graphcore-rahult's picture
End of training
fe7a965
raw
history blame
8.14 kB
{
"best_metric": 0.9825148809523809,
"best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-eurosat/checkpoint-567",
"epoch": 3.0,
"global_step": 567,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.05,
"learning_rate": 8.771929824561403e-06,
"loss": 2.3141,
"step": 10
},
{
"epoch": 0.11,
"learning_rate": 1.7543859649122806e-05,
"loss": 2.2088,
"step": 20
},
{
"epoch": 0.16,
"learning_rate": 2.6315789473684212e-05,
"loss": 2.1541,
"step": 30
},
{
"epoch": 0.21,
"learning_rate": 3.508771929824561e-05,
"loss": 1.8185,
"step": 40
},
{
"epoch": 0.26,
"learning_rate": 4.3859649122807014e-05,
"loss": 1.468,
"step": 50
},
{
"epoch": 0.32,
"learning_rate": 4.970588235294118e-05,
"loss": 1.3207,
"step": 60
},
{
"epoch": 0.37,
"learning_rate": 4.872549019607843e-05,
"loss": 0.9924,
"step": 70
},
{
"epoch": 0.42,
"learning_rate": 4.774509803921569e-05,
"loss": 0.7575,
"step": 80
},
{
"epoch": 0.48,
"learning_rate": 4.6764705882352944e-05,
"loss": 0.4987,
"step": 90
},
{
"epoch": 0.53,
"learning_rate": 4.5784313725490194e-05,
"loss": 0.5602,
"step": 100
},
{
"epoch": 0.58,
"learning_rate": 4.480392156862745e-05,
"loss": 0.6816,
"step": 110
},
{
"epoch": 0.63,
"learning_rate": 4.382352941176471e-05,
"loss": 0.6875,
"step": 120
},
{
"epoch": 0.69,
"learning_rate": 4.284313725490196e-05,
"loss": 0.4998,
"step": 130
},
{
"epoch": 0.74,
"learning_rate": 4.1862745098039217e-05,
"loss": 0.4879,
"step": 140
},
{
"epoch": 0.79,
"learning_rate": 4.0882352941176474e-05,
"loss": 0.3393,
"step": 150
},
{
"epoch": 0.85,
"learning_rate": 3.990196078431373e-05,
"loss": 0.2399,
"step": 160
},
{
"epoch": 0.9,
"learning_rate": 3.892156862745098e-05,
"loss": 0.3542,
"step": 170
},
{
"epoch": 0.95,
"learning_rate": 3.794117647058824e-05,
"loss": 0.2284,
"step": 180
},
{
"epoch": 1.0,
"eval_accuracy": 0.9657738095238095,
"eval_loss": 0.27978515625,
"eval_runtime": 7.7904,
"eval_samples_per_second": 345.039,
"eval_steps_per_second": 21.565,
"step": 189
},
{
"epoch": 1.01,
"learning_rate": 3.6960784313725496e-05,
"loss": 0.2552,
"step": 190
},
{
"epoch": 1.06,
"learning_rate": 3.5980392156862746e-05,
"loss": 0.2004,
"step": 200
},
{
"epoch": 1.11,
"learning_rate": 3.5e-05,
"loss": 0.2532,
"step": 210
},
{
"epoch": 1.16,
"learning_rate": 3.401960784313726e-05,
"loss": 0.3075,
"step": 220
},
{
"epoch": 1.22,
"learning_rate": 3.303921568627451e-05,
"loss": 0.2281,
"step": 230
},
{
"epoch": 1.27,
"learning_rate": 3.205882352941177e-05,
"loss": 0.3189,
"step": 240
},
{
"epoch": 1.32,
"learning_rate": 3.107843137254902e-05,
"loss": 0.3038,
"step": 250
},
{
"epoch": 1.38,
"learning_rate": 3.0098039215686273e-05,
"loss": 0.1736,
"step": 260
},
{
"epoch": 1.43,
"learning_rate": 2.9117647058823534e-05,
"loss": 0.2718,
"step": 270
},
{
"epoch": 1.48,
"learning_rate": 2.8137254901960784e-05,
"loss": 0.3009,
"step": 280
},
{
"epoch": 1.53,
"learning_rate": 2.7156862745098038e-05,
"loss": 0.2359,
"step": 290
},
{
"epoch": 1.59,
"learning_rate": 2.6176470588235295e-05,
"loss": 0.2563,
"step": 300
},
{
"epoch": 1.64,
"learning_rate": 2.519607843137255e-05,
"loss": 0.2496,
"step": 310
},
{
"epoch": 1.69,
"learning_rate": 2.4215686274509806e-05,
"loss": 0.324,
"step": 320
},
{
"epoch": 1.75,
"learning_rate": 2.323529411764706e-05,
"loss": 0.2806,
"step": 330
},
{
"epoch": 1.8,
"learning_rate": 2.2254901960784314e-05,
"loss": 0.3569,
"step": 340
},
{
"epoch": 1.85,
"learning_rate": 2.1274509803921568e-05,
"loss": 0.2734,
"step": 350
},
{
"epoch": 1.9,
"learning_rate": 2.0294117647058825e-05,
"loss": 0.2582,
"step": 360
},
{
"epoch": 1.96,
"learning_rate": 1.931372549019608e-05,
"loss": 0.2262,
"step": 370
},
{
"epoch": 2.0,
"eval_accuracy": 0.9799107142857143,
"eval_loss": 0.1624755859375,
"eval_runtime": 7.8583,
"eval_samples_per_second": 342.057,
"eval_steps_per_second": 21.379,
"step": 378
},
{
"epoch": 2.01,
"learning_rate": 1.8333333333333333e-05,
"loss": 0.4627,
"step": 380
},
{
"epoch": 2.06,
"learning_rate": 1.735294117647059e-05,
"loss": 0.2553,
"step": 390
},
{
"epoch": 2.12,
"learning_rate": 1.6372549019607844e-05,
"loss": 0.1291,
"step": 400
},
{
"epoch": 2.17,
"learning_rate": 1.53921568627451e-05,
"loss": 0.1432,
"step": 410
},
{
"epoch": 2.22,
"learning_rate": 1.4411764705882352e-05,
"loss": 0.3603,
"step": 420
},
{
"epoch": 2.28,
"learning_rate": 1.3431372549019607e-05,
"loss": 0.2984,
"step": 430
},
{
"epoch": 2.33,
"learning_rate": 1.2450980392156863e-05,
"loss": 0.1232,
"step": 440
},
{
"epoch": 2.38,
"learning_rate": 1.1470588235294118e-05,
"loss": 0.215,
"step": 450
},
{
"epoch": 2.43,
"learning_rate": 1.0490196078431374e-05,
"loss": 0.1089,
"step": 460
},
{
"epoch": 2.49,
"learning_rate": 9.509803921568628e-06,
"loss": 0.2441,
"step": 470
},
{
"epoch": 2.54,
"learning_rate": 8.529411764705883e-06,
"loss": 0.127,
"step": 480
},
{
"epoch": 2.59,
"learning_rate": 7.549019607843137e-06,
"loss": 0.1098,
"step": 490
},
{
"epoch": 2.65,
"learning_rate": 6.568627450980393e-06,
"loss": 0.1042,
"step": 500
},
{
"epoch": 2.7,
"learning_rate": 5.588235294117647e-06,
"loss": 0.1162,
"step": 510
},
{
"epoch": 2.75,
"learning_rate": 4.607843137254902e-06,
"loss": 0.1821,
"step": 520
},
{
"epoch": 2.8,
"learning_rate": 3.6274509803921573e-06,
"loss": 0.3014,
"step": 530
},
{
"epoch": 2.86,
"learning_rate": 2.647058823529412e-06,
"loss": 0.181,
"step": 540
},
{
"epoch": 2.91,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.1033,
"step": 550
},
{
"epoch": 2.96,
"learning_rate": 6.862745098039216e-07,
"loss": 0.1693,
"step": 560
},
{
"epoch": 3.0,
"eval_accuracy": 0.9825148809523809,
"eval_loss": 0.1412353515625,
"eval_runtime": 7.3548,
"eval_samples_per_second": 365.475,
"eval_steps_per_second": 22.842,
"step": 567
},
{
"epoch": 3.0,
"step": 567,
"total_flos": 7.199311221168632e+20,
"train_loss": 0.46718859798693785,
"train_runtime": 815.2101,
"train_samples_per_second": 89.027,
"train_steps_per_second": 0.696
}
],
"max_steps": 567,
"num_train_epochs": 3,
"start_time": 1670273012.1493745,
"total_flos": 7.199311221168632e+20,
"trial_name": null,
"trial_params": null
}