siddharth963's picture
End of training
6d01da9
raw
history blame
21 kB
{
"best_metric": 0.8705607476635514,
"best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-cassava/checkpoint-1200",
"epoch": 9.996677740863788,
"global_step": 1500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.5923,
"step": 10
},
{
"epoch": 0.13,
"learning_rate": 6.666666666666667e-06,
"loss": 1.546,
"step": 20
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 1.425,
"step": 30
},
{
"epoch": 0.27,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.2868,
"step": 40
},
{
"epoch": 0.33,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.0968,
"step": 50
},
{
"epoch": 0.4,
"learning_rate": 2e-05,
"loss": 1.0634,
"step": 60
},
{
"epoch": 0.47,
"learning_rate": 2.3333333333333336e-05,
"loss": 0.9365,
"step": 70
},
{
"epoch": 0.53,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.854,
"step": 80
},
{
"epoch": 0.6,
"learning_rate": 3e-05,
"loss": 0.7884,
"step": 90
},
{
"epoch": 0.66,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.7399,
"step": 100
},
{
"epoch": 0.73,
"learning_rate": 3.6666666666666666e-05,
"loss": 0.6745,
"step": 110
},
{
"epoch": 0.8,
"learning_rate": 4e-05,
"loss": 0.6399,
"step": 120
},
{
"epoch": 0.86,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.6273,
"step": 130
},
{
"epoch": 0.93,
"learning_rate": 4.666666666666667e-05,
"loss": 0.5644,
"step": 140
},
{
"epoch": 1.0,
"learning_rate": 5e-05,
"loss": 0.5628,
"step": 150
},
{
"epoch": 1.0,
"eval_accuracy": 0.8308411214953271,
"eval_loss": 0.5357394218444824,
"eval_runtime": 44.4006,
"eval_samples_per_second": 48.198,
"eval_steps_per_second": 1.509,
"step": 150
},
{
"epoch": 1.07,
"learning_rate": 4.962962962962963e-05,
"loss": 0.5665,
"step": 160
},
{
"epoch": 1.13,
"learning_rate": 4.925925925925926e-05,
"loss": 0.5119,
"step": 170
},
{
"epoch": 1.2,
"learning_rate": 4.888888888888889e-05,
"loss": 0.4606,
"step": 180
},
{
"epoch": 1.27,
"learning_rate": 4.851851851851852e-05,
"loss": 0.5033,
"step": 190
},
{
"epoch": 1.33,
"learning_rate": 4.814814814814815e-05,
"loss": 0.4933,
"step": 200
},
{
"epoch": 1.4,
"learning_rate": 4.7777777777777784e-05,
"loss": 0.4688,
"step": 210
},
{
"epoch": 1.47,
"learning_rate": 4.740740740740741e-05,
"loss": 0.514,
"step": 220
},
{
"epoch": 1.53,
"learning_rate": 4.703703703703704e-05,
"loss": 0.4439,
"step": 230
},
{
"epoch": 1.6,
"learning_rate": 4.666666666666667e-05,
"loss": 0.5142,
"step": 240
},
{
"epoch": 1.66,
"learning_rate": 4.62962962962963e-05,
"loss": 0.4206,
"step": 250
},
{
"epoch": 1.73,
"learning_rate": 4.592592592592593e-05,
"loss": 0.4254,
"step": 260
},
{
"epoch": 1.8,
"learning_rate": 4.555555555555556e-05,
"loss": 0.4284,
"step": 270
},
{
"epoch": 1.86,
"learning_rate": 4.518518518518519e-05,
"loss": 0.4446,
"step": 280
},
{
"epoch": 1.93,
"learning_rate": 4.481481481481482e-05,
"loss": 0.422,
"step": 290
},
{
"epoch": 2.0,
"learning_rate": 4.4444444444444447e-05,
"loss": 0.4398,
"step": 300
},
{
"epoch": 2.0,
"eval_accuracy": 0.8598130841121495,
"eval_loss": 0.43114611506462097,
"eval_runtime": 46.3574,
"eval_samples_per_second": 46.163,
"eval_steps_per_second": 1.445,
"step": 300
},
{
"epoch": 2.07,
"learning_rate": 4.4074074074074076e-05,
"loss": 0.4463,
"step": 310
},
{
"epoch": 2.13,
"learning_rate": 4.3703703703703705e-05,
"loss": 0.3933,
"step": 320
},
{
"epoch": 2.2,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.4323,
"step": 330
},
{
"epoch": 2.27,
"learning_rate": 4.296296296296296e-05,
"loss": 0.4332,
"step": 340
},
{
"epoch": 2.33,
"learning_rate": 4.259259259259259e-05,
"loss": 0.4394,
"step": 350
},
{
"epoch": 2.4,
"learning_rate": 4.222222222222222e-05,
"loss": 0.4084,
"step": 360
},
{
"epoch": 2.47,
"learning_rate": 4.185185185185185e-05,
"loss": 0.3772,
"step": 370
},
{
"epoch": 2.53,
"learning_rate": 4.148148148148148e-05,
"loss": 0.4324,
"step": 380
},
{
"epoch": 2.6,
"learning_rate": 4.111111111111111e-05,
"loss": 0.3999,
"step": 390
},
{
"epoch": 2.66,
"learning_rate": 4.074074074074074e-05,
"loss": 0.425,
"step": 400
},
{
"epoch": 2.73,
"learning_rate": 4.0370370370370374e-05,
"loss": 0.42,
"step": 410
},
{
"epoch": 2.8,
"learning_rate": 4e-05,
"loss": 0.431,
"step": 420
},
{
"epoch": 2.86,
"learning_rate": 3.962962962962963e-05,
"loss": 0.3784,
"step": 430
},
{
"epoch": 2.93,
"learning_rate": 3.925925925925926e-05,
"loss": 0.3684,
"step": 440
},
{
"epoch": 3.0,
"learning_rate": 3.888888888888889e-05,
"loss": 0.4022,
"step": 450
},
{
"epoch": 3.0,
"eval_accuracy": 0.866822429906542,
"eval_loss": 0.39579352736473083,
"eval_runtime": 44.3798,
"eval_samples_per_second": 48.22,
"eval_steps_per_second": 1.51,
"step": 450
},
{
"epoch": 3.07,
"learning_rate": 3.851851851851852e-05,
"loss": 0.4017,
"step": 460
},
{
"epoch": 3.13,
"learning_rate": 3.814814814814815e-05,
"loss": 0.4008,
"step": 470
},
{
"epoch": 3.2,
"learning_rate": 3.777777777777778e-05,
"loss": 0.3794,
"step": 480
},
{
"epoch": 3.27,
"learning_rate": 3.740740740740741e-05,
"loss": 0.3789,
"step": 490
},
{
"epoch": 3.33,
"learning_rate": 3.7037037037037037e-05,
"loss": 0.3759,
"step": 500
},
{
"epoch": 3.4,
"learning_rate": 3.6666666666666666e-05,
"loss": 0.3652,
"step": 510
},
{
"epoch": 3.47,
"learning_rate": 3.62962962962963e-05,
"loss": 0.3496,
"step": 520
},
{
"epoch": 3.53,
"learning_rate": 3.592592592592593e-05,
"loss": 0.3565,
"step": 530
},
{
"epoch": 3.6,
"learning_rate": 3.555555555555556e-05,
"loss": 0.3943,
"step": 540
},
{
"epoch": 3.66,
"learning_rate": 3.518518518518519e-05,
"loss": 0.3959,
"step": 550
},
{
"epoch": 3.73,
"learning_rate": 3.481481481481482e-05,
"loss": 0.3871,
"step": 560
},
{
"epoch": 3.8,
"learning_rate": 3.444444444444445e-05,
"loss": 0.3636,
"step": 570
},
{
"epoch": 3.86,
"learning_rate": 3.4074074074074077e-05,
"loss": 0.3487,
"step": 580
},
{
"epoch": 3.93,
"learning_rate": 3.3703703703703706e-05,
"loss": 0.4183,
"step": 590
},
{
"epoch": 4.0,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.3855,
"step": 600
},
{
"epoch": 4.0,
"eval_accuracy": 0.8598130841121495,
"eval_loss": 0.4029967784881592,
"eval_runtime": 42.4817,
"eval_samples_per_second": 50.375,
"eval_steps_per_second": 1.577,
"step": 600
},
{
"epoch": 4.07,
"learning_rate": 3.2962962962962964e-05,
"loss": 0.3929,
"step": 610
},
{
"epoch": 4.13,
"learning_rate": 3.25925925925926e-05,
"loss": 0.3455,
"step": 620
},
{
"epoch": 4.2,
"learning_rate": 3.222222222222223e-05,
"loss": 0.351,
"step": 630
},
{
"epoch": 4.27,
"learning_rate": 3.185185185185185e-05,
"loss": 0.3597,
"step": 640
},
{
"epoch": 4.33,
"learning_rate": 3.148148148148148e-05,
"loss": 0.3524,
"step": 650
},
{
"epoch": 4.4,
"learning_rate": 3.111111111111111e-05,
"loss": 0.3514,
"step": 660
},
{
"epoch": 4.47,
"learning_rate": 3.074074074074074e-05,
"loss": 0.3362,
"step": 670
},
{
"epoch": 4.53,
"learning_rate": 3.037037037037037e-05,
"loss": 0.3549,
"step": 680
},
{
"epoch": 4.6,
"learning_rate": 3e-05,
"loss": 0.3175,
"step": 690
},
{
"epoch": 4.66,
"learning_rate": 2.962962962962963e-05,
"loss": 0.3253,
"step": 700
},
{
"epoch": 4.73,
"learning_rate": 2.925925925925926e-05,
"loss": 0.3367,
"step": 710
},
{
"epoch": 4.8,
"learning_rate": 2.8888888888888888e-05,
"loss": 0.3571,
"step": 720
},
{
"epoch": 4.86,
"learning_rate": 2.851851851851852e-05,
"loss": 0.3256,
"step": 730
},
{
"epoch": 4.93,
"learning_rate": 2.814814814814815e-05,
"loss": 0.3747,
"step": 740
},
{
"epoch": 5.0,
"learning_rate": 2.777777777777778e-05,
"loss": 0.3659,
"step": 750
},
{
"epoch": 5.0,
"eval_accuracy": 0.8616822429906542,
"eval_loss": 0.4124689996242523,
"eval_runtime": 43.3097,
"eval_samples_per_second": 49.412,
"eval_steps_per_second": 1.547,
"step": 750
},
{
"epoch": 5.07,
"learning_rate": 2.7407407407407408e-05,
"loss": 0.3325,
"step": 760
},
{
"epoch": 5.13,
"learning_rate": 2.7037037037037037e-05,
"loss": 0.3088,
"step": 770
},
{
"epoch": 5.2,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.3233,
"step": 780
},
{
"epoch": 5.27,
"learning_rate": 2.6296296296296296e-05,
"loss": 0.2802,
"step": 790
},
{
"epoch": 5.33,
"learning_rate": 2.5925925925925925e-05,
"loss": 0.3253,
"step": 800
},
{
"epoch": 5.4,
"learning_rate": 2.5555555555555554e-05,
"loss": 0.3135,
"step": 810
},
{
"epoch": 5.47,
"learning_rate": 2.5185185185185183e-05,
"loss": 0.3415,
"step": 820
},
{
"epoch": 5.53,
"learning_rate": 2.4814814814814816e-05,
"loss": 0.2925,
"step": 830
},
{
"epoch": 5.6,
"learning_rate": 2.4444444444444445e-05,
"loss": 0.3586,
"step": 840
},
{
"epoch": 5.66,
"learning_rate": 2.4074074074074074e-05,
"loss": 0.3392,
"step": 850
},
{
"epoch": 5.73,
"learning_rate": 2.3703703703703707e-05,
"loss": 0.3423,
"step": 860
},
{
"epoch": 5.8,
"learning_rate": 2.3333333333333336e-05,
"loss": 0.3371,
"step": 870
},
{
"epoch": 5.86,
"learning_rate": 2.2962962962962965e-05,
"loss": 0.3295,
"step": 880
},
{
"epoch": 5.93,
"learning_rate": 2.2592592592592594e-05,
"loss": 0.345,
"step": 890
},
{
"epoch": 6.0,
"learning_rate": 2.2222222222222223e-05,
"loss": 0.3393,
"step": 900
},
{
"epoch": 6.0,
"eval_accuracy": 0.8672897196261682,
"eval_loss": 0.3839728534221649,
"eval_runtime": 43.7964,
"eval_samples_per_second": 48.862,
"eval_steps_per_second": 1.53,
"step": 900
},
{
"epoch": 6.07,
"learning_rate": 2.1851851851851852e-05,
"loss": 0.3399,
"step": 910
},
{
"epoch": 6.13,
"learning_rate": 2.148148148148148e-05,
"loss": 0.3209,
"step": 920
},
{
"epoch": 6.2,
"learning_rate": 2.111111111111111e-05,
"loss": 0.3055,
"step": 930
},
{
"epoch": 6.27,
"learning_rate": 2.074074074074074e-05,
"loss": 0.2946,
"step": 940
},
{
"epoch": 6.33,
"learning_rate": 2.037037037037037e-05,
"loss": 0.3265,
"step": 950
},
{
"epoch": 6.4,
"learning_rate": 2e-05,
"loss": 0.3245,
"step": 960
},
{
"epoch": 6.47,
"learning_rate": 1.962962962962963e-05,
"loss": 0.2877,
"step": 970
},
{
"epoch": 6.53,
"learning_rate": 1.925925925925926e-05,
"loss": 0.3212,
"step": 980
},
{
"epoch": 6.6,
"learning_rate": 1.888888888888889e-05,
"loss": 0.2927,
"step": 990
},
{
"epoch": 6.66,
"learning_rate": 1.8518518518518518e-05,
"loss": 0.3031,
"step": 1000
},
{
"epoch": 6.73,
"learning_rate": 1.814814814814815e-05,
"loss": 0.2875,
"step": 1010
},
{
"epoch": 6.8,
"learning_rate": 1.777777777777778e-05,
"loss": 0.3567,
"step": 1020
},
{
"epoch": 6.86,
"learning_rate": 1.740740740740741e-05,
"loss": 0.3145,
"step": 1030
},
{
"epoch": 6.93,
"learning_rate": 1.7037037037037038e-05,
"loss": 0.3167,
"step": 1040
},
{
"epoch": 7.0,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.3022,
"step": 1050
},
{
"epoch": 7.0,
"eval_accuracy": 0.8672897196261682,
"eval_loss": 0.37750476598739624,
"eval_runtime": 42.1228,
"eval_samples_per_second": 50.804,
"eval_steps_per_second": 1.591,
"step": 1050
},
{
"epoch": 7.07,
"learning_rate": 1.62962962962963e-05,
"loss": 0.286,
"step": 1060
},
{
"epoch": 7.13,
"learning_rate": 1.5925925925925926e-05,
"loss": 0.292,
"step": 1070
},
{
"epoch": 7.2,
"learning_rate": 1.5555555555555555e-05,
"loss": 0.2974,
"step": 1080
},
{
"epoch": 7.27,
"learning_rate": 1.5185185185185186e-05,
"loss": 0.3276,
"step": 1090
},
{
"epoch": 7.33,
"learning_rate": 1.4814814814814815e-05,
"loss": 0.3145,
"step": 1100
},
{
"epoch": 7.4,
"learning_rate": 1.4444444444444444e-05,
"loss": 0.2889,
"step": 1110
},
{
"epoch": 7.47,
"learning_rate": 1.4074074074074075e-05,
"loss": 0.2937,
"step": 1120
},
{
"epoch": 7.53,
"learning_rate": 1.3703703703703704e-05,
"loss": 0.2652,
"step": 1130
},
{
"epoch": 7.6,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.2759,
"step": 1140
},
{
"epoch": 7.66,
"learning_rate": 1.2962962962962962e-05,
"loss": 0.3055,
"step": 1150
},
{
"epoch": 7.73,
"learning_rate": 1.2592592592592592e-05,
"loss": 0.3161,
"step": 1160
},
{
"epoch": 7.8,
"learning_rate": 1.2222222222222222e-05,
"loss": 0.3121,
"step": 1170
},
{
"epoch": 7.86,
"learning_rate": 1.1851851851851853e-05,
"loss": 0.2599,
"step": 1180
},
{
"epoch": 7.93,
"learning_rate": 1.1481481481481482e-05,
"loss": 0.2658,
"step": 1190
},
{
"epoch": 8.0,
"learning_rate": 1.1111111111111112e-05,
"loss": 0.2941,
"step": 1200
},
{
"epoch": 8.0,
"eval_accuracy": 0.8705607476635514,
"eval_loss": 0.3742366135120392,
"eval_runtime": 42.2299,
"eval_samples_per_second": 50.675,
"eval_steps_per_second": 1.587,
"step": 1200
},
{
"epoch": 8.07,
"learning_rate": 1.074074074074074e-05,
"loss": 0.2998,
"step": 1210
},
{
"epoch": 8.13,
"learning_rate": 1.037037037037037e-05,
"loss": 0.3095,
"step": 1220
},
{
"epoch": 8.2,
"learning_rate": 1e-05,
"loss": 0.2833,
"step": 1230
},
{
"epoch": 8.27,
"learning_rate": 9.62962962962963e-06,
"loss": 0.2792,
"step": 1240
},
{
"epoch": 8.33,
"learning_rate": 9.259259259259259e-06,
"loss": 0.2776,
"step": 1250
},
{
"epoch": 8.4,
"learning_rate": 8.88888888888889e-06,
"loss": 0.3238,
"step": 1260
},
{
"epoch": 8.47,
"learning_rate": 8.518518518518519e-06,
"loss": 0.2764,
"step": 1270
},
{
"epoch": 8.53,
"learning_rate": 8.14814814814815e-06,
"loss": 0.2875,
"step": 1280
},
{
"epoch": 8.6,
"learning_rate": 7.777777777777777e-06,
"loss": 0.2545,
"step": 1290
},
{
"epoch": 8.66,
"learning_rate": 7.4074074074074075e-06,
"loss": 0.2551,
"step": 1300
},
{
"epoch": 8.73,
"learning_rate": 7.0370370370370375e-06,
"loss": 0.2611,
"step": 1310
},
{
"epoch": 8.8,
"learning_rate": 6.666666666666667e-06,
"loss": 0.251,
"step": 1320
},
{
"epoch": 8.86,
"learning_rate": 6.296296296296296e-06,
"loss": 0.2864,
"step": 1330
},
{
"epoch": 8.93,
"learning_rate": 5.925925925925927e-06,
"loss": 0.2921,
"step": 1340
},
{
"epoch": 9.0,
"learning_rate": 5.555555555555556e-06,
"loss": 0.2903,
"step": 1350
},
{
"epoch": 9.0,
"eval_accuracy": 0.8696261682242991,
"eval_loss": 0.3808652460575104,
"eval_runtime": 42.311,
"eval_samples_per_second": 50.578,
"eval_steps_per_second": 1.584,
"step": 1350
},
{
"epoch": 9.07,
"learning_rate": 5.185185185185185e-06,
"loss": 0.3301,
"step": 1360
},
{
"epoch": 9.13,
"learning_rate": 4.814814814814815e-06,
"loss": 0.3025,
"step": 1370
},
{
"epoch": 9.2,
"learning_rate": 4.444444444444445e-06,
"loss": 0.2471,
"step": 1380
},
{
"epoch": 9.27,
"learning_rate": 4.074074074074075e-06,
"loss": 0.2805,
"step": 1390
},
{
"epoch": 9.33,
"learning_rate": 3.7037037037037037e-06,
"loss": 0.2679,
"step": 1400
},
{
"epoch": 9.4,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.2784,
"step": 1410
},
{
"epoch": 9.47,
"learning_rate": 2.9629629629629633e-06,
"loss": 0.2785,
"step": 1420
},
{
"epoch": 9.53,
"learning_rate": 2.5925925925925925e-06,
"loss": 0.2662,
"step": 1430
},
{
"epoch": 9.6,
"learning_rate": 2.2222222222222225e-06,
"loss": 0.2679,
"step": 1440
},
{
"epoch": 9.66,
"learning_rate": 1.8518518518518519e-06,
"loss": 0.2702,
"step": 1450
},
{
"epoch": 9.73,
"learning_rate": 1.4814814814814817e-06,
"loss": 0.2737,
"step": 1460
},
{
"epoch": 9.8,
"learning_rate": 1.1111111111111112e-06,
"loss": 0.2747,
"step": 1470
},
{
"epoch": 9.86,
"learning_rate": 7.407407407407408e-07,
"loss": 0.2418,
"step": 1480
},
{
"epoch": 9.93,
"learning_rate": 3.703703703703704e-07,
"loss": 0.2568,
"step": 1490
},
{
"epoch": 10.0,
"learning_rate": 0.0,
"loss": 0.2584,
"step": 1500
},
{
"epoch": 10.0,
"eval_accuracy": 0.8696261682242991,
"eval_loss": 0.37556955218315125,
"eval_runtime": 42.3603,
"eval_samples_per_second": 50.519,
"eval_steps_per_second": 1.582,
"step": 1500
},
{
"epoch": 10.0,
"step": 1500,
"total_flos": 1.4918616518411741e+19,
"train_loss": 0.4060697093009949,
"train_runtime": 6764.8563,
"train_samples_per_second": 28.466,
"train_steps_per_second": 0.222
}
],
"max_steps": 1500,
"num_train_epochs": 10,
"total_flos": 1.4918616518411741e+19,
"trial_name": null,
"trial_params": null
}