Upload 8 files
Browse files- config.json +42 -0
- optimizer.pt +3 -0
- preprocessor_config.json +22 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +361 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-base-patch16-224-in21k",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "plastic",
|
| 13 |
+
"1": "glass",
|
| 14 |
+
"2": "trash",
|
| 15 |
+
"3": "cardboard",
|
| 16 |
+
"4": ".DS_Store",
|
| 17 |
+
"5": "paper",
|
| 18 |
+
"6": "metal"
|
| 19 |
+
},
|
| 20 |
+
"image_size": 224,
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"label2id": {
|
| 24 |
+
".DS_Store": "4",
|
| 25 |
+
"cardboard": "3",
|
| 26 |
+
"glass": "1",
|
| 27 |
+
"metal": "6",
|
| 28 |
+
"paper": "5",
|
| 29 |
+
"plastic": "0",
|
| 30 |
+
"trash": "2"
|
| 31 |
+
},
|
| 32 |
+
"layer_norm_eps": 1e-12,
|
| 33 |
+
"model_type": "vit",
|
| 34 |
+
"num_attention_heads": 12,
|
| 35 |
+
"num_channels": 3,
|
| 36 |
+
"num_hidden_layers": 12,
|
| 37 |
+
"patch_size": 16,
|
| 38 |
+
"problem_type": "single_label_classification",
|
| 39 |
+
"qkv_bias": true,
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.25.1"
|
| 42 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e03ceb6e24b0b5cf000f1ef927bbf50eba06f9a882e4bed352ea5410017500e3
|
| 3 |
+
size 686544005
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"do_rescale": true,
|
| 4 |
+
"do_resize": true,
|
| 5 |
+
"image_mean": [
|
| 6 |
+
0.5,
|
| 7 |
+
0.5,
|
| 8 |
+
0.5
|
| 9 |
+
],
|
| 10 |
+
"image_processor_type": "ViTImageProcessor",
|
| 11 |
+
"image_std": [
|
| 12 |
+
0.5,
|
| 13 |
+
0.5,
|
| 14 |
+
0.5
|
| 15 |
+
],
|
| 16 |
+
"resample": 2,
|
| 17 |
+
"rescale_factor": 0.00392156862745098,
|
| 18 |
+
"size": {
|
| 19 |
+
"height": 224,
|
| 20 |
+
"width": 224
|
| 21 |
+
}
|
| 22 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d63eac47e2c75876b5a76418b1e3462c88f3d9c9749a7d3f3d80763f825898a
|
| 3 |
+
size 343281325
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7115fcd198f9a024a0ba7f8ad5b53f017f3dbc232c0c00d507816579b5a2d8c5
|
| 3 |
+
size 13553
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cb6ba3c28ea151722bc52e13c7e9e8cdb9aa05bb1331d326ceed3acb5445e3f
|
| 3 |
+
size 627
|
trainer_state.json
ADDED
|
@@ -0,0 +1,361 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.22444555163383484,
|
| 3 |
+
"best_model_checkpoint": "vit-base-tarsh/checkpoint-500",
|
| 4 |
+
"epoch": 3.937007874015748,
|
| 5 |
+
"global_step": 500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.08,
|
| 12 |
+
"learning_rate": 0.00019606299212598428,
|
| 13 |
+
"loss": 1.623,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.16,
|
| 18 |
+
"learning_rate": 0.0001921259842519685,
|
| 19 |
+
"loss": 1.2641,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.24,
|
| 24 |
+
"learning_rate": 0.00018818897637795277,
|
| 25 |
+
"loss": 0.9268,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.31,
|
| 30 |
+
"learning_rate": 0.000184251968503937,
|
| 31 |
+
"loss": 0.7366,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.39,
|
| 36 |
+
"learning_rate": 0.00018031496062992125,
|
| 37 |
+
"loss": 0.7007,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.47,
|
| 42 |
+
"learning_rate": 0.00017637795275590552,
|
| 43 |
+
"loss": 0.557,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.55,
|
| 48 |
+
"learning_rate": 0.00017244094488188977,
|
| 49 |
+
"loss": 0.6111,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.63,
|
| 54 |
+
"learning_rate": 0.000168503937007874,
|
| 55 |
+
"loss": 0.6436,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.71,
|
| 60 |
+
"learning_rate": 0.00016456692913385828,
|
| 61 |
+
"loss": 0.4778,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.79,
|
| 66 |
+
"learning_rate": 0.00016062992125984252,
|
| 67 |
+
"loss": 0.4692,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.79,
|
| 72 |
+
"eval_accuracy": 0.9018567639257294,
|
| 73 |
+
"eval_loss": 0.4172811806201935,
|
| 74 |
+
"eval_runtime": 347.1671,
|
| 75 |
+
"eval_samples_per_second": 1.086,
|
| 76 |
+
"eval_steps_per_second": 0.138,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.87,
|
| 81 |
+
"learning_rate": 0.0001566929133858268,
|
| 82 |
+
"loss": 0.2606,
|
| 83 |
+
"step": 110
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.94,
|
| 87 |
+
"learning_rate": 0.00015275590551181104,
|
| 88 |
+
"loss": 0.2834,
|
| 89 |
+
"step": 120
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 1.02,
|
| 93 |
+
"learning_rate": 0.00014881889763779528,
|
| 94 |
+
"loss": 0.2804,
|
| 95 |
+
"step": 130
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.1,
|
| 99 |
+
"learning_rate": 0.00014488188976377955,
|
| 100 |
+
"loss": 0.2504,
|
| 101 |
+
"step": 140
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.18,
|
| 105 |
+
"learning_rate": 0.00014094488188976377,
|
| 106 |
+
"loss": 0.184,
|
| 107 |
+
"step": 150
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.26,
|
| 111 |
+
"learning_rate": 0.00013700787401574804,
|
| 112 |
+
"loss": 0.213,
|
| 113 |
+
"step": 160
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 1.34,
|
| 117 |
+
"learning_rate": 0.0001330708661417323,
|
| 118 |
+
"loss": 0.2614,
|
| 119 |
+
"step": 170
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.42,
|
| 123 |
+
"learning_rate": 0.00012913385826771653,
|
| 124 |
+
"loss": 0.2567,
|
| 125 |
+
"step": 180
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 1.5,
|
| 129 |
+
"learning_rate": 0.0001251968503937008,
|
| 130 |
+
"loss": 0.1094,
|
| 131 |
+
"step": 190
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.57,
|
| 135 |
+
"learning_rate": 0.00012125984251968505,
|
| 136 |
+
"loss": 0.2064,
|
| 137 |
+
"step": 200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.57,
|
| 141 |
+
"eval_accuracy": 0.9257294429708223,
|
| 142 |
+
"eval_loss": 0.29030969738960266,
|
| 143 |
+
"eval_runtime": 349.8071,
|
| 144 |
+
"eval_samples_per_second": 1.078,
|
| 145 |
+
"eval_steps_per_second": 0.137,
|
| 146 |
+
"step": 200
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.65,
|
| 150 |
+
"learning_rate": 0.00011732283464566928,
|
| 151 |
+
"loss": 0.1707,
|
| 152 |
+
"step": 210
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.73,
|
| 156 |
+
"learning_rate": 0.00011338582677165355,
|
| 157 |
+
"loss": 0.2581,
|
| 158 |
+
"step": 220
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.81,
|
| 162 |
+
"learning_rate": 0.00010944881889763781,
|
| 163 |
+
"loss": 0.0885,
|
| 164 |
+
"step": 230
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.89,
|
| 168 |
+
"learning_rate": 0.00010551181102362204,
|
| 169 |
+
"loss": 0.1996,
|
| 170 |
+
"step": 240
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.97,
|
| 174 |
+
"learning_rate": 0.0001015748031496063,
|
| 175 |
+
"loss": 0.1614,
|
| 176 |
+
"step": 250
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 2.05,
|
| 180 |
+
"learning_rate": 9.763779527559055e-05,
|
| 181 |
+
"loss": 0.1308,
|
| 182 |
+
"step": 260
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 2.13,
|
| 186 |
+
"learning_rate": 9.370078740157481e-05,
|
| 187 |
+
"loss": 0.0973,
|
| 188 |
+
"step": 270
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.2,
|
| 192 |
+
"learning_rate": 8.976377952755905e-05,
|
| 193 |
+
"loss": 0.0764,
|
| 194 |
+
"step": 280
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.28,
|
| 198 |
+
"learning_rate": 8.582677165354331e-05,
|
| 199 |
+
"loss": 0.039,
|
| 200 |
+
"step": 290
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.36,
|
| 204 |
+
"learning_rate": 8.188976377952757e-05,
|
| 205 |
+
"loss": 0.0748,
|
| 206 |
+
"step": 300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.36,
|
| 210 |
+
"eval_accuracy": 0.9363395225464191,
|
| 211 |
+
"eval_loss": 0.2725047767162323,
|
| 212 |
+
"eval_runtime": 349.7967,
|
| 213 |
+
"eval_samples_per_second": 1.078,
|
| 214 |
+
"eval_steps_per_second": 0.137,
|
| 215 |
+
"step": 300
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 2.44,
|
| 219 |
+
"learning_rate": 7.795275590551181e-05,
|
| 220 |
+
"loss": 0.0508,
|
| 221 |
+
"step": 310
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.52,
|
| 225 |
+
"learning_rate": 7.401574803149607e-05,
|
| 226 |
+
"loss": 0.0781,
|
| 227 |
+
"step": 320
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 2.6,
|
| 231 |
+
"learning_rate": 7.007874015748031e-05,
|
| 232 |
+
"loss": 0.1238,
|
| 233 |
+
"step": 330
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.68,
|
| 237 |
+
"learning_rate": 6.614173228346457e-05,
|
| 238 |
+
"loss": 0.0506,
|
| 239 |
+
"step": 340
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 2.76,
|
| 243 |
+
"learning_rate": 6.220472440944882e-05,
|
| 244 |
+
"loss": 0.05,
|
| 245 |
+
"step": 350
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 2.83,
|
| 249 |
+
"learning_rate": 5.826771653543307e-05,
|
| 250 |
+
"loss": 0.0748,
|
| 251 |
+
"step": 360
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.91,
|
| 255 |
+
"learning_rate": 5.433070866141733e-05,
|
| 256 |
+
"loss": 0.0257,
|
| 257 |
+
"step": 370
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 2.99,
|
| 261 |
+
"learning_rate": 5.0393700787401575e-05,
|
| 262 |
+
"loss": 0.0861,
|
| 263 |
+
"step": 380
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 3.07,
|
| 267 |
+
"learning_rate": 4.645669291338583e-05,
|
| 268 |
+
"loss": 0.029,
|
| 269 |
+
"step": 390
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 3.15,
|
| 273 |
+
"learning_rate": 4.251968503937008e-05,
|
| 274 |
+
"loss": 0.0343,
|
| 275 |
+
"step": 400
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 3.15,
|
| 279 |
+
"eval_accuracy": 0.9283819628647215,
|
| 280 |
+
"eval_loss": 0.2931118607521057,
|
| 281 |
+
"eval_runtime": 343.908,
|
| 282 |
+
"eval_samples_per_second": 1.096,
|
| 283 |
+
"eval_steps_per_second": 0.14,
|
| 284 |
+
"step": 400
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 3.23,
|
| 288 |
+
"learning_rate": 3.858267716535433e-05,
|
| 289 |
+
"loss": 0.0234,
|
| 290 |
+
"step": 410
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.31,
|
| 294 |
+
"learning_rate": 3.464566929133858e-05,
|
| 295 |
+
"loss": 0.0208,
|
| 296 |
+
"step": 420
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 3.39,
|
| 300 |
+
"learning_rate": 3.070866141732284e-05,
|
| 301 |
+
"loss": 0.0186,
|
| 302 |
+
"step": 430
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.46,
|
| 306 |
+
"learning_rate": 2.677165354330709e-05,
|
| 307 |
+
"loss": 0.0398,
|
| 308 |
+
"step": 440
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.54,
|
| 312 |
+
"learning_rate": 2.283464566929134e-05,
|
| 313 |
+
"loss": 0.0262,
|
| 314 |
+
"step": 450
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.62,
|
| 318 |
+
"learning_rate": 1.889763779527559e-05,
|
| 319 |
+
"loss": 0.0482,
|
| 320 |
+
"step": 460
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.7,
|
| 324 |
+
"learning_rate": 1.4960629921259845e-05,
|
| 325 |
+
"loss": 0.0191,
|
| 326 |
+
"step": 470
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.78,
|
| 330 |
+
"learning_rate": 1.1023622047244095e-05,
|
| 331 |
+
"loss": 0.02,
|
| 332 |
+
"step": 480
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.86,
|
| 336 |
+
"learning_rate": 7.086614173228347e-06,
|
| 337 |
+
"loss": 0.0501,
|
| 338 |
+
"step": 490
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.94,
|
| 342 |
+
"learning_rate": 3.1496062992125985e-06,
|
| 343 |
+
"loss": 0.0176,
|
| 344 |
+
"step": 500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.94,
|
| 348 |
+
"eval_accuracy": 0.9442970822281167,
|
| 349 |
+
"eval_loss": 0.22444555163383484,
|
| 350 |
+
"eval_runtime": 342.5831,
|
| 351 |
+
"eval_samples_per_second": 1.1,
|
| 352 |
+
"eval_steps_per_second": 0.14,
|
| 353 |
+
"step": 500
|
| 354 |
+
}
|
| 355 |
+
],
|
| 356 |
+
"max_steps": 508,
|
| 357 |
+
"num_train_epochs": 4,
|
| 358 |
+
"total_flos": 6.169413753348895e+17,
|
| 359 |
+
"trial_name": null,
|
| 360 |
+
"trial_params": null
|
| 361 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2976684c675b1fa729f4d41a7428291976937fa933a1accdb268533f66efd1e3
|
| 3 |
+
size 3387
|