step": 2920
},
{
"epoch": 6.81,
"learning_rate": 3.188372093023256e-05,
"loss": 0.0138,
"step": 2930
},
{
"epoch": 6.84,
"learning_rate": 3.1651162790697676e-05,
"loss": 0.0394,
"step": 2940
},
{
"epoch": 6.86,
"learning_rate": 3.1418604651162794e-05,
"loss": 0.0136,
"step": 2950
},
{
"epoch": 6.88,
"learning_rate": 3.1186046511627905e-05,
"loss": 0.0181,
"step": 2960
},
{
"epoch": 6.91,
"learning_rate": 3.0953488372093024e-05,
"loss": 0.0152,
"step": 2970
},
{
"epoch": 6.93,
"learning_rate": 3.072093023255814e-05,
"loss": 0.0148,
"step": 2980
},
{
"epoch": 6.95,
"learning_rate": 3.048837209302326e-05,
"loss": 0.0138,
"step": 2990
},
{
"epoch": 6.98,
"learning_rate": 3.025581395348837e-05,
"loss": 0.0134,
"step": 3000
},
{
"epoch": 6.98,
"eval_accuracy": 0.8520710059171598,
"eval_loss": 0.599067211151123,
"eval_runtime": 19.2584,
"eval_samples_per_second": 87.754,
"eval_steps_per_second": 11.008,
"step": 3000
},
{
"epoch": 7.0,
"learning_rate": 3.002325581395349e-05,
"loss": 0.0404,
"step": 3010
},
{
"epoch": 7.02,
"learning_rate": 2.9790697674418604e-05,
"loss": 0.013,
"step": 3020
},
{
"epoch": 7.05,
"learning_rate": 2.9558139534883723e-05,
"loss": 0.014,
"step": 3030
},
{
"epoch": 7.07,
"learning_rate": 2.9325581395348837e-05,
"loss": 0.0128,
"step": 3040
},
{
"epoch": 7.09,
"learning_rate": 2.9093023255813956e-05,
"loss": 0.0134,
"step": 3050
},
{
"epoch": 7.12,
"learning_rate": 2.8860465116279074e-05,
"loss": 0.0132,
"step": 3060
},
{
"epoch": 7.14,
"learning_rate": 2.862790697674419e-05,
"loss": 0.0135,
"step": 3070
},
{
"epoch": 7.16,
"learning_rate": 2.83953488372093e-05,
"loss": 0.0128,
"step": 3080
},
{
"epoch": 7.19,
"learning_rate": 2.8162790697674418e-05,
"loss": 0.0279,
"step": 3090
},
{
"epoch": 7.21,
"learning_rate": 2.7930232558139536e-05,
"loss": 0.0271,
"step": 3100
},
{
"epoch": 7.21,
"eval_accuracy": 0.8502958579881656,
"eval_loss": 0.5971536636352539,
"eval_runtime": 19.4411,
"eval_samples_per_second": 86.929,
"eval_steps_per_second": 10.905,
"step": 3100
},
{
"epoch": 7.23,
"learning_rate": 2.769767441860465e-05,
"loss": 0.0219,
"step": 3110
},
{
"epoch": 7.26,
"learning_rate": 2.746511627906977e-05,
"loss": 0.0225,
"step": 3120
},
{
"epoch": 7.28,
"learning_rate": 2.7232558139534888e-05,
"loss": 0.0147,
"step": 3130
},
{
"epoch": 7.3,
"learning_rate": 2.7000000000000002e-05,
"loss": 0.0123,
"step": 3140
},
{
"epoch": 7.33,
"learning_rate": 2.676744186046512e-05,
"loss": 0.0145,
"step": 3150
},
{
"epoch": 7.35,
"learning_rate": 2.6534883720930232e-05,
"loss": 0.0121,
"step": 3160
},
{
"epoch": 7.37,
"learning_rate": 2.6302325581395347e-05,
"loss": 0.0123,
"step": 3170
},
{
"epoch": 7.4,
"learning_rate": 2.6069767441860465e-05,
"loss": 0.0106,
"step": 3180
},
{
"epoch": 7.42,
"learning_rate": 2.5837209302325583e-05,
"loss": 0.0124,
"step": 3190
},
{
"epoch": 7.44,
"learning_rate": 2.5604651162790698e-05,
"loss": 0.0128,
"step": 3200
},
{
"epoch": 7.44,
"eval_accuracy": 0.8520710059171598,
"eval_loss": 0.5910876989364624,
"eval_runtime": 19.7261,
"eval_samples_per_second": 85.673,
"eval_steps_per_second": 10.747,
"step": 3200
},
{
"epoch": 7.47,
"learning_rate": 2.5372093023255816e-05,
"loss": 0.0117,
"step": 3210
},
{
"epoch": 7.49,
"learning_rate": 2.5139534883720934e-05,
"loss": 0.0117,
"step": 3220
},
{
"epoch": 7.51,
"learning_rate": 2.4906976744186046e-05,
"loss": 0.0117,
"step": 3230
},
{
"epoch": 7.53,
"learning_rate": 2.4674418604651164e-05,
"loss": 0.0106,
"step": 3240
},
{
"epoch": 7.56,
"learning_rate": 2.4441860465116282e-05,
"loss": 0.0107,
"step": 3250
},
{
"epoch": 7.58,
"learning_rate": 2.4209302325581397e-05,
"loss": 0.0119,
"step": 3260
},
{
"epoch": 7.6,
"learning_rate": 2.3976744186046512e-05,
"loss": 0.0123,
"step": 3270
},
{
"epoch": 7.63,
"learning_rate": 2.374418604651163e-05,
"loss": 0.0123,
"step": 3280
},
{
"epoch": 7.65,
"learning_rate": 2.3511627906976748e-05,
"loss": 0.0105,
"step": 3290
},
{
"epoch": 7.67,
"learning_rate": 2.327906976744186e-05,
"loss": 0.0123,
"step": 3300
},
{
"epoch": 7.67,
"eval_accuracy": 0.8538461538461538,
"eval_loss": 0.5888754725456238,
"eval_runtime": 20.1627,
"eval_samples_per_second": 83.818,
"eval_steps_per_second": 10.514,
"step": 3300
},
{
"epoch": 7.7,
"learning_rate": 2.3046511627906978e-05,
"loss": 0.011,
"step": 3310
},
{
"epoch": 7.72,
"learning_rate": 2.2813953488372096e-05,
"loss": 0.0115,
"step": 3320
},
{
"epoch": 7.74,
"learning_rate": 2.258139534883721e-05,
"loss": 0.0108,
"step": 3330
},
{
"epoch": 7.77,
"learning_rate": 2.2348837209302326e-05,
"loss": 0.0111,
"step": 3340
},
{
"epoch": 7.79,
"learning_rate": 2.2116279069767444e-05,
"loss": 0.011,
"step": 3350
},
{
"epoch": 7.81,
"learning_rate": 2.188372093023256e-05,
"loss": 0.0104,
"step": 3360
},
{
"epoch": 7.84,
"learning_rate": 2.1651162790697677e-05,
"loss": 0.01,
"step": 3370
},
{
"epoch": 7.86,
"learning_rate": 2.141860465116279e-05,
"loss": 0.0136,
"step": 3380
},
{
"epoch": 7.88,
"learning_rate": 2.1186046511627906e-05,
"loss": 0.0115,
"step": 3390
},
{
"epoch": 7.91,
"learning_rate": 2.0953488372093025e-05,
"loss": 0.0278,
"step": 3400
},
{
"epoch": 7.91,
"eval_accuracy": 0.849112426035503,
"eval_loss": 0.6134743094444275,
"eval_runtime": 19.6151,
"eval_samples_per_second": 86.158,
"eval_steps_per_second": 10.808,
"step": 3400
},
{
"epoch": 7.93,
"learning_rate": 2.0720930232558143e-05,
"loss": 0.0174,
"step": 3410
},
{
"epoch": 7.95,
"learning_rate": 2.0488372093023258e-05,
"loss": 0.011,
"step": 3420
},
{
"epoch": 7.98,
"learning_rate": 2.0255813953488372e-05,
"loss": 0.0118,
"step": 3430
},
{
"epoch": 8.0,
"learning_rate": 2.002325581395349e-05,
"loss": 0.0107,
"step": 3440
},
{
"epoch": 8.02,
"learning_rate": 1.9790697674418605e-05,
"loss": 0.0104,
"step": 3450
},
{
"epoch": 8.05,
"learning_rate": 1.955813953488372e-05,
"loss": 0.023,
"step": 3460
},
{
"epoch": 8.07,
"learning_rate": 1.9325581395348838e-05,
"loss": 0.0107,
"step": 3470
},
{
"epoch": 8.09,
"learning_rate": 1.9093023255813956e-05,
"loss": 0.0107,
"step": 3480
},
{
"epoch": 8.12,
"learning_rate": 1.886046511627907e-05,
"loss": 0.0103,
"step": 3490
},
{
"epoch": 8.14,
"learning_rate": 1.8627906976744186e-05,
"loss": 0.0106,
"step": 3500
},
{
"epoch": 8.14,
"eval_accuracy": 0.8532544378698225,
"eval_loss": 0.5933935642242432,
"eval_runtime": 19.5044,
"eval_samples_per_second": 86.647,
"eval_steps_per_second": 10.869,
"step": 3500
},
{
"epoch": 8.16,
"learning_rate": 1.8395348837209304e-05,
"loss": 0.0105,
"step": 3510
},
{
"epoch": 8.19,
"learning_rate": 1.816279069767442e-05,
"loss": 0.0095,
"step": 3520
},
{
"epoch": 8.21,
"learning_rate": 1.7930232558139537e-05,
"loss": 0.0097,
"step": 3530
},
{
"epoch": 8.23,
"learning_rate": 1.7697674418604652e-05,
"loss": 0.0089,
"step": 3540
},
{
"epoch": 8.26,
"learning_rate": 1.7465116279069767e-05,
"loss": 0.0094,
"step": 3550
},
{
"epoch": 8.28,
"learning_rate": 1.7232558139534885e-05,
"loss": 0.01,
"step": 3560
},
{
"epoch": 8.3,
"learning_rate": 1.7000000000000003e-05,
"loss": 0.0094,
"step": 3570
},
{
"epoch": 8.33,
"learning_rate": 1.6767441860465115e-05,
"loss": 0.0097,
"step": 3580
},
{
"epoch": 8.35,
"learning_rate": 1.6534883720930233e-05,
"loss": 0.0091,
"step": 3590
},
{
"epoch": 8.37,
"learning_rate": 1.630232558139535e-05,
"loss": 0.0109,
"step": 3600
},
{
"epoch": 8.37,
"eval_accuracy": 0.8532544378698225,
"eval_loss": 0.5928947925567627,
"eval_runtime": 19.301,
"eval_samples_per_second": 87.56,
"eval_steps_per_second": 10.984,
"step": 3600
},
{
"epoch": 8.4,
"learning_rate": 1.6069767441860466e-05,
"loss": 0.0102,
"step": 3610
},
{
"epoch": 8.42,
"learning_rate": 1.583720930232558e-05,
"loss": 0.0095,
"step": 3620
},
{
"epoch": 8.44,
"learning_rate": 1.56046511627907e-05,
"loss": 0.0096,
"step": 3630
},
{
"epoch": 8.47,
"learning_rate": 1.5372093023255814e-05,
"loss": 0.01,
"step": 3640
},
{
"epoch": 8.49,
"learning_rate": 1.5139534883720932e-05,
"loss": 0.0094,
"step": 3650
},
{
"epoch": 8.51,
"learning_rate": 1.4906976744186047e-05,
"loss": 0.0097,
"step": 3660
},
{
"epoch": 8.53,
"learning_rate": 1.4674418604651163e-05,
"loss": 0.0121,
"step": 3670
},
{
"epoch": 8.56,
"learning_rate": 1.444186046511628e-05,
"loss": 0.0102,
"step": 3680
},
{
"epoch": 8.58,
"learning_rate": 1.4209302325581398e-05,
"loss": 0.0099,
"step": 3690
},
{
"epoch": 8.6,
"learning_rate": 1.3976744186046511e-05,
"loss": 0.0095,
"step": 3700
},
{
"epoch": 8.6,
"eval_accuracy": 0.8550295857988166,
"eval_loss": 0.5952585935592651,
"eval_runtime": 19.2704,
"eval_samples_per_second": 87.699,
"eval_steps_per_second": 11.001,
"step": 3700
},
{
"epoch": 8.63,
"learning_rate": 1.3744186046511629e-05,
"loss": 0.0098,
"step": 3710
},
{
"epoch": 8.65,
"learning_rate": 1.3511627906976746e-05,
"loss": 0.0104,
"step": 3720
},
{
"epoch": 8.67,
"learning_rate": 1.3279069767441862e-05,
"loss": 0.009,
"step": 3730
},
{
"epoch": 8.7,
"learning_rate": 1.3046511627906977e-05,
"loss": 0.0092,
"step": 3740
},
{
"epoch": 8.72,
"learning_rate": 1.2813953488372093e-05,
"loss": 0.0089,
"step": 3750
},
{
"epoch": 8.74,
"learning_rate": 1.258139534883721e-05,
"loss": 0.009,
"step": 3760
},
{
"epoch": 8.77,
"learning_rate": 1.2348837209302326e-05,
"loss": 0.0098,
"step": 3770
},
{
"epoch": 8.79,
"learning_rate": 1.2116279069767443e-05,
"loss": 0.0086,
"step": 3780
},
{
"epoch": 8.81,
"learning_rate": 1.188372093023256e-05,
"loss": 0.023,
"step": 3790
},
{
"epoch": 8.84,
"learning_rate": 1.1651162790697676e-05,
"loss": 0.009,
"step": 3800
},
{
"epoch": 8.84,
"eval_accuracy": 0.857396449704142,
"eval_loss": 0.5933314561843872,
"eval_runtime": 19.4586,
"eval_samples_per_second": 86.851,
"eval_steps_per_second": 10.895,
"step": 3800
},
{
"epoch": 8.86,
"learning_rate": 1.141860465116279e-05,
"loss": 0.0089,
"step": 3810
},
{
"epoch": 8.88,
"learning_rate": 1.1186046511627909e-05,
"loss": 0.0089,
"step": 3820
},
{
"epoch": 8.91,
"learning_rate": 1.0953488372093024e-05,
"loss": 0.0097,
"step": 3830
},
{
"epoch": 8.93,
"learning_rate": 1.072093023255814e-05,
"loss": 0.0092,
"step": 3840
},
{
"epoch": 8.95,
"learning_rate": 1.0488372093023257e-05,
"loss": 0.0099,
"step": 3850
},
{
"epoch": 8.98,
"learning_rate": 1.0255813953488373e-05,
"loss": 0.0099,
"step": 3860
},
{
"epoch": 9.0,
"learning_rate": 1.0023255813953488e-05,
"loss": 0.0089,
"step": 3870
},
{
"epoch": 9.02,
"learning_rate": 9.790697674418606e-06,
"loss": 0.0091,
"step": 3880
},
{
"epoch": 9.05,
"learning_rate": 9.558139534883721e-06,
"loss": 0.0115,
"step": 3890
},
{
"epoch": 9.07,
"learning_rate": 9.325581395348837e-06,
"loss": 0.009,
"step": 3900
},
{
"epoch": 9.07,
"eval_accuracy": 0.8550295857988166,
"eval_loss": 0.5948178768157959,
"eval_runtime": 19.3608,
"eval_samples_per_second": 87.29,
"eval_steps_per_second": 10.95,
"step": 3900
},
{
"epoch": 9.09,
"learning_rate": 9.093023255813954e-06,
"loss": 0.0088,
"step": 3910
},
{
"epoch": 9.12,
"learning_rate": 8.86046511627907e-06,
"loss": 0.0085,
"step": 3920
},
{
"epoch": 9.14,
"learning_rate": 8.627906976744187e-06,
"loss": 0.0088,
"step": 3930
},
{
"epoch": 9.16,
"learning_rate": 8.395348837209303e-06,
"loss": 0.0084,
"step": 3940
},
{
"epoch": 9.19,
"learning_rate": 8.162790697674418e-06,
"loss": 0.0085,
"step": 3950
},
{
"epoch": 9.21,
"learning_rate": 7.930232558139536e-06,
"loss": 0.0082,
"step": 3960
},
{
"epoch": 9.23,
"learning_rate": 7.697674418604651e-06,
"loss": 0.0088,
"step": 3970
},
{
"epoch": 9.26,
"learning_rate": 7.465116279069769e-06,
"loss": 0.0085,
"step": 3980
},
{
"epoch": 9.28,
"learning_rate": 7.232558139534884e-06,
"loss": 0.0085,
"step": 3990
},
{
"epoch": 9.3,
"learning_rate": 7.000000000000001e-06,
"loss": 0.0089,
"step": 4000
},
{
"epoch": 9.3,
"eval_accuracy": 0.855621301775148,
"eval_loss": 0.5952693223953247,
"eval_runtime": 19.9361,
"eval_samples_per_second": 84.771,
"eval_steps_per_second": 10.634,
"step": 4000
},
{
"epoch": 9.33,
"learning_rate": 6.767441860465116e-06,
"loss": 0.0091,
"step": 4010
},
{
"epoch": 9.35,
"learning_rate": 6.534883720930234e-06,
"loss": 0.0086,
"step": 4020
},
{
"epoch": 9.37,
"learning_rate": 6.302325581395349e-06,
"loss": 0.0091,
"step": 4030
},
{
"epoch": 9.4,
"learning_rate": 6.069767441860465e-06,
"loss": 0.0088,
"step": 4040
},
{
"epoch": 9.42,
"learning_rate": 5.8372093023255815e-06,
"loss": 0.0088,
"step": 4050
},
{
"epoch": 9.44,
"learning_rate": 5.604651162790698e-06,
"loss": 0.0085,
"step": 4060
},
{
"epoch": 9.47,
"learning_rate": 5.3720930232558145e-06,
"loss": 0.0087,
"step": 4070
},
{
"epoch": 9.49,
"learning_rate": 5.13953488372093e-06,
"loss": 0.0079,
"step": 4080
},
{
"epoch": 9.51,
"learning_rate": 4.906976744186047e-06,
"loss": 0.0085,
"step": 4090
},
{
"epoch": 9.53,
"learning_rate": 4.674418604651163e-06,
"loss": 0.0086,
"step": 4100
},
{
"epoch": 9.53,
"eval_accuracy": 0.8544378698224852,
"eval_loss": 0.5956243276596069,
"eval_runtime": 19.6785,
"eval_samples_per_second": 85.881,
"eval_steps_per_second": 10.773,
"step": 4100
},
{
"epoch": 9.56,
"learning_rate": 4.44186046511628e-06,
"loss": 0.0079,
"step": 4110
},
{
"epoch": 9.58,
"learning_rate": 4.209302325581395e-06,
"loss": 0.0082,
"step": 4120
},
{
"epoch": 9.6,
"learning_rate": 3.976744186046512e-06,
"loss": 0.0089,
"step": 4130
},
{
"epoch": 9.63,
"learning_rate": 3.7441860465116283e-06,
"loss": 0.0082,
"step": 4140
},
{
"epoch": 9.65,
"learning_rate": 3.5116279069767444e-06,
"loss": 0.0089,
"step": 4150
},
{
"epoch": 9.67,
"learning_rate": 3.2790697674418604e-06,
"loss": 0.0084,
"step": 4160
},
{
"epoch": 9.7,
"learning_rate": 3.046511627906977e-06,
"loss": 0.0085,
"step": 4170
},
{
"epoch": 9.72,
"learning_rate": 2.813953488372093e-06,
"loss": 0.0193,
"step": 4180
},
{
"epoch": 9.74,
"learning_rate": 2.5813953488372095e-06,
"loss": 0.0082,
"step": 4190
},
{
"epoch": 9.77,
"learning_rate": 2.3488372093023256e-06,
"loss": 0.0085,
"step": 4200
},
{
"epoch": 9.77,
"eval_accuracy": 0.855621301775148,
"eval_loss": 0.5954723954200745,
"eval_runtime": 20.1652,
"eval_samples_per_second": 83.808,
"eval_steps_per_second": 10.513,
"step": 4200
},
{
"epoch": 9.79,
"learning_rate": 2.116279069767442e-06,
"loss": 0.0085,
"step": 4210
},
{
"epoch": 9.81,
"learning_rate": 1.8837209302325584e-06,
"loss": 0.0085,
"step": 4220
},
{
"epoch": 9.84,
"learning_rate": 1.6511627906976744e-06,
"loss": 0.0096,
"step": 4230
},
{
"epoch": 9.86,
"learning_rate": 1.4186046511627907e-06,
"loss": 0.0082,
"step": 4240
},
{
"epoch": 9.88,
"learning_rate": 1.186046511627907e-06,
"loss": 0.0208,
"step": 4250
},
{
"epoch": 9.91,
"learning_rate": 9.534883720930234e-07,
"loss": 0.008,
"step": 4260
},
{
"epoch": 9.93,
"learning_rate": 7.209302325581396e-07,
"loss": 0.0093,
"step": 4270
},
{
"epoch": 9.95,
"learning_rate": 4.883720930232559e-07,
"loss": 0.008,
"step": 4280
},
{
"epoch": 9.98,
"learning_rate": 2.558139534883721e-07,
"loss": 0.0085,
"step": 4290
},
{
"epoch": 10.0,
"learning_rate": 2.3255813953488376e-08,
"loss": 0.0087,
"step": 4300
},
{
"epoch": 10.0,
"eval_accuracy": 0.8538461538461538,
"eval_loss": 0.5954486727714539,
"eval_runtime": 19.7825,
"eval_samples_per_second": 85.429,
"eval_steps_per_second": 10.717,
"step": 4300
},
{
"epoch": 10.0,
"step": 4300,
"total_flos": 5.338473146715341e+18,
"train_loss": 0.6670613629041716,
"train_runtime": 2574.0231,
"train_samples_per_second": 26.729,
"train_steps_per_second": 1.671
}
],
"max_steps": 4300,
"num_train_epochs": 10,
"total_flos": 5.338473146715341e+18,
"trial_name": null,
"trial_params": null
}