Initial model upload
Browse files- all_results.json +8 -0
- checkpoint-350/config.json +33 -0
- checkpoint-350/model.safetensors +3 -0
- checkpoint-350/optimizer.pt +3 -0
- checkpoint-350/preprocessor_config.json +23 -0
- checkpoint-350/rng_state.pth +3 -0
- checkpoint-350/scaler.pt +3 -0
- checkpoint-350/scheduler.pt +3 -0
- checkpoint-350/trainer_state.json +342 -0
- checkpoint-350/training_args.bin +3 -0
- checkpoint-375/config.json +33 -0
- checkpoint-375/model.safetensors +3 -0
- checkpoint-375/optimizer.pt +3 -0
- checkpoint-375/preprocessor_config.json +23 -0
- checkpoint-375/rng_state.pth +3 -0
- checkpoint-375/scaler.pt +3 -0
- checkpoint-375/scheduler.pt +3 -0
- checkpoint-375/trainer_state.json +356 -0
- checkpoint-375/training_args.bin +3 -0
- config.json +33 -0
- model.safetensors +3 -0
- preprocessor_config.json +23 -0
- train_results.json +8 -0
- trainer_state.json +365 -0
- training_args.bin +3 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 4.64951937687552e+17,
|
| 4 |
+
"train_loss": 0.3516346867879232,
|
| 5 |
+
"train_runtime": 232.2268,
|
| 6 |
+
"train_samples_per_second": 25.837,
|
| 7 |
+
"train_steps_per_second": 1.615
|
| 8 |
+
}
|
checkpoint-350/config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ViTForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "Normal",
|
| 13 |
+
"1": "Stroke"
|
| 14 |
+
},
|
| 15 |
+
"image_size": 224,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 3072,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"Normal": "0",
|
| 20 |
+
"Stroke": "1"
|
| 21 |
+
},
|
| 22 |
+
"layer_norm_eps": 1e-12,
|
| 23 |
+
"model_type": "vit",
|
| 24 |
+
"num_attention_heads": 12,
|
| 25 |
+
"num_channels": 3,
|
| 26 |
+
"num_hidden_layers": 12,
|
| 27 |
+
"patch_size": 16,
|
| 28 |
+
"pooler_act": "tanh",
|
| 29 |
+
"pooler_output_size": 768,
|
| 30 |
+
"problem_type": "single_label_classification",
|
| 31 |
+
"qkv_bias": true,
|
| 32 |
+
"transformers_version": "4.56.1"
|
| 33 |
+
}
|
checkpoint-350/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cbf166d41eee416d013784388cb6329a093caa28fe2a65c40b639264d1cb02c
|
| 3 |
+
size 343223968
|
checkpoint-350/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dbbe42e500d13c5b12208fbb0188e4a13c3ea5ab070c26df14185484b5d88f63
|
| 3 |
+
size 686572107
|
checkpoint-350/preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"resample": 2,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-350/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17cd930da9783ca70bad4b9cdeee6a06c0acea8f34645a333c93341f487f66a3
|
| 3 |
+
size 14645
|
checkpoint-350/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7a3b7c7f6ae3d4b9e2fcb0774d91cea73c152177e3f6818dfc11db3a9a74275
|
| 3 |
+
size 1383
|
checkpoint-350/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47cfe82f13a5be2b56585c903bfd0022928fc105dd431787360f5e72fbf89c56
|
| 3 |
+
size 1465
|
checkpoint-350/trainer_state.json
ADDED
|
@@ -0,0 +1,342 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 350,
|
| 3 |
+
"best_metric": 0.19462016224861145,
|
| 4 |
+
"best_model_checkpoint": "./vit-stroke-detector/checkpoint-350",
|
| 5 |
+
"epoch": 2.8,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 350,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08,
|
| 14 |
+
"grad_norm": 0.828596830368042,
|
| 15 |
+
"learning_rate": 0.0001952,
|
| 16 |
+
"loss": 0.6849,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16,
|
| 21 |
+
"grad_norm": 1.6220552921295166,
|
| 22 |
+
"learning_rate": 0.00018986666666666668,
|
| 23 |
+
"loss": 0.6632,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.24,
|
| 28 |
+
"grad_norm": 1.931017279624939,
|
| 29 |
+
"learning_rate": 0.00018453333333333334,
|
| 30 |
+
"loss": 0.6737,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.32,
|
| 35 |
+
"grad_norm": 0.5325367450714111,
|
| 36 |
+
"learning_rate": 0.00017920000000000002,
|
| 37 |
+
"loss": 0.6441,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4,
|
| 42 |
+
"grad_norm": 0.20815162360668182,
|
| 43 |
+
"learning_rate": 0.00017386666666666667,
|
| 44 |
+
"loss": 0.6959,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4,
|
| 49 |
+
"eval_accuracy": 0.6007984031936128,
|
| 50 |
+
"eval_loss": 0.6659401059150696,
|
| 51 |
+
"eval_runtime": 12.6028,
|
| 52 |
+
"eval_samples_per_second": 39.753,
|
| 53 |
+
"eval_steps_per_second": 4.999,
|
| 54 |
+
"step": 50
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.48,
|
| 58 |
+
"grad_norm": 0.6144075393676758,
|
| 59 |
+
"learning_rate": 0.00016853333333333336,
|
| 60 |
+
"loss": 0.6542,
|
| 61 |
+
"step": 60
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.56,
|
| 65 |
+
"grad_norm": 0.3622696101665497,
|
| 66 |
+
"learning_rate": 0.0001632,
|
| 67 |
+
"loss": 0.6336,
|
| 68 |
+
"step": 70
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.64,
|
| 72 |
+
"grad_norm": 1.7294650077819824,
|
| 73 |
+
"learning_rate": 0.00015786666666666666,
|
| 74 |
+
"loss": 0.6705,
|
| 75 |
+
"step": 80
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.72,
|
| 79 |
+
"grad_norm": 1.2596920728683472,
|
| 80 |
+
"learning_rate": 0.00015253333333333335,
|
| 81 |
+
"loss": 0.6144,
|
| 82 |
+
"step": 90
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.8,
|
| 86 |
+
"grad_norm": 2.432494640350342,
|
| 87 |
+
"learning_rate": 0.0001472,
|
| 88 |
+
"loss": 0.6038,
|
| 89 |
+
"step": 100
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.8,
|
| 93 |
+
"eval_accuracy": 0.624750499001996,
|
| 94 |
+
"eval_loss": 0.6349905133247375,
|
| 95 |
+
"eval_runtime": 4.8798,
|
| 96 |
+
"eval_samples_per_second": 102.668,
|
| 97 |
+
"eval_steps_per_second": 12.91,
|
| 98 |
+
"step": 100
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.88,
|
| 102 |
+
"grad_norm": 0.8303850293159485,
|
| 103 |
+
"learning_rate": 0.00014186666666666668,
|
| 104 |
+
"loss": 0.5625,
|
| 105 |
+
"step": 110
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.96,
|
| 109 |
+
"grad_norm": 1.7314858436584473,
|
| 110 |
+
"learning_rate": 0.00013653333333333334,
|
| 111 |
+
"loss": 0.4656,
|
| 112 |
+
"step": 120
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.04,
|
| 116 |
+
"grad_norm": 2.2635459899902344,
|
| 117 |
+
"learning_rate": 0.00013120000000000002,
|
| 118 |
+
"loss": 0.401,
|
| 119 |
+
"step": 130
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.12,
|
| 123 |
+
"grad_norm": 1.9528553485870361,
|
| 124 |
+
"learning_rate": 0.00012586666666666667,
|
| 125 |
+
"loss": 0.5874,
|
| 126 |
+
"step": 140
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.2,
|
| 130 |
+
"grad_norm": 2.2214958667755127,
|
| 131 |
+
"learning_rate": 0.00012053333333333334,
|
| 132 |
+
"loss": 0.4041,
|
| 133 |
+
"step": 150
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.2,
|
| 137 |
+
"eval_accuracy": 0.7904191616766467,
|
| 138 |
+
"eval_loss": 0.5236000418663025,
|
| 139 |
+
"eval_runtime": 6.2426,
|
| 140 |
+
"eval_samples_per_second": 80.255,
|
| 141 |
+
"eval_steps_per_second": 10.092,
|
| 142 |
+
"step": 150
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.28,
|
| 146 |
+
"grad_norm": 1.578341007232666,
|
| 147 |
+
"learning_rate": 0.0001152,
|
| 148 |
+
"loss": 0.4256,
|
| 149 |
+
"step": 160
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3599999999999999,
|
| 153 |
+
"grad_norm": 0.521945059299469,
|
| 154 |
+
"learning_rate": 0.00010986666666666668,
|
| 155 |
+
"loss": 0.3168,
|
| 156 |
+
"step": 170
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.44,
|
| 160 |
+
"grad_norm": 1.1707364320755005,
|
| 161 |
+
"learning_rate": 0.00010453333333333333,
|
| 162 |
+
"loss": 0.4316,
|
| 163 |
+
"step": 180
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.52,
|
| 167 |
+
"grad_norm": 1.4519349336624146,
|
| 168 |
+
"learning_rate": 9.92e-05,
|
| 169 |
+
"loss": 0.2944,
|
| 170 |
+
"step": 190
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.6,
|
| 174 |
+
"grad_norm": 1.0237702131271362,
|
| 175 |
+
"learning_rate": 9.386666666666667e-05,
|
| 176 |
+
"loss": 0.3243,
|
| 177 |
+
"step": 200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6,
|
| 181 |
+
"eval_accuracy": 0.874251497005988,
|
| 182 |
+
"eval_loss": 0.3280556797981262,
|
| 183 |
+
"eval_runtime": 4.9813,
|
| 184 |
+
"eval_samples_per_second": 100.577,
|
| 185 |
+
"eval_steps_per_second": 12.647,
|
| 186 |
+
"step": 200
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.6800000000000002,
|
| 190 |
+
"grad_norm": 4.305712699890137,
|
| 191 |
+
"learning_rate": 8.853333333333333e-05,
|
| 192 |
+
"loss": 0.3118,
|
| 193 |
+
"step": 210
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.76,
|
| 197 |
+
"grad_norm": 1.6635700464248657,
|
| 198 |
+
"learning_rate": 8.32e-05,
|
| 199 |
+
"loss": 0.2534,
|
| 200 |
+
"step": 220
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.8399999999999999,
|
| 204 |
+
"grad_norm": 1.5301051139831543,
|
| 205 |
+
"learning_rate": 7.786666666666667e-05,
|
| 206 |
+
"loss": 0.2702,
|
| 207 |
+
"step": 230
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.92,
|
| 211 |
+
"grad_norm": 0.7002310156822205,
|
| 212 |
+
"learning_rate": 7.253333333333334e-05,
|
| 213 |
+
"loss": 0.1926,
|
| 214 |
+
"step": 240
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 2.0,
|
| 218 |
+
"grad_norm": 0.4949961304664612,
|
| 219 |
+
"learning_rate": 6.720000000000001e-05,
|
| 220 |
+
"loss": 0.2041,
|
| 221 |
+
"step": 250
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.0,
|
| 225 |
+
"eval_accuracy": 0.8922155688622755,
|
| 226 |
+
"eval_loss": 0.28237032890319824,
|
| 227 |
+
"eval_runtime": 4.9486,
|
| 228 |
+
"eval_samples_per_second": 101.242,
|
| 229 |
+
"eval_steps_per_second": 12.731,
|
| 230 |
+
"step": 250
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.08,
|
| 234 |
+
"grad_norm": 13.5546236038208,
|
| 235 |
+
"learning_rate": 6.186666666666668e-05,
|
| 236 |
+
"loss": 0.173,
|
| 237 |
+
"step": 260
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.16,
|
| 241 |
+
"grad_norm": 7.708662986755371,
|
| 242 |
+
"learning_rate": 5.6533333333333336e-05,
|
| 243 |
+
"loss": 0.1206,
|
| 244 |
+
"step": 270
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.24,
|
| 248 |
+
"grad_norm": 0.44631195068359375,
|
| 249 |
+
"learning_rate": 5.1200000000000004e-05,
|
| 250 |
+
"loss": 0.1615,
|
| 251 |
+
"step": 280
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.32,
|
| 255 |
+
"grad_norm": 1.9430702924728394,
|
| 256 |
+
"learning_rate": 4.5866666666666666e-05,
|
| 257 |
+
"loss": 0.1196,
|
| 258 |
+
"step": 290
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.4,
|
| 262 |
+
"grad_norm": 2.0821869373321533,
|
| 263 |
+
"learning_rate": 4.0533333333333334e-05,
|
| 264 |
+
"loss": 0.117,
|
| 265 |
+
"step": 300
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.4,
|
| 269 |
+
"eval_accuracy": 0.9161676646706587,
|
| 270 |
+
"eval_loss": 0.22871683537960052,
|
| 271 |
+
"eval_runtime": 4.8973,
|
| 272 |
+
"eval_samples_per_second": 102.302,
|
| 273 |
+
"eval_steps_per_second": 12.864,
|
| 274 |
+
"step": 300
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.48,
|
| 278 |
+
"grad_norm": 9.634779930114746,
|
| 279 |
+
"learning_rate": 3.52e-05,
|
| 280 |
+
"loss": 0.0606,
|
| 281 |
+
"step": 310
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.56,
|
| 285 |
+
"grad_norm": 6.319236755371094,
|
| 286 |
+
"learning_rate": 2.986666666666667e-05,
|
| 287 |
+
"loss": 0.0774,
|
| 288 |
+
"step": 320
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.64,
|
| 292 |
+
"grad_norm": 9.121068000793457,
|
| 293 |
+
"learning_rate": 2.4533333333333334e-05,
|
| 294 |
+
"loss": 0.06,
|
| 295 |
+
"step": 330
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.7199999999999998,
|
| 299 |
+
"grad_norm": 3.4029245376586914,
|
| 300 |
+
"learning_rate": 1.9200000000000003e-05,
|
| 301 |
+
"loss": 0.0997,
|
| 302 |
+
"step": 340
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.8,
|
| 306 |
+
"grad_norm": 0.8972381353378296,
|
| 307 |
+
"learning_rate": 1.3866666666666667e-05,
|
| 308 |
+
"loss": 0.0353,
|
| 309 |
+
"step": 350
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.8,
|
| 313 |
+
"eval_accuracy": 0.9401197604790419,
|
| 314 |
+
"eval_loss": 0.19462016224861145,
|
| 315 |
+
"eval_runtime": 5.5189,
|
| 316 |
+
"eval_samples_per_second": 90.78,
|
| 317 |
+
"eval_steps_per_second": 11.415,
|
| 318 |
+
"step": 350
|
| 319 |
+
}
|
| 320 |
+
],
|
| 321 |
+
"logging_steps": 10,
|
| 322 |
+
"max_steps": 375,
|
| 323 |
+
"num_input_tokens_seen": 0,
|
| 324 |
+
"num_train_epochs": 3,
|
| 325 |
+
"save_steps": 50,
|
| 326 |
+
"stateful_callbacks": {
|
| 327 |
+
"TrainerControl": {
|
| 328 |
+
"args": {
|
| 329 |
+
"should_epoch_stop": false,
|
| 330 |
+
"should_evaluate": false,
|
| 331 |
+
"should_log": false,
|
| 332 |
+
"should_save": true,
|
| 333 |
+
"should_training_stop": false
|
| 334 |
+
},
|
| 335 |
+
"attributes": {}
|
| 336 |
+
}
|
| 337 |
+
},
|
| 338 |
+
"total_flos": 4.339551418417152e+17,
|
| 339 |
+
"train_batch_size": 16,
|
| 340 |
+
"trial_name": null,
|
| 341 |
+
"trial_params": null
|
| 342 |
+
}
|
checkpoint-350/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8317ceaa60875571340e1a38d19c2d04367ed0d503c395ed80f409547047289
|
| 3 |
+
size 5777
|
checkpoint-375/config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ViTForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "Normal",
|
| 13 |
+
"1": "Stroke"
|
| 14 |
+
},
|
| 15 |
+
"image_size": 224,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 3072,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"Normal": "0",
|
| 20 |
+
"Stroke": "1"
|
| 21 |
+
},
|
| 22 |
+
"layer_norm_eps": 1e-12,
|
| 23 |
+
"model_type": "vit",
|
| 24 |
+
"num_attention_heads": 12,
|
| 25 |
+
"num_channels": 3,
|
| 26 |
+
"num_hidden_layers": 12,
|
| 27 |
+
"patch_size": 16,
|
| 28 |
+
"pooler_act": "tanh",
|
| 29 |
+
"pooler_output_size": 768,
|
| 30 |
+
"problem_type": "single_label_classification",
|
| 31 |
+
"qkv_bias": true,
|
| 32 |
+
"transformers_version": "4.56.1"
|
| 33 |
+
}
|
checkpoint-375/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a067b33266d306c455b262c2efccd9cc3fdf3a1dff8cb6dcd5f5446c36d89ea
|
| 3 |
+
size 343223968
|
checkpoint-375/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8193b1640fc6ef97c9dbc111117e2e562c7212f79c0bc5e5c084a6ca5310ca9
|
| 3 |
+
size 686572107
|
checkpoint-375/preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"resample": 2,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-375/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17cd930da9783ca70bad4b9cdeee6a06c0acea8f34645a333c93341f487f66a3
|
| 3 |
+
size 14645
|
checkpoint-375/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac58ef7454dc327cdd1497d560e3c1cadc5881183c106e69fda6701af10248e8
|
| 3 |
+
size 1383
|
checkpoint-375/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:842c9cfc377474f3cc400aa68eb12f46fcbb5333ee717560db8d34b4443de63e
|
| 3 |
+
size 1465
|
checkpoint-375/trainer_state.json
ADDED
|
@@ -0,0 +1,356 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 350,
|
| 3 |
+
"best_metric": 0.19462016224861145,
|
| 4 |
+
"best_model_checkpoint": "./vit-stroke-detector/checkpoint-350",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08,
|
| 14 |
+
"grad_norm": 0.828596830368042,
|
| 15 |
+
"learning_rate": 0.0001952,
|
| 16 |
+
"loss": 0.6849,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16,
|
| 21 |
+
"grad_norm": 1.6220552921295166,
|
| 22 |
+
"learning_rate": 0.00018986666666666668,
|
| 23 |
+
"loss": 0.6632,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.24,
|
| 28 |
+
"grad_norm": 1.931017279624939,
|
| 29 |
+
"learning_rate": 0.00018453333333333334,
|
| 30 |
+
"loss": 0.6737,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.32,
|
| 35 |
+
"grad_norm": 0.5325367450714111,
|
| 36 |
+
"learning_rate": 0.00017920000000000002,
|
| 37 |
+
"loss": 0.6441,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4,
|
| 42 |
+
"grad_norm": 0.20815162360668182,
|
| 43 |
+
"learning_rate": 0.00017386666666666667,
|
| 44 |
+
"loss": 0.6959,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4,
|
| 49 |
+
"eval_accuracy": 0.6007984031936128,
|
| 50 |
+
"eval_loss": 0.6659401059150696,
|
| 51 |
+
"eval_runtime": 12.6028,
|
| 52 |
+
"eval_samples_per_second": 39.753,
|
| 53 |
+
"eval_steps_per_second": 4.999,
|
| 54 |
+
"step": 50
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.48,
|
| 58 |
+
"grad_norm": 0.6144075393676758,
|
| 59 |
+
"learning_rate": 0.00016853333333333336,
|
| 60 |
+
"loss": 0.6542,
|
| 61 |
+
"step": 60
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.56,
|
| 65 |
+
"grad_norm": 0.3622696101665497,
|
| 66 |
+
"learning_rate": 0.0001632,
|
| 67 |
+
"loss": 0.6336,
|
| 68 |
+
"step": 70
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.64,
|
| 72 |
+
"grad_norm": 1.7294650077819824,
|
| 73 |
+
"learning_rate": 0.00015786666666666666,
|
| 74 |
+
"loss": 0.6705,
|
| 75 |
+
"step": 80
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.72,
|
| 79 |
+
"grad_norm": 1.2596920728683472,
|
| 80 |
+
"learning_rate": 0.00015253333333333335,
|
| 81 |
+
"loss": 0.6144,
|
| 82 |
+
"step": 90
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.8,
|
| 86 |
+
"grad_norm": 2.432494640350342,
|
| 87 |
+
"learning_rate": 0.0001472,
|
| 88 |
+
"loss": 0.6038,
|
| 89 |
+
"step": 100
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.8,
|
| 93 |
+
"eval_accuracy": 0.624750499001996,
|
| 94 |
+
"eval_loss": 0.6349905133247375,
|
| 95 |
+
"eval_runtime": 4.8798,
|
| 96 |
+
"eval_samples_per_second": 102.668,
|
| 97 |
+
"eval_steps_per_second": 12.91,
|
| 98 |
+
"step": 100
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.88,
|
| 102 |
+
"grad_norm": 0.8303850293159485,
|
| 103 |
+
"learning_rate": 0.00014186666666666668,
|
| 104 |
+
"loss": 0.5625,
|
| 105 |
+
"step": 110
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.96,
|
| 109 |
+
"grad_norm": 1.7314858436584473,
|
| 110 |
+
"learning_rate": 0.00013653333333333334,
|
| 111 |
+
"loss": 0.4656,
|
| 112 |
+
"step": 120
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.04,
|
| 116 |
+
"grad_norm": 2.2635459899902344,
|
| 117 |
+
"learning_rate": 0.00013120000000000002,
|
| 118 |
+
"loss": 0.401,
|
| 119 |
+
"step": 130
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.12,
|
| 123 |
+
"grad_norm": 1.9528553485870361,
|
| 124 |
+
"learning_rate": 0.00012586666666666667,
|
| 125 |
+
"loss": 0.5874,
|
| 126 |
+
"step": 140
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.2,
|
| 130 |
+
"grad_norm": 2.2214958667755127,
|
| 131 |
+
"learning_rate": 0.00012053333333333334,
|
| 132 |
+
"loss": 0.4041,
|
| 133 |
+
"step": 150
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.2,
|
| 137 |
+
"eval_accuracy": 0.7904191616766467,
|
| 138 |
+
"eval_loss": 0.5236000418663025,
|
| 139 |
+
"eval_runtime": 6.2426,
|
| 140 |
+
"eval_samples_per_second": 80.255,
|
| 141 |
+
"eval_steps_per_second": 10.092,
|
| 142 |
+
"step": 150
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.28,
|
| 146 |
+
"grad_norm": 1.578341007232666,
|
| 147 |
+
"learning_rate": 0.0001152,
|
| 148 |
+
"loss": 0.4256,
|
| 149 |
+
"step": 160
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3599999999999999,
|
| 153 |
+
"grad_norm": 0.521945059299469,
|
| 154 |
+
"learning_rate": 0.00010986666666666668,
|
| 155 |
+
"loss": 0.3168,
|
| 156 |
+
"step": 170
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.44,
|
| 160 |
+
"grad_norm": 1.1707364320755005,
|
| 161 |
+
"learning_rate": 0.00010453333333333333,
|
| 162 |
+
"loss": 0.4316,
|
| 163 |
+
"step": 180
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.52,
|
| 167 |
+
"grad_norm": 1.4519349336624146,
|
| 168 |
+
"learning_rate": 9.92e-05,
|
| 169 |
+
"loss": 0.2944,
|
| 170 |
+
"step": 190
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.6,
|
| 174 |
+
"grad_norm": 1.0237702131271362,
|
| 175 |
+
"learning_rate": 9.386666666666667e-05,
|
| 176 |
+
"loss": 0.3243,
|
| 177 |
+
"step": 200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6,
|
| 181 |
+
"eval_accuracy": 0.874251497005988,
|
| 182 |
+
"eval_loss": 0.3280556797981262,
|
| 183 |
+
"eval_runtime": 4.9813,
|
| 184 |
+
"eval_samples_per_second": 100.577,
|
| 185 |
+
"eval_steps_per_second": 12.647,
|
| 186 |
+
"step": 200
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.6800000000000002,
|
| 190 |
+
"grad_norm": 4.305712699890137,
|
| 191 |
+
"learning_rate": 8.853333333333333e-05,
|
| 192 |
+
"loss": 0.3118,
|
| 193 |
+
"step": 210
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.76,
|
| 197 |
+
"grad_norm": 1.6635700464248657,
|
| 198 |
+
"learning_rate": 8.32e-05,
|
| 199 |
+
"loss": 0.2534,
|
| 200 |
+
"step": 220
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.8399999999999999,
|
| 204 |
+
"grad_norm": 1.5301051139831543,
|
| 205 |
+
"learning_rate": 7.786666666666667e-05,
|
| 206 |
+
"loss": 0.2702,
|
| 207 |
+
"step": 230
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.92,
|
| 211 |
+
"grad_norm": 0.7002310156822205,
|
| 212 |
+
"learning_rate": 7.253333333333334e-05,
|
| 213 |
+
"loss": 0.1926,
|
| 214 |
+
"step": 240
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 2.0,
|
| 218 |
+
"grad_norm": 0.4949961304664612,
|
| 219 |
+
"learning_rate": 6.720000000000001e-05,
|
| 220 |
+
"loss": 0.2041,
|
| 221 |
+
"step": 250
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.0,
|
| 225 |
+
"eval_accuracy": 0.8922155688622755,
|
| 226 |
+
"eval_loss": 0.28237032890319824,
|
| 227 |
+
"eval_runtime": 4.9486,
|
| 228 |
+
"eval_samples_per_second": 101.242,
|
| 229 |
+
"eval_steps_per_second": 12.731,
|
| 230 |
+
"step": 250
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.08,
|
| 234 |
+
"grad_norm": 13.5546236038208,
|
| 235 |
+
"learning_rate": 6.186666666666668e-05,
|
| 236 |
+
"loss": 0.173,
|
| 237 |
+
"step": 260
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.16,
|
| 241 |
+
"grad_norm": 7.708662986755371,
|
| 242 |
+
"learning_rate": 5.6533333333333336e-05,
|
| 243 |
+
"loss": 0.1206,
|
| 244 |
+
"step": 270
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.24,
|
| 248 |
+
"grad_norm": 0.44631195068359375,
|
| 249 |
+
"learning_rate": 5.1200000000000004e-05,
|
| 250 |
+
"loss": 0.1615,
|
| 251 |
+
"step": 280
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.32,
|
| 255 |
+
"grad_norm": 1.9430702924728394,
|
| 256 |
+
"learning_rate": 4.5866666666666666e-05,
|
| 257 |
+
"loss": 0.1196,
|
| 258 |
+
"step": 290
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.4,
|
| 262 |
+
"grad_norm": 2.0821869373321533,
|
| 263 |
+
"learning_rate": 4.0533333333333334e-05,
|
| 264 |
+
"loss": 0.117,
|
| 265 |
+
"step": 300
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.4,
|
| 269 |
+
"eval_accuracy": 0.9161676646706587,
|
| 270 |
+
"eval_loss": 0.22871683537960052,
|
| 271 |
+
"eval_runtime": 4.8973,
|
| 272 |
+
"eval_samples_per_second": 102.302,
|
| 273 |
+
"eval_steps_per_second": 12.864,
|
| 274 |
+
"step": 300
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.48,
|
| 278 |
+
"grad_norm": 9.634779930114746,
|
| 279 |
+
"learning_rate": 3.52e-05,
|
| 280 |
+
"loss": 0.0606,
|
| 281 |
+
"step": 310
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.56,
|
| 285 |
+
"grad_norm": 6.319236755371094,
|
| 286 |
+
"learning_rate": 2.986666666666667e-05,
|
| 287 |
+
"loss": 0.0774,
|
| 288 |
+
"step": 320
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.64,
|
| 292 |
+
"grad_norm": 9.121068000793457,
|
| 293 |
+
"learning_rate": 2.4533333333333334e-05,
|
| 294 |
+
"loss": 0.06,
|
| 295 |
+
"step": 330
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.7199999999999998,
|
| 299 |
+
"grad_norm": 3.4029245376586914,
|
| 300 |
+
"learning_rate": 1.9200000000000003e-05,
|
| 301 |
+
"loss": 0.0997,
|
| 302 |
+
"step": 340
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.8,
|
| 306 |
+
"grad_norm": 0.8972381353378296,
|
| 307 |
+
"learning_rate": 1.3866666666666667e-05,
|
| 308 |
+
"loss": 0.0353,
|
| 309 |
+
"step": 350
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.8,
|
| 313 |
+
"eval_accuracy": 0.9401197604790419,
|
| 314 |
+
"eval_loss": 0.19462016224861145,
|
| 315 |
+
"eval_runtime": 5.5189,
|
| 316 |
+
"eval_samples_per_second": 90.78,
|
| 317 |
+
"eval_steps_per_second": 11.415,
|
| 318 |
+
"step": 350
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.88,
|
| 322 |
+
"grad_norm": 0.1985252946615219,
|
| 323 |
+
"learning_rate": 8.533333333333334e-06,
|
| 324 |
+
"loss": 0.0833,
|
| 325 |
+
"step": 360
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.96,
|
| 329 |
+
"grad_norm": 0.129100501537323,
|
| 330 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 331 |
+
"loss": 0.0775,
|
| 332 |
+
"step": 370
|
| 333 |
+
}
|
| 334 |
+
],
|
| 335 |
+
"logging_steps": 10,
|
| 336 |
+
"max_steps": 375,
|
| 337 |
+
"num_input_tokens_seen": 0,
|
| 338 |
+
"num_train_epochs": 3,
|
| 339 |
+
"save_steps": 50,
|
| 340 |
+
"stateful_callbacks": {
|
| 341 |
+
"TrainerControl": {
|
| 342 |
+
"args": {
|
| 343 |
+
"should_epoch_stop": false,
|
| 344 |
+
"should_evaluate": false,
|
| 345 |
+
"should_log": false,
|
| 346 |
+
"should_save": true,
|
| 347 |
+
"should_training_stop": true
|
| 348 |
+
},
|
| 349 |
+
"attributes": {}
|
| 350 |
+
}
|
| 351 |
+
},
|
| 352 |
+
"total_flos": 4.64951937687552e+17,
|
| 353 |
+
"train_batch_size": 16,
|
| 354 |
+
"trial_name": null,
|
| 355 |
+
"trial_params": null
|
| 356 |
+
}
|
checkpoint-375/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8317ceaa60875571340e1a38d19c2d04367ed0d503c395ed80f409547047289
|
| 3 |
+
size 5777
|
config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ViTForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "Normal",
|
| 13 |
+
"1": "Stroke"
|
| 14 |
+
},
|
| 15 |
+
"image_size": 224,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 3072,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"Normal": "0",
|
| 20 |
+
"Stroke": "1"
|
| 21 |
+
},
|
| 22 |
+
"layer_norm_eps": 1e-12,
|
| 23 |
+
"model_type": "vit",
|
| 24 |
+
"num_attention_heads": 12,
|
| 25 |
+
"num_channels": 3,
|
| 26 |
+
"num_hidden_layers": 12,
|
| 27 |
+
"patch_size": 16,
|
| 28 |
+
"pooler_act": "tanh",
|
| 29 |
+
"pooler_output_size": 768,
|
| 30 |
+
"problem_type": "single_label_classification",
|
| 31 |
+
"qkv_bias": true,
|
| 32 |
+
"transformers_version": "4.56.1"
|
| 33 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cbf166d41eee416d013784388cb6329a093caa28fe2a65c40b639264d1cb02c
|
| 3 |
+
size 343223968
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"resample": 2,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 4.64951937687552e+17,
|
| 4 |
+
"train_loss": 0.3516346867879232,
|
| 5 |
+
"train_runtime": 232.2268,
|
| 6 |
+
"train_samples_per_second": 25.837,
|
| 7 |
+
"train_steps_per_second": 1.615
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,365 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 350,
|
| 3 |
+
"best_metric": 0.19462016224861145,
|
| 4 |
+
"best_model_checkpoint": "./vit-stroke-detector/checkpoint-350",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08,
|
| 14 |
+
"grad_norm": 0.828596830368042,
|
| 15 |
+
"learning_rate": 0.0001952,
|
| 16 |
+
"loss": 0.6849,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16,
|
| 21 |
+
"grad_norm": 1.6220552921295166,
|
| 22 |
+
"learning_rate": 0.00018986666666666668,
|
| 23 |
+
"loss": 0.6632,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.24,
|
| 28 |
+
"grad_norm": 1.931017279624939,
|
| 29 |
+
"learning_rate": 0.00018453333333333334,
|
| 30 |
+
"loss": 0.6737,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.32,
|
| 35 |
+
"grad_norm": 0.5325367450714111,
|
| 36 |
+
"learning_rate": 0.00017920000000000002,
|
| 37 |
+
"loss": 0.6441,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4,
|
| 42 |
+
"grad_norm": 0.20815162360668182,
|
| 43 |
+
"learning_rate": 0.00017386666666666667,
|
| 44 |
+
"loss": 0.6959,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4,
|
| 49 |
+
"eval_accuracy": 0.6007984031936128,
|
| 50 |
+
"eval_loss": 0.6659401059150696,
|
| 51 |
+
"eval_runtime": 12.6028,
|
| 52 |
+
"eval_samples_per_second": 39.753,
|
| 53 |
+
"eval_steps_per_second": 4.999,
|
| 54 |
+
"step": 50
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.48,
|
| 58 |
+
"grad_norm": 0.6144075393676758,
|
| 59 |
+
"learning_rate": 0.00016853333333333336,
|
| 60 |
+
"loss": 0.6542,
|
| 61 |
+
"step": 60
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.56,
|
| 65 |
+
"grad_norm": 0.3622696101665497,
|
| 66 |
+
"learning_rate": 0.0001632,
|
| 67 |
+
"loss": 0.6336,
|
| 68 |
+
"step": 70
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.64,
|
| 72 |
+
"grad_norm": 1.7294650077819824,
|
| 73 |
+
"learning_rate": 0.00015786666666666666,
|
| 74 |
+
"loss": 0.6705,
|
| 75 |
+
"step": 80
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.72,
|
| 79 |
+
"grad_norm": 1.2596920728683472,
|
| 80 |
+
"learning_rate": 0.00015253333333333335,
|
| 81 |
+
"loss": 0.6144,
|
| 82 |
+
"step": 90
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.8,
|
| 86 |
+
"grad_norm": 2.432494640350342,
|
| 87 |
+
"learning_rate": 0.0001472,
|
| 88 |
+
"loss": 0.6038,
|
| 89 |
+
"step": 100
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.8,
|
| 93 |
+
"eval_accuracy": 0.624750499001996,
|
| 94 |
+
"eval_loss": 0.6349905133247375,
|
| 95 |
+
"eval_runtime": 4.8798,
|
| 96 |
+
"eval_samples_per_second": 102.668,
|
| 97 |
+
"eval_steps_per_second": 12.91,
|
| 98 |
+
"step": 100
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.88,
|
| 102 |
+
"grad_norm": 0.8303850293159485,
|
| 103 |
+
"learning_rate": 0.00014186666666666668,
|
| 104 |
+
"loss": 0.5625,
|
| 105 |
+
"step": 110
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.96,
|
| 109 |
+
"grad_norm": 1.7314858436584473,
|
| 110 |
+
"learning_rate": 0.00013653333333333334,
|
| 111 |
+
"loss": 0.4656,
|
| 112 |
+
"step": 120
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.04,
|
| 116 |
+
"grad_norm": 2.2635459899902344,
|
| 117 |
+
"learning_rate": 0.00013120000000000002,
|
| 118 |
+
"loss": 0.401,
|
| 119 |
+
"step": 130
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.12,
|
| 123 |
+
"grad_norm": 1.9528553485870361,
|
| 124 |
+
"learning_rate": 0.00012586666666666667,
|
| 125 |
+
"loss": 0.5874,
|
| 126 |
+
"step": 140
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.2,
|
| 130 |
+
"grad_norm": 2.2214958667755127,
|
| 131 |
+
"learning_rate": 0.00012053333333333334,
|
| 132 |
+
"loss": 0.4041,
|
| 133 |
+
"step": 150
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.2,
|
| 137 |
+
"eval_accuracy": 0.7904191616766467,
|
| 138 |
+
"eval_loss": 0.5236000418663025,
|
| 139 |
+
"eval_runtime": 6.2426,
|
| 140 |
+
"eval_samples_per_second": 80.255,
|
| 141 |
+
"eval_steps_per_second": 10.092,
|
| 142 |
+
"step": 150
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.28,
|
| 146 |
+
"grad_norm": 1.578341007232666,
|
| 147 |
+
"learning_rate": 0.0001152,
|
| 148 |
+
"loss": 0.4256,
|
| 149 |
+
"step": 160
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3599999999999999,
|
| 153 |
+
"grad_norm": 0.521945059299469,
|
| 154 |
+
"learning_rate": 0.00010986666666666668,
|
| 155 |
+
"loss": 0.3168,
|
| 156 |
+
"step": 170
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.44,
|
| 160 |
+
"grad_norm": 1.1707364320755005,
|
| 161 |
+
"learning_rate": 0.00010453333333333333,
|
| 162 |
+
"loss": 0.4316,
|
| 163 |
+
"step": 180
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.52,
|
| 167 |
+
"grad_norm": 1.4519349336624146,
|
| 168 |
+
"learning_rate": 9.92e-05,
|
| 169 |
+
"loss": 0.2944,
|
| 170 |
+
"step": 190
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.6,
|
| 174 |
+
"grad_norm": 1.0237702131271362,
|
| 175 |
+
"learning_rate": 9.386666666666667e-05,
|
| 176 |
+
"loss": 0.3243,
|
| 177 |
+
"step": 200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6,
|
| 181 |
+
"eval_accuracy": 0.874251497005988,
|
| 182 |
+
"eval_loss": 0.3280556797981262,
|
| 183 |
+
"eval_runtime": 4.9813,
|
| 184 |
+
"eval_samples_per_second": 100.577,
|
| 185 |
+
"eval_steps_per_second": 12.647,
|
| 186 |
+
"step": 200
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.6800000000000002,
|
| 190 |
+
"grad_norm": 4.305712699890137,
|
| 191 |
+
"learning_rate": 8.853333333333333e-05,
|
| 192 |
+
"loss": 0.3118,
|
| 193 |
+
"step": 210
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.76,
|
| 197 |
+
"grad_norm": 1.6635700464248657,
|
| 198 |
+
"learning_rate": 8.32e-05,
|
| 199 |
+
"loss": 0.2534,
|
| 200 |
+
"step": 220
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.8399999999999999,
|
| 204 |
+
"grad_norm": 1.5301051139831543,
|
| 205 |
+
"learning_rate": 7.786666666666667e-05,
|
| 206 |
+
"loss": 0.2702,
|
| 207 |
+
"step": 230
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.92,
|
| 211 |
+
"grad_norm": 0.7002310156822205,
|
| 212 |
+
"learning_rate": 7.253333333333334e-05,
|
| 213 |
+
"loss": 0.1926,
|
| 214 |
+
"step": 240
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 2.0,
|
| 218 |
+
"grad_norm": 0.4949961304664612,
|
| 219 |
+
"learning_rate": 6.720000000000001e-05,
|
| 220 |
+
"loss": 0.2041,
|
| 221 |
+
"step": 250
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.0,
|
| 225 |
+
"eval_accuracy": 0.8922155688622755,
|
| 226 |
+
"eval_loss": 0.28237032890319824,
|
| 227 |
+
"eval_runtime": 4.9486,
|
| 228 |
+
"eval_samples_per_second": 101.242,
|
| 229 |
+
"eval_steps_per_second": 12.731,
|
| 230 |
+
"step": 250
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.08,
|
| 234 |
+
"grad_norm": 13.5546236038208,
|
| 235 |
+
"learning_rate": 6.186666666666668e-05,
|
| 236 |
+
"loss": 0.173,
|
| 237 |
+
"step": 260
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.16,
|
| 241 |
+
"grad_norm": 7.708662986755371,
|
| 242 |
+
"learning_rate": 5.6533333333333336e-05,
|
| 243 |
+
"loss": 0.1206,
|
| 244 |
+
"step": 270
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.24,
|
| 248 |
+
"grad_norm": 0.44631195068359375,
|
| 249 |
+
"learning_rate": 5.1200000000000004e-05,
|
| 250 |
+
"loss": 0.1615,
|
| 251 |
+
"step": 280
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.32,
|
| 255 |
+
"grad_norm": 1.9430702924728394,
|
| 256 |
+
"learning_rate": 4.5866666666666666e-05,
|
| 257 |
+
"loss": 0.1196,
|
| 258 |
+
"step": 290
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.4,
|
| 262 |
+
"grad_norm": 2.0821869373321533,
|
| 263 |
+
"learning_rate": 4.0533333333333334e-05,
|
| 264 |
+
"loss": 0.117,
|
| 265 |
+
"step": 300
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.4,
|
| 269 |
+
"eval_accuracy": 0.9161676646706587,
|
| 270 |
+
"eval_loss": 0.22871683537960052,
|
| 271 |
+
"eval_runtime": 4.8973,
|
| 272 |
+
"eval_samples_per_second": 102.302,
|
| 273 |
+
"eval_steps_per_second": 12.864,
|
| 274 |
+
"step": 300
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.48,
|
| 278 |
+
"grad_norm": 9.634779930114746,
|
| 279 |
+
"learning_rate": 3.52e-05,
|
| 280 |
+
"loss": 0.0606,
|
| 281 |
+
"step": 310
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.56,
|
| 285 |
+
"grad_norm": 6.319236755371094,
|
| 286 |
+
"learning_rate": 2.986666666666667e-05,
|
| 287 |
+
"loss": 0.0774,
|
| 288 |
+
"step": 320
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.64,
|
| 292 |
+
"grad_norm": 9.121068000793457,
|
| 293 |
+
"learning_rate": 2.4533333333333334e-05,
|
| 294 |
+
"loss": 0.06,
|
| 295 |
+
"step": 330
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.7199999999999998,
|
| 299 |
+
"grad_norm": 3.4029245376586914,
|
| 300 |
+
"learning_rate": 1.9200000000000003e-05,
|
| 301 |
+
"loss": 0.0997,
|
| 302 |
+
"step": 340
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.8,
|
| 306 |
+
"grad_norm": 0.8972381353378296,
|
| 307 |
+
"learning_rate": 1.3866666666666667e-05,
|
| 308 |
+
"loss": 0.0353,
|
| 309 |
+
"step": 350
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.8,
|
| 313 |
+
"eval_accuracy": 0.9401197604790419,
|
| 314 |
+
"eval_loss": 0.19462016224861145,
|
| 315 |
+
"eval_runtime": 5.5189,
|
| 316 |
+
"eval_samples_per_second": 90.78,
|
| 317 |
+
"eval_steps_per_second": 11.415,
|
| 318 |
+
"step": 350
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.88,
|
| 322 |
+
"grad_norm": 0.1985252946615219,
|
| 323 |
+
"learning_rate": 8.533333333333334e-06,
|
| 324 |
+
"loss": 0.0833,
|
| 325 |
+
"step": 360
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.96,
|
| 329 |
+
"grad_norm": 0.129100501537323,
|
| 330 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 331 |
+
"loss": 0.0775,
|
| 332 |
+
"step": 370
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.0,
|
| 336 |
+
"step": 375,
|
| 337 |
+
"total_flos": 4.64951937687552e+17,
|
| 338 |
+
"train_loss": 0.3516346867879232,
|
| 339 |
+
"train_runtime": 232.2268,
|
| 340 |
+
"train_samples_per_second": 25.837,
|
| 341 |
+
"train_steps_per_second": 1.615
|
| 342 |
+
}
|
| 343 |
+
],
|
| 344 |
+
"logging_steps": 10,
|
| 345 |
+
"max_steps": 375,
|
| 346 |
+
"num_input_tokens_seen": 0,
|
| 347 |
+
"num_train_epochs": 3,
|
| 348 |
+
"save_steps": 50,
|
| 349 |
+
"stateful_callbacks": {
|
| 350 |
+
"TrainerControl": {
|
| 351 |
+
"args": {
|
| 352 |
+
"should_epoch_stop": false,
|
| 353 |
+
"should_evaluate": false,
|
| 354 |
+
"should_log": false,
|
| 355 |
+
"should_save": true,
|
| 356 |
+
"should_training_stop": true
|
| 357 |
+
},
|
| 358 |
+
"attributes": {}
|
| 359 |
+
}
|
| 360 |
+
},
|
| 361 |
+
"total_flos": 4.64951937687552e+17,
|
| 362 |
+
"train_batch_size": 16,
|
| 363 |
+
"trial_name": null,
|
| 364 |
+
"trial_params": null
|
| 365 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8317ceaa60875571340e1a38d19c2d04367ed0d503c395ed80f409547047289
|
| 3 |
+
size 5777
|