End of training
Browse files- README.md +14 -13
- all_results.json +9 -9
- config.json +8 -8
- eval_results.json +5 -5
- model.safetensors +1 -1
- runs/Aug01_12-35-15_3c8cb0b5712e/events.out.tfevents.1722515747.3c8cb0b5712e.503.0 +3 -0
- runs/Aug01_12-35-15_3c8cb0b5712e/events.out.tfevents.1722516355.3c8cb0b5712e.503.1 +3 -0
- train_results.json +4 -4
- trainer_state.json +255 -255
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -2,6 +2,7 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -15,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# vit-base-images
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -50,16 +51,16 @@ The following hyperparameters were used during training:
|
|
| 50 |
|
| 51 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 52 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 53 |
-
| 0.
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
| 62 |
-
| 0.
|
| 63 |
|
| 64 |
|
| 65 |
### Framework versions
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
|
|
| 16 |
|
| 17 |
# vit-base-images
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the marmal88/skin_cancer dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.0918
|
| 22 |
+
- Accuracy: 0.981
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 54 |
+
| 0.8785 | 0.4 | 100 | 0.7795 | 0.711 |
|
| 55 |
+
| 0.7076 | 0.8 | 200 | 0.5421 | 0.818 |
|
| 56 |
+
| 0.4283 | 1.2 | 300 | 0.3951 | 0.876 |
|
| 57 |
+
| 0.4251 | 1.6 | 400 | 0.3818 | 0.864 |
|
| 58 |
+
| 0.335 | 2.0 | 500 | 0.2474 | 0.924 |
|
| 59 |
+
| 0.2286 | 2.4 | 600 | 0.1675 | 0.952 |
|
| 60 |
+
| 0.1523 | 2.8 | 700 | 0.1641 | 0.954 |
|
| 61 |
+
| 0.1346 | 3.2 | 800 | 0.1120 | 0.969 |
|
| 62 |
+
| 0.0638 | 3.6 | 900 | 0.1025 | 0.978 |
|
| 63 |
+
| 0.0574 | 4.0 | 1000 | 0.0918 | 0.981 |
|
| 64 |
|
| 65 |
|
| 66 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
"total_flos": 1.239905171570688e+18,
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 1.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.981,
|
| 4 |
+
"eval_loss": 0.09177211672067642,
|
| 5 |
+
"eval_runtime": 15.1575,
|
| 6 |
+
"eval_samples_per_second": 65.974,
|
| 7 |
+
"eval_steps_per_second": 8.247,
|
| 8 |
"total_flos": 1.239905171570688e+18,
|
| 9 |
+
"train_loss": 0.3459155881404877,
|
| 10 |
+
"train_runtime": 562.8746,
|
| 11 |
+
"train_samples_per_second": 28.426,
|
| 12 |
+
"train_steps_per_second": 1.777
|
| 13 |
}
|
config.json
CHANGED
|
@@ -9,21 +9,21 @@
|
|
| 9 |
"hidden_dropout_prob": 0.0,
|
| 10 |
"hidden_size": 768,
|
| 11 |
"id2label": {
|
| 12 |
-
"0": "
|
| 13 |
-
"1": "
|
| 14 |
-
"2": "
|
| 15 |
"3": "dermatofibroma",
|
| 16 |
-
"4": "
|
| 17 |
},
|
| 18 |
"image_size": 224,
|
| 19 |
"initializer_range": 0.02,
|
| 20 |
"intermediate_size": 3072,
|
| 21 |
"label2id": {
|
| 22 |
-
"actinic_keratoses": "
|
| 23 |
-
"basal_cell_carcinoma": "
|
| 24 |
-
"benign_keratosis-like_lesions": "
|
| 25 |
"dermatofibroma": "3",
|
| 26 |
-
"melanocytic_Nevi": "
|
| 27 |
},
|
| 28 |
"layer_norm_eps": 1e-12,
|
| 29 |
"model_type": "vit",
|
|
|
|
| 9 |
"hidden_dropout_prob": 0.0,
|
| 10 |
"hidden_size": 768,
|
| 11 |
"id2label": {
|
| 12 |
+
"0": "benign_keratosis-like_lesions",
|
| 13 |
+
"1": "basal_cell_carcinoma",
|
| 14 |
+
"2": "actinic_keratoses",
|
| 15 |
"3": "dermatofibroma",
|
| 16 |
+
"4": "melanocytic_Nevi"
|
| 17 |
},
|
| 18 |
"image_size": 224,
|
| 19 |
"initializer_range": 0.02,
|
| 20 |
"intermediate_size": 3072,
|
| 21 |
"label2id": {
|
| 22 |
+
"actinic_keratoses": "2",
|
| 23 |
+
"basal_cell_carcinoma": "1",
|
| 24 |
+
"benign_keratosis-like_lesions": "0",
|
| 25 |
"dermatofibroma": "3",
|
| 26 |
+
"melanocytic_Nevi": "4"
|
| 27 |
},
|
| 28 |
"layer_norm_eps": 1e-12,
|
| 29 |
"model_type": "vit",
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.981,
|
| 4 |
+
"eval_loss": 0.09177211672067642,
|
| 5 |
+
"eval_runtime": 15.1575,
|
| 6 |
+
"eval_samples_per_second": 65.974,
|
| 7 |
+
"eval_steps_per_second": 8.247
|
| 8 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 343233204
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a857093be94afc442b9b7de17ca6322749916293529da0a7d0c659ac6aae89e0
|
| 3 |
size 343233204
|
runs/Aug01_12-35-15_3c8cb0b5712e/events.out.tfevents.1722515747.3c8cb0b5712e.503.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b76bef5ca525d67c1ea03da9dde2800a9daed78991fc38127a05ec85f52dd32c
|
| 3 |
+
size 29704
|
runs/Aug01_12-35-15_3c8cb0b5712e/events.out.tfevents.1722516355.3c8cb0b5712e.503.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7b156180a1f6f187a556dc3a746cc4aba1daa990ceb96032c97066cd12133e0
|
| 3 |
+
size 411
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
"total_flos": 1.239905171570688e+18,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
"total_flos": 1.239905171570688e+18,
|
| 4 |
+
"train_loss": 0.3459155881404877,
|
| 5 |
+
"train_runtime": 562.8746,
|
| 6 |
+
"train_samples_per_second": 28.426,
|
| 7 |
+
"train_steps_per_second": 1.777
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "./vit-base-images/checkpoint-1000",
|
| 4 |
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
|
@@ -10,802 +10,802 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.04,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 0.00019800000000000002,
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.08,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 0.000196,
|
| 22 |
-
"loss":
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.12,
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
"learning_rate": 0.000194,
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.16,
|
| 34 |
-
"grad_norm": 1.
|
| 35 |
"learning_rate": 0.000192,
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.2,
|
| 41 |
-
"grad_norm":
|
| 42 |
"learning_rate": 0.00019,
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.24,
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 0.000188,
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.28,
|
| 55 |
-
"grad_norm":
|
| 56 |
"learning_rate": 0.00018600000000000002,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.32,
|
| 62 |
-
"grad_norm":
|
| 63 |
"learning_rate": 0.00018400000000000003,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.36,
|
| 69 |
-
"grad_norm": 2.
|
| 70 |
"learning_rate": 0.000182,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.4,
|
| 76 |
-
"grad_norm": 2.
|
| 77 |
"learning_rate": 0.00018,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.4,
|
| 83 |
-
"eval_accuracy": 0.
|
| 84 |
-
"eval_loss": 0.
|
| 85 |
-
"eval_runtime":
|
| 86 |
-
"eval_samples_per_second":
|
| 87 |
-
"eval_steps_per_second":
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
{
|
| 91 |
"epoch": 0.44,
|
| 92 |
-
"grad_norm": 1.
|
| 93 |
"learning_rate": 0.00017800000000000002,
|
| 94 |
-
"loss": 0.
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 0.48,
|
| 99 |
-
"grad_norm": 1.
|
| 100 |
"learning_rate": 0.00017600000000000002,
|
| 101 |
-
"loss": 0.
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 0.52,
|
| 106 |
-
"grad_norm":
|
| 107 |
"learning_rate": 0.000174,
|
| 108 |
-
"loss": 0.
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 0.56,
|
| 113 |
-
"grad_norm":
|
| 114 |
"learning_rate": 0.000172,
|
| 115 |
-
"loss": 0.
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.6,
|
| 120 |
-
"grad_norm":
|
| 121 |
"learning_rate": 0.00017,
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 0.64,
|
| 127 |
-
"grad_norm":
|
| 128 |
"learning_rate": 0.000168,
|
| 129 |
-
"loss": 0.
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 0.68,
|
| 134 |
-
"grad_norm":
|
| 135 |
"learning_rate": 0.000166,
|
| 136 |
-
"loss": 0.
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 0.72,
|
| 141 |
-
"grad_norm":
|
| 142 |
"learning_rate": 0.000164,
|
| 143 |
-
"loss": 0.
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.76,
|
| 148 |
-
"grad_norm":
|
| 149 |
"learning_rate": 0.000162,
|
| 150 |
-
"loss": 0.
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 0.8,
|
| 155 |
-
"grad_norm":
|
| 156 |
"learning_rate": 0.00016,
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.8,
|
| 162 |
-
"eval_accuracy": 0.
|
| 163 |
-
"eval_loss": 0.
|
| 164 |
-
"eval_runtime": 14.
|
| 165 |
-
"eval_samples_per_second":
|
| 166 |
-
"eval_steps_per_second": 8.
|
| 167 |
"step": 200
|
| 168 |
},
|
| 169 |
{
|
| 170 |
"epoch": 0.84,
|
| 171 |
-
"grad_norm":
|
| 172 |
"learning_rate": 0.00015800000000000002,
|
| 173 |
-
"loss": 0.
|
| 174 |
"step": 210
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.88,
|
| 178 |
-
"grad_norm":
|
| 179 |
"learning_rate": 0.00015600000000000002,
|
| 180 |
-
"loss": 0.
|
| 181 |
"step": 220
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 0.92,
|
| 185 |
-
"grad_norm":
|
| 186 |
"learning_rate": 0.000154,
|
| 187 |
-
"loss": 0.
|
| 188 |
"step": 230
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.96,
|
| 192 |
-
"grad_norm": 1.
|
| 193 |
"learning_rate": 0.000152,
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 240
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 1.0,
|
| 199 |
-
"grad_norm": 1.
|
| 200 |
"learning_rate": 0.00015000000000000001,
|
| 201 |
-
"loss": 0.
|
| 202 |
"step": 250
|
| 203 |
},
|
| 204 |
{
|
| 205 |
"epoch": 1.04,
|
| 206 |
-
"grad_norm":
|
| 207 |
"learning_rate": 0.000148,
|
| 208 |
-
"loss": 0.
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 1.08,
|
| 213 |
-
"grad_norm":
|
| 214 |
"learning_rate": 0.000146,
|
| 215 |
-
"loss": 0.
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
"epoch": 1.12,
|
| 220 |
-
"grad_norm": 1.
|
| 221 |
"learning_rate": 0.000144,
|
| 222 |
-
"loss": 0.
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
"epoch": 1.16,
|
| 227 |
-
"grad_norm": 1.
|
| 228 |
"learning_rate": 0.000142,
|
| 229 |
-
"loss": 0.
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 1.2,
|
| 234 |
-
"grad_norm":
|
| 235 |
"learning_rate": 0.00014,
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 1.2,
|
| 241 |
-
"eval_accuracy": 0.
|
| 242 |
-
"eval_loss": 0.
|
| 243 |
-
"eval_runtime": 14.
|
| 244 |
-
"eval_samples_per_second": 70.
|
| 245 |
-
"eval_steps_per_second": 8.
|
| 246 |
"step": 300
|
| 247 |
},
|
| 248 |
{
|
| 249 |
"epoch": 1.24,
|
| 250 |
-
"grad_norm":
|
| 251 |
"learning_rate": 0.000138,
|
| 252 |
-
"loss": 0.
|
| 253 |
"step": 310
|
| 254 |
},
|
| 255 |
{
|
| 256 |
"epoch": 1.28,
|
| 257 |
-
"grad_norm": 1.
|
| 258 |
"learning_rate": 0.00013600000000000003,
|
| 259 |
-
"loss": 0.
|
| 260 |
"step": 320
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 1.32,
|
| 264 |
-
"grad_norm":
|
| 265 |
"learning_rate": 0.000134,
|
| 266 |
-
"loss": 0.
|
| 267 |
"step": 330
|
| 268 |
},
|
| 269 |
{
|
| 270 |
"epoch": 1.3599999999999999,
|
| 271 |
-
"grad_norm":
|
| 272 |
"learning_rate": 0.000132,
|
| 273 |
-
"loss": 0.
|
| 274 |
"step": 340
|
| 275 |
},
|
| 276 |
{
|
| 277 |
"epoch": 1.4,
|
| 278 |
-
"grad_norm":
|
| 279 |
"learning_rate": 0.00013000000000000002,
|
| 280 |
-
"loss": 0.
|
| 281 |
"step": 350
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 1.44,
|
| 285 |
-
"grad_norm":
|
| 286 |
"learning_rate": 0.00012800000000000002,
|
| 287 |
-
"loss": 0.
|
| 288 |
"step": 360
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"epoch": 1.48,
|
| 292 |
-
"grad_norm":
|
| 293 |
"learning_rate": 0.000126,
|
| 294 |
-
"loss": 0.
|
| 295 |
"step": 370
|
| 296 |
},
|
| 297 |
{
|
| 298 |
"epoch": 1.52,
|
| 299 |
-
"grad_norm":
|
| 300 |
"learning_rate": 0.000124,
|
| 301 |
-
"loss": 0.
|
| 302 |
"step": 380
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 1.56,
|
| 306 |
-
"grad_norm": 1.
|
| 307 |
"learning_rate": 0.000122,
|
| 308 |
-
"loss": 0.
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 1.6,
|
| 313 |
-
"grad_norm": 3.
|
| 314 |
"learning_rate": 0.00012,
|
| 315 |
-
"loss": 0.
|
| 316 |
"step": 400
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 1.6,
|
| 320 |
-
"eval_accuracy": 0.
|
| 321 |
-
"eval_loss": 0.
|
| 322 |
-
"eval_runtime": 14.
|
| 323 |
-
"eval_samples_per_second":
|
| 324 |
-
"eval_steps_per_second": 8.
|
| 325 |
"step": 400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.6400000000000001,
|
| 329 |
-
"grad_norm":
|
| 330 |
"learning_rate": 0.000118,
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 410
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.6800000000000002,
|
| 336 |
-
"grad_norm":
|
| 337 |
"learning_rate": 0.000116,
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 420
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.72,
|
| 343 |
-
"grad_norm":
|
| 344 |
"learning_rate": 0.00011399999999999999,
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 430
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.76,
|
| 350 |
-
"grad_norm":
|
| 351 |
"learning_rate": 0.00011200000000000001,
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 440
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.8,
|
| 357 |
-
"grad_norm":
|
| 358 |
"learning_rate": 0.00011000000000000002,
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 450
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.8399999999999999,
|
| 364 |
-
"grad_norm":
|
| 365 |
"learning_rate": 0.00010800000000000001,
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 460
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.88,
|
| 371 |
-
"grad_norm":
|
| 372 |
"learning_rate": 0.00010600000000000002,
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 470
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.92,
|
| 378 |
-
"grad_norm": 2.
|
| 379 |
"learning_rate": 0.00010400000000000001,
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 480
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.96,
|
| 385 |
-
"grad_norm":
|
| 386 |
"learning_rate": 0.00010200000000000001,
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 490
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 2.0,
|
| 392 |
-
"grad_norm":
|
| 393 |
"learning_rate": 0.0001,
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 2.0,
|
| 399 |
-
"eval_accuracy": 0.
|
| 400 |
-
"eval_loss": 0.
|
| 401 |
-
"eval_runtime":
|
| 402 |
-
"eval_samples_per_second":
|
| 403 |
-
"eval_steps_per_second": 8.
|
| 404 |
"step": 500
|
| 405 |
},
|
| 406 |
{
|
| 407 |
"epoch": 2.04,
|
| 408 |
-
"grad_norm":
|
| 409 |
"learning_rate": 9.8e-05,
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 510
|
| 412 |
},
|
| 413 |
{
|
| 414 |
"epoch": 2.08,
|
| 415 |
-
"grad_norm":
|
| 416 |
"learning_rate": 9.6e-05,
|
| 417 |
-
"loss": 0.
|
| 418 |
"step": 520
|
| 419 |
},
|
| 420 |
{
|
| 421 |
"epoch": 2.12,
|
| 422 |
-
"grad_norm":
|
| 423 |
"learning_rate": 9.4e-05,
|
| 424 |
-
"loss": 0.
|
| 425 |
"step": 530
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"epoch": 2.16,
|
| 429 |
-
"grad_norm":
|
| 430 |
"learning_rate": 9.200000000000001e-05,
|
| 431 |
-
"loss": 0.
|
| 432 |
"step": 540
|
| 433 |
},
|
| 434 |
{
|
| 435 |
"epoch": 2.2,
|
| 436 |
-
"grad_norm":
|
| 437 |
"learning_rate": 9e-05,
|
| 438 |
-
"loss": 0.
|
| 439 |
"step": 550
|
| 440 |
},
|
| 441 |
{
|
| 442 |
"epoch": 2.24,
|
| 443 |
-
"grad_norm":
|
| 444 |
"learning_rate": 8.800000000000001e-05,
|
| 445 |
-
"loss": 0.
|
| 446 |
"step": 560
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 2.2800000000000002,
|
| 450 |
-
"grad_norm":
|
| 451 |
"learning_rate": 8.6e-05,
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 570
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 2.32,
|
| 457 |
-
"grad_norm":
|
| 458 |
"learning_rate": 8.4e-05,
|
| 459 |
-
"loss": 0.
|
| 460 |
"step": 580
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 2.36,
|
| 464 |
-
"grad_norm":
|
| 465 |
"learning_rate": 8.2e-05,
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 590
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 2.4,
|
| 471 |
-
"grad_norm":
|
| 472 |
"learning_rate": 8e-05,
|
| 473 |
-
"loss": 0.
|
| 474 |
"step": 600
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 2.4,
|
| 478 |
-
"eval_accuracy": 0.
|
| 479 |
-
"eval_loss": 0.
|
| 480 |
-
"eval_runtime":
|
| 481 |
-
"eval_samples_per_second":
|
| 482 |
-
"eval_steps_per_second":
|
| 483 |
"step": 600
|
| 484 |
},
|
| 485 |
{
|
| 486 |
"epoch": 2.44,
|
| 487 |
-
"grad_norm":
|
| 488 |
"learning_rate": 7.800000000000001e-05,
|
| 489 |
-
"loss": 0.
|
| 490 |
"step": 610
|
| 491 |
},
|
| 492 |
{
|
| 493 |
"epoch": 2.48,
|
| 494 |
-
"grad_norm":
|
| 495 |
"learning_rate": 7.6e-05,
|
| 496 |
-
"loss": 0.
|
| 497 |
"step": 620
|
| 498 |
},
|
| 499 |
{
|
| 500 |
"epoch": 2.52,
|
| 501 |
-
"grad_norm":
|
| 502 |
"learning_rate": 7.4e-05,
|
| 503 |
-
"loss": 0.
|
| 504 |
"step": 630
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 2.56,
|
| 508 |
-
"grad_norm":
|
| 509 |
"learning_rate": 7.2e-05,
|
| 510 |
-
"loss": 0.
|
| 511 |
"step": 640
|
| 512 |
},
|
| 513 |
{
|
| 514 |
"epoch": 2.6,
|
| 515 |
-
"grad_norm":
|
| 516 |
"learning_rate": 7e-05,
|
| 517 |
-
"loss": 0.
|
| 518 |
"step": 650
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 2.64,
|
| 522 |
-
"grad_norm":
|
| 523 |
"learning_rate": 6.800000000000001e-05,
|
| 524 |
-
"loss": 0.
|
| 525 |
"step": 660
|
| 526 |
},
|
| 527 |
{
|
| 528 |
"epoch": 2.68,
|
| 529 |
-
"grad_norm":
|
| 530 |
"learning_rate": 6.6e-05,
|
| 531 |
-
"loss": 0.
|
| 532 |
"step": 670
|
| 533 |
},
|
| 534 |
{
|
| 535 |
"epoch": 2.7199999999999998,
|
| 536 |
-
"grad_norm":
|
| 537 |
"learning_rate": 6.400000000000001e-05,
|
| 538 |
-
"loss": 0.
|
| 539 |
"step": 680
|
| 540 |
},
|
| 541 |
{
|
| 542 |
"epoch": 2.76,
|
| 543 |
-
"grad_norm":
|
| 544 |
"learning_rate": 6.2e-05,
|
| 545 |
-
"loss": 0.
|
| 546 |
"step": 690
|
| 547 |
},
|
| 548 |
{
|
| 549 |
"epoch": 2.8,
|
| 550 |
-
"grad_norm": 5.
|
| 551 |
"learning_rate": 6e-05,
|
| 552 |
-
"loss": 0.
|
| 553 |
"step": 700
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"epoch": 2.8,
|
| 557 |
-
"eval_accuracy": 0.
|
| 558 |
-
"eval_loss": 0.
|
| 559 |
-
"eval_runtime": 14.
|
| 560 |
-
"eval_samples_per_second":
|
| 561 |
-
"eval_steps_per_second": 8.
|
| 562 |
"step": 700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 2.84,
|
| 566 |
-
"grad_norm":
|
| 567 |
"learning_rate": 5.8e-05,
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 710
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 2.88,
|
| 573 |
-
"grad_norm":
|
| 574 |
"learning_rate": 5.6000000000000006e-05,
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 720
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 2.92,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 5.4000000000000005e-05,
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 730
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 2.96,
|
| 587 |
-
"grad_norm":
|
| 588 |
"learning_rate": 5.2000000000000004e-05,
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 740
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 3.0,
|
| 594 |
-
"grad_norm":
|
| 595 |
"learning_rate": 5e-05,
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 750
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 3.04,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 4.8e-05,
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 760
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 3.08,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 4.600000000000001e-05,
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 770
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 3.12,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 4.4000000000000006e-05,
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 780
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 3.16,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
"learning_rate": 4.2e-05,
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 790
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 3.2,
|
| 629 |
-
"grad_norm":
|
| 630 |
"learning_rate": 4e-05,
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 800
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 3.2,
|
| 636 |
-
"eval_accuracy": 0.
|
| 637 |
-
"eval_loss": 0.
|
| 638 |
-
"eval_runtime": 14.
|
| 639 |
-
"eval_samples_per_second": 70.
|
| 640 |
-
"eval_steps_per_second": 8.
|
| 641 |
"step": 800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 3.24,
|
| 645 |
-
"grad_norm": 0.
|
| 646 |
"learning_rate": 3.8e-05,
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 810
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 3.2800000000000002,
|
| 652 |
-
"grad_norm":
|
| 653 |
"learning_rate": 3.6e-05,
|
| 654 |
-
"loss": 0.
|
| 655 |
"step": 820
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 3.32,
|
| 659 |
-
"grad_norm": 0.
|
| 660 |
"learning_rate": 3.4000000000000007e-05,
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 830
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 3.36,
|
| 666 |
-
"grad_norm": 0.
|
| 667 |
"learning_rate": 3.2000000000000005e-05,
|
| 668 |
-
"loss": 0.
|
| 669 |
"step": 840
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 3.4,
|
| 673 |
-
"grad_norm": 0.
|
| 674 |
"learning_rate": 3e-05,
|
| 675 |
-
"loss": 0.
|
| 676 |
"step": 850
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 3.44,
|
| 680 |
-
"grad_norm":
|
| 681 |
"learning_rate": 2.8000000000000003e-05,
|
| 682 |
-
"loss": 0.
|
| 683 |
"step": 860
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 3.48,
|
| 687 |
-
"grad_norm":
|
| 688 |
"learning_rate": 2.6000000000000002e-05,
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 870
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 3.52,
|
| 694 |
-
"grad_norm":
|
| 695 |
"learning_rate": 2.4e-05,
|
| 696 |
-
"loss": 0.
|
| 697 |
"step": 880
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 3.56,
|
| 701 |
-
"grad_norm": 0.
|
| 702 |
"learning_rate": 2.2000000000000003e-05,
|
| 703 |
-
"loss": 0.
|
| 704 |
"step": 890
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 3.6,
|
| 708 |
-
"grad_norm":
|
| 709 |
"learning_rate": 2e-05,
|
| 710 |
-
"loss": 0.
|
| 711 |
"step": 900
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 3.6,
|
| 715 |
-
"eval_accuracy": 0.
|
| 716 |
-
"eval_loss": 0.
|
| 717 |
-
"eval_runtime": 14.
|
| 718 |
-
"eval_samples_per_second": 69.
|
| 719 |
-
"eval_steps_per_second": 8.
|
| 720 |
"step": 900
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 3.64,
|
| 724 |
-
"grad_norm":
|
| 725 |
"learning_rate": 1.8e-05,
|
| 726 |
-
"loss": 0.
|
| 727 |
"step": 910
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 3.68,
|
| 731 |
-
"grad_norm": 0.
|
| 732 |
"learning_rate": 1.6000000000000003e-05,
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 920
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 3.7199999999999998,
|
| 738 |
-
"grad_norm": 0.
|
| 739 |
"learning_rate": 1.4000000000000001e-05,
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 930
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 3.76,
|
| 745 |
-
"grad_norm": 0.
|
| 746 |
"learning_rate": 1.2e-05,
|
| 747 |
-
"loss": 0.
|
| 748 |
"step": 940
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 3.8,
|
| 752 |
-
"grad_norm": 0.
|
| 753 |
"learning_rate": 1e-05,
|
| 754 |
-
"loss": 0.
|
| 755 |
"step": 950
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 3.84,
|
| 759 |
-
"grad_norm": 0.
|
| 760 |
"learning_rate": 8.000000000000001e-06,
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 960
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 3.88,
|
| 766 |
-
"grad_norm": 0.
|
| 767 |
"learning_rate": 6e-06,
|
| 768 |
-
"loss": 0.
|
| 769 |
"step": 970
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 3.92,
|
| 773 |
-
"grad_norm": 0.
|
| 774 |
"learning_rate": 4.000000000000001e-06,
|
| 775 |
-
"loss": 0.
|
| 776 |
"step": 980
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 3.96,
|
| 780 |
-
"grad_norm": 0.
|
| 781 |
"learning_rate": 2.0000000000000003e-06,
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 990
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 4.0,
|
| 787 |
-
"grad_norm": 0.
|
| 788 |
"learning_rate": 0.0,
|
| 789 |
-
"loss": 0.
|
| 790 |
"step": 1000
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 4.0,
|
| 794 |
-
"eval_accuracy": 0.
|
| 795 |
-
"eval_loss": 0.
|
| 796 |
-
"eval_runtime":
|
| 797 |
-
"eval_samples_per_second":
|
| 798 |
-
"eval_steps_per_second": 8.
|
| 799 |
"step": 1000
|
| 800 |
},
|
| 801 |
{
|
| 802 |
"epoch": 4.0,
|
| 803 |
"step": 1000,
|
| 804 |
"total_flos": 1.239905171570688e+18,
|
| 805 |
-
"train_loss": 0.
|
| 806 |
-
"train_runtime":
|
| 807 |
-
"train_samples_per_second":
|
| 808 |
-
"train_steps_per_second": 1.
|
| 809 |
}
|
| 810 |
],
|
| 811 |
"logging_steps": 10,
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.09177211672067642,
|
| 3 |
"best_model_checkpoint": "./vit-base-images/checkpoint-1000",
|
| 4 |
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.04,
|
| 13 |
+
"grad_norm": 2.1739912033081055,
|
| 14 |
"learning_rate": 0.00019800000000000002,
|
| 15 |
+
"loss": 1.2716,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.08,
|
| 20 |
+
"grad_norm": 2.4669973850250244,
|
| 21 |
"learning_rate": 0.000196,
|
| 22 |
+
"loss": 1.0636,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.12,
|
| 27 |
+
"grad_norm": 1.3152661323547363,
|
| 28 |
"learning_rate": 0.000194,
|
| 29 |
+
"loss": 0.8686,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.16,
|
| 34 |
+
"grad_norm": 1.30520761013031,
|
| 35 |
"learning_rate": 0.000192,
|
| 36 |
+
"loss": 0.8709,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.2,
|
| 41 |
+
"grad_norm": 3.462522506713867,
|
| 42 |
"learning_rate": 0.00019,
|
| 43 |
+
"loss": 0.8241,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.24,
|
| 48 |
+
"grad_norm": 2.773179531097412,
|
| 49 |
"learning_rate": 0.000188,
|
| 50 |
+
"loss": 0.743,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.28,
|
| 55 |
+
"grad_norm": 0.8740523457527161,
|
| 56 |
"learning_rate": 0.00018600000000000002,
|
| 57 |
+
"loss": 0.7111,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.32,
|
| 62 |
+
"grad_norm": 3.179422378540039,
|
| 63 |
"learning_rate": 0.00018400000000000003,
|
| 64 |
+
"loss": 0.7327,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.36,
|
| 69 |
+
"grad_norm": 2.8877387046813965,
|
| 70 |
"learning_rate": 0.000182,
|
| 71 |
+
"loss": 0.8438,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.4,
|
| 76 |
+
"grad_norm": 2.021406888961792,
|
| 77 |
"learning_rate": 0.00018,
|
| 78 |
+
"loss": 0.8785,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.4,
|
| 83 |
+
"eval_accuracy": 0.711,
|
| 84 |
+
"eval_loss": 0.779495358467102,
|
| 85 |
+
"eval_runtime": 14.6516,
|
| 86 |
+
"eval_samples_per_second": 68.252,
|
| 87 |
+
"eval_steps_per_second": 8.531,
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
{
|
| 91 |
"epoch": 0.44,
|
| 92 |
+
"grad_norm": 1.4613615274429321,
|
| 93 |
"learning_rate": 0.00017800000000000002,
|
| 94 |
+
"loss": 0.767,
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 0.48,
|
| 99 |
+
"grad_norm": 1.6331909894943237,
|
| 100 |
"learning_rate": 0.00017600000000000002,
|
| 101 |
+
"loss": 0.9023,
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 0.52,
|
| 106 |
+
"grad_norm": 1.2065010070800781,
|
| 107 |
"learning_rate": 0.000174,
|
| 108 |
+
"loss": 0.7347,
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 0.56,
|
| 113 |
+
"grad_norm": 1.609735369682312,
|
| 114 |
"learning_rate": 0.000172,
|
| 115 |
+
"loss": 0.5344,
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.6,
|
| 120 |
+
"grad_norm": 3.425642251968384,
|
| 121 |
"learning_rate": 0.00017,
|
| 122 |
+
"loss": 0.5897,
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 0.64,
|
| 127 |
+
"grad_norm": 0.7095292210578918,
|
| 128 |
"learning_rate": 0.000168,
|
| 129 |
+
"loss": 0.6487,
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 0.68,
|
| 134 |
+
"grad_norm": 1.6857857704162598,
|
| 135 |
"learning_rate": 0.000166,
|
| 136 |
+
"loss": 0.479,
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 0.72,
|
| 141 |
+
"grad_norm": 3.3737733364105225,
|
| 142 |
"learning_rate": 0.000164,
|
| 143 |
+
"loss": 0.6547,
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.76,
|
| 148 |
+
"grad_norm": 2.8827691078186035,
|
| 149 |
"learning_rate": 0.000162,
|
| 150 |
+
"loss": 0.667,
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 0.8,
|
| 155 |
+
"grad_norm": 0.998505711555481,
|
| 156 |
"learning_rate": 0.00016,
|
| 157 |
+
"loss": 0.7076,
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.8,
|
| 162 |
+
"eval_accuracy": 0.818,
|
| 163 |
+
"eval_loss": 0.5420999526977539,
|
| 164 |
+
"eval_runtime": 14.8716,
|
| 165 |
+
"eval_samples_per_second": 67.242,
|
| 166 |
+
"eval_steps_per_second": 8.405,
|
| 167 |
"step": 200
|
| 168 |
},
|
| 169 |
{
|
| 170 |
"epoch": 0.84,
|
| 171 |
+
"grad_norm": 1.6601933240890503,
|
| 172 |
"learning_rate": 0.00015800000000000002,
|
| 173 |
+
"loss": 0.5418,
|
| 174 |
"step": 210
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.88,
|
| 178 |
+
"grad_norm": 2.71398663520813,
|
| 179 |
"learning_rate": 0.00015600000000000002,
|
| 180 |
+
"loss": 0.6696,
|
| 181 |
"step": 220
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 0.92,
|
| 185 |
+
"grad_norm": 3.1173503398895264,
|
| 186 |
"learning_rate": 0.000154,
|
| 187 |
+
"loss": 0.5191,
|
| 188 |
"step": 230
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.96,
|
| 192 |
+
"grad_norm": 1.7604912519454956,
|
| 193 |
"learning_rate": 0.000152,
|
| 194 |
+
"loss": 0.5596,
|
| 195 |
"step": 240
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 1.0,
|
| 199 |
+
"grad_norm": 1.8370306491851807,
|
| 200 |
"learning_rate": 0.00015000000000000001,
|
| 201 |
+
"loss": 0.4941,
|
| 202 |
"step": 250
|
| 203 |
},
|
| 204 |
{
|
| 205 |
"epoch": 1.04,
|
| 206 |
+
"grad_norm": 3.0072226524353027,
|
| 207 |
"learning_rate": 0.000148,
|
| 208 |
+
"loss": 0.417,
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 1.08,
|
| 213 |
+
"grad_norm": 2.4395639896392822,
|
| 214 |
"learning_rate": 0.000146,
|
| 215 |
+
"loss": 0.5287,
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
"epoch": 1.12,
|
| 220 |
+
"grad_norm": 1.1612118482589722,
|
| 221 |
"learning_rate": 0.000144,
|
| 222 |
+
"loss": 0.4771,
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
"epoch": 1.16,
|
| 227 |
+
"grad_norm": 1.5060698986053467,
|
| 228 |
"learning_rate": 0.000142,
|
| 229 |
+
"loss": 0.4211,
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 1.2,
|
| 234 |
+
"grad_norm": 2.4940316677093506,
|
| 235 |
"learning_rate": 0.00014,
|
| 236 |
+
"loss": 0.4283,
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 1.2,
|
| 241 |
+
"eval_accuracy": 0.876,
|
| 242 |
+
"eval_loss": 0.3951060175895691,
|
| 243 |
+
"eval_runtime": 14.0948,
|
| 244 |
+
"eval_samples_per_second": 70.948,
|
| 245 |
+
"eval_steps_per_second": 8.869,
|
| 246 |
"step": 300
|
| 247 |
},
|
| 248 |
{
|
| 249 |
"epoch": 1.24,
|
| 250 |
+
"grad_norm": 3.643969774246216,
|
| 251 |
"learning_rate": 0.000138,
|
| 252 |
+
"loss": 0.4377,
|
| 253 |
"step": 310
|
| 254 |
},
|
| 255 |
{
|
| 256 |
"epoch": 1.28,
|
| 257 |
+
"grad_norm": 1.954455852508545,
|
| 258 |
"learning_rate": 0.00013600000000000003,
|
| 259 |
+
"loss": 0.4311,
|
| 260 |
"step": 320
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 1.32,
|
| 264 |
+
"grad_norm": 1.4906481504440308,
|
| 265 |
"learning_rate": 0.000134,
|
| 266 |
+
"loss": 0.3393,
|
| 267 |
"step": 330
|
| 268 |
},
|
| 269 |
{
|
| 270 |
"epoch": 1.3599999999999999,
|
| 271 |
+
"grad_norm": 2.0885210037231445,
|
| 272 |
"learning_rate": 0.000132,
|
| 273 |
+
"loss": 0.4909,
|
| 274 |
"step": 340
|
| 275 |
},
|
| 276 |
{
|
| 277 |
"epoch": 1.4,
|
| 278 |
+
"grad_norm": 4.9808173179626465,
|
| 279 |
"learning_rate": 0.00013000000000000002,
|
| 280 |
+
"loss": 0.4071,
|
| 281 |
"step": 350
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 1.44,
|
| 285 |
+
"grad_norm": 2.143996477127075,
|
| 286 |
"learning_rate": 0.00012800000000000002,
|
| 287 |
+
"loss": 0.2979,
|
| 288 |
"step": 360
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"epoch": 1.48,
|
| 292 |
+
"grad_norm": 5.164979457855225,
|
| 293 |
"learning_rate": 0.000126,
|
| 294 |
+
"loss": 0.4576,
|
| 295 |
"step": 370
|
| 296 |
},
|
| 297 |
{
|
| 298 |
"epoch": 1.52,
|
| 299 |
+
"grad_norm": 6.777462482452393,
|
| 300 |
"learning_rate": 0.000124,
|
| 301 |
+
"loss": 0.446,
|
| 302 |
"step": 380
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 1.56,
|
| 306 |
+
"grad_norm": 1.1243666410446167,
|
| 307 |
"learning_rate": 0.000122,
|
| 308 |
+
"loss": 0.3795,
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 1.6,
|
| 313 |
+
"grad_norm": 3.663288116455078,
|
| 314 |
"learning_rate": 0.00012,
|
| 315 |
+
"loss": 0.4251,
|
| 316 |
"step": 400
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 1.6,
|
| 320 |
+
"eval_accuracy": 0.864,
|
| 321 |
+
"eval_loss": 0.38176068663597107,
|
| 322 |
+
"eval_runtime": 14.993,
|
| 323 |
+
"eval_samples_per_second": 66.698,
|
| 324 |
+
"eval_steps_per_second": 8.337,
|
| 325 |
"step": 400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.6400000000000001,
|
| 329 |
+
"grad_norm": 2.137402296066284,
|
| 330 |
"learning_rate": 0.000118,
|
| 331 |
+
"loss": 0.4443,
|
| 332 |
"step": 410
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.6800000000000002,
|
| 336 |
+
"grad_norm": 3.767395257949829,
|
| 337 |
"learning_rate": 0.000116,
|
| 338 |
+
"loss": 0.5193,
|
| 339 |
"step": 420
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.72,
|
| 343 |
+
"grad_norm": 1.152035117149353,
|
| 344 |
"learning_rate": 0.00011399999999999999,
|
| 345 |
+
"loss": 0.3036,
|
| 346 |
"step": 430
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.76,
|
| 350 |
+
"grad_norm": 5.046035289764404,
|
| 351 |
"learning_rate": 0.00011200000000000001,
|
| 352 |
+
"loss": 0.2667,
|
| 353 |
"step": 440
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.8,
|
| 357 |
+
"grad_norm": 1.6602838039398193,
|
| 358 |
"learning_rate": 0.00011000000000000002,
|
| 359 |
+
"loss": 0.3035,
|
| 360 |
"step": 450
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.8399999999999999,
|
| 364 |
+
"grad_norm": 2.508718729019165,
|
| 365 |
"learning_rate": 0.00010800000000000001,
|
| 366 |
+
"loss": 0.4932,
|
| 367 |
"step": 460
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.88,
|
| 371 |
+
"grad_norm": 2.9392433166503906,
|
| 372 |
"learning_rate": 0.00010600000000000002,
|
| 373 |
+
"loss": 0.4191,
|
| 374 |
"step": 470
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.92,
|
| 378 |
+
"grad_norm": 2.791663885116577,
|
| 379 |
"learning_rate": 0.00010400000000000001,
|
| 380 |
+
"loss": 0.3038,
|
| 381 |
"step": 480
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.96,
|
| 385 |
+
"grad_norm": 2.3044326305389404,
|
| 386 |
"learning_rate": 0.00010200000000000001,
|
| 387 |
+
"loss": 0.4298,
|
| 388 |
"step": 490
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 2.0,
|
| 392 |
+
"grad_norm": 3.1489250659942627,
|
| 393 |
"learning_rate": 0.0001,
|
| 394 |
+
"loss": 0.335,
|
| 395 |
"step": 500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 2.0,
|
| 399 |
+
"eval_accuracy": 0.924,
|
| 400 |
+
"eval_loss": 0.24736037850379944,
|
| 401 |
+
"eval_runtime": 15.0667,
|
| 402 |
+
"eval_samples_per_second": 66.372,
|
| 403 |
+
"eval_steps_per_second": 8.296,
|
| 404 |
"step": 500
|
| 405 |
},
|
| 406 |
{
|
| 407 |
"epoch": 2.04,
|
| 408 |
+
"grad_norm": 3.3036346435546875,
|
| 409 |
"learning_rate": 9.8e-05,
|
| 410 |
+
"loss": 0.1649,
|
| 411 |
"step": 510
|
| 412 |
},
|
| 413 |
{
|
| 414 |
"epoch": 2.08,
|
| 415 |
+
"grad_norm": 2.8792152404785156,
|
| 416 |
"learning_rate": 9.6e-05,
|
| 417 |
+
"loss": 0.1627,
|
| 418 |
"step": 520
|
| 419 |
},
|
| 420 |
{
|
| 421 |
"epoch": 2.12,
|
| 422 |
+
"grad_norm": 0.28887757658958435,
|
| 423 |
"learning_rate": 9.4e-05,
|
| 424 |
+
"loss": 0.3599,
|
| 425 |
"step": 530
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"epoch": 2.16,
|
| 429 |
+
"grad_norm": 5.17996883392334,
|
| 430 |
"learning_rate": 9.200000000000001e-05,
|
| 431 |
+
"loss": 0.3233,
|
| 432 |
"step": 540
|
| 433 |
},
|
| 434 |
{
|
| 435 |
"epoch": 2.2,
|
| 436 |
+
"grad_norm": 2.1433322429656982,
|
| 437 |
"learning_rate": 9e-05,
|
| 438 |
+
"loss": 0.222,
|
| 439 |
"step": 550
|
| 440 |
},
|
| 441 |
{
|
| 442 |
"epoch": 2.24,
|
| 443 |
+
"grad_norm": 3.143852949142456,
|
| 444 |
"learning_rate": 8.800000000000001e-05,
|
| 445 |
+
"loss": 0.2713,
|
| 446 |
"step": 560
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 2.2800000000000002,
|
| 450 |
+
"grad_norm": 0.4215773344039917,
|
| 451 |
"learning_rate": 8.6e-05,
|
| 452 |
+
"loss": 0.1708,
|
| 453 |
"step": 570
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 2.32,
|
| 457 |
+
"grad_norm": 1.9217822551727295,
|
| 458 |
"learning_rate": 8.4e-05,
|
| 459 |
+
"loss": 0.198,
|
| 460 |
"step": 580
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 2.36,
|
| 464 |
+
"grad_norm": 2.4554295539855957,
|
| 465 |
"learning_rate": 8.2e-05,
|
| 466 |
+
"loss": 0.1617,
|
| 467 |
"step": 590
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 2.4,
|
| 471 |
+
"grad_norm": 1.2291343212127686,
|
| 472 |
"learning_rate": 8e-05,
|
| 473 |
+
"loss": 0.2286,
|
| 474 |
"step": 600
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 2.4,
|
| 478 |
+
"eval_accuracy": 0.952,
|
| 479 |
+
"eval_loss": 0.16752035915851593,
|
| 480 |
+
"eval_runtime": 13.8882,
|
| 481 |
+
"eval_samples_per_second": 72.003,
|
| 482 |
+
"eval_steps_per_second": 9.0,
|
| 483 |
"step": 600
|
| 484 |
},
|
| 485 |
{
|
| 486 |
"epoch": 2.44,
|
| 487 |
+
"grad_norm": 0.3957996964454651,
|
| 488 |
"learning_rate": 7.800000000000001e-05,
|
| 489 |
+
"loss": 0.1758,
|
| 490 |
"step": 610
|
| 491 |
},
|
| 492 |
{
|
| 493 |
"epoch": 2.48,
|
| 494 |
+
"grad_norm": 2.093458414077759,
|
| 495 |
"learning_rate": 7.6e-05,
|
| 496 |
+
"loss": 0.138,
|
| 497 |
"step": 620
|
| 498 |
},
|
| 499 |
{
|
| 500 |
"epoch": 2.52,
|
| 501 |
+
"grad_norm": 1.130835771560669,
|
| 502 |
"learning_rate": 7.4e-05,
|
| 503 |
+
"loss": 0.1194,
|
| 504 |
"step": 630
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 2.56,
|
| 508 |
+
"grad_norm": 1.3829611539840698,
|
| 509 |
"learning_rate": 7.2e-05,
|
| 510 |
+
"loss": 0.1629,
|
| 511 |
"step": 640
|
| 512 |
},
|
| 513 |
{
|
| 514 |
"epoch": 2.6,
|
| 515 |
+
"grad_norm": 0.17628225684165955,
|
| 516 |
"learning_rate": 7e-05,
|
| 517 |
+
"loss": 0.0954,
|
| 518 |
"step": 650
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 2.64,
|
| 522 |
+
"grad_norm": 1.7156352996826172,
|
| 523 |
"learning_rate": 6.800000000000001e-05,
|
| 524 |
+
"loss": 0.172,
|
| 525 |
"step": 660
|
| 526 |
},
|
| 527 |
{
|
| 528 |
"epoch": 2.68,
|
| 529 |
+
"grad_norm": 3.950498342514038,
|
| 530 |
"learning_rate": 6.6e-05,
|
| 531 |
+
"loss": 0.1414,
|
| 532 |
"step": 670
|
| 533 |
},
|
| 534 |
{
|
| 535 |
"epoch": 2.7199999999999998,
|
| 536 |
+
"grad_norm": 3.134085178375244,
|
| 537 |
"learning_rate": 6.400000000000001e-05,
|
| 538 |
+
"loss": 0.1734,
|
| 539 |
"step": 680
|
| 540 |
},
|
| 541 |
{
|
| 542 |
"epoch": 2.76,
|
| 543 |
+
"grad_norm": 4.362244129180908,
|
| 544 |
"learning_rate": 6.2e-05,
|
| 545 |
+
"loss": 0.1806,
|
| 546 |
"step": 690
|
| 547 |
},
|
| 548 |
{
|
| 549 |
"epoch": 2.8,
|
| 550 |
+
"grad_norm": 5.472875118255615,
|
| 551 |
"learning_rate": 6e-05,
|
| 552 |
+
"loss": 0.1523,
|
| 553 |
"step": 700
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"epoch": 2.8,
|
| 557 |
+
"eval_accuracy": 0.954,
|
| 558 |
+
"eval_loss": 0.1640758216381073,
|
| 559 |
+
"eval_runtime": 14.3922,
|
| 560 |
+
"eval_samples_per_second": 69.482,
|
| 561 |
+
"eval_steps_per_second": 8.685,
|
| 562 |
"step": 700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 2.84,
|
| 566 |
+
"grad_norm": 1.1265697479248047,
|
| 567 |
"learning_rate": 5.8e-05,
|
| 568 |
+
"loss": 0.1939,
|
| 569 |
"step": 710
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 2.88,
|
| 573 |
+
"grad_norm": 3.1513569355010986,
|
| 574 |
"learning_rate": 5.6000000000000006e-05,
|
| 575 |
+
"loss": 0.1369,
|
| 576 |
"step": 720
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 2.92,
|
| 580 |
+
"grad_norm": 0.14964208006858826,
|
| 581 |
"learning_rate": 5.4000000000000005e-05,
|
| 582 |
+
"loss": 0.1372,
|
| 583 |
"step": 730
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 2.96,
|
| 587 |
+
"grad_norm": 0.1409606784582138,
|
| 588 |
"learning_rate": 5.2000000000000004e-05,
|
| 589 |
+
"loss": 0.1503,
|
| 590 |
"step": 740
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 3.0,
|
| 594 |
+
"grad_norm": 0.1397903859615326,
|
| 595 |
"learning_rate": 5e-05,
|
| 596 |
+
"loss": 0.1655,
|
| 597 |
"step": 750
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 3.04,
|
| 601 |
+
"grad_norm": 0.6654576063156128,
|
| 602 |
"learning_rate": 4.8e-05,
|
| 603 |
+
"loss": 0.0662,
|
| 604 |
"step": 760
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 3.08,
|
| 608 |
+
"grad_norm": 0.11212094128131866,
|
| 609 |
"learning_rate": 4.600000000000001e-05,
|
| 610 |
+
"loss": 0.0362,
|
| 611 |
"step": 770
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 3.12,
|
| 615 |
+
"grad_norm": 0.8599291443824768,
|
| 616 |
"learning_rate": 4.4000000000000006e-05,
|
| 617 |
+
"loss": 0.0545,
|
| 618 |
"step": 780
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 3.16,
|
| 622 |
+
"grad_norm": 0.0967254787683487,
|
| 623 |
"learning_rate": 4.2e-05,
|
| 624 |
+
"loss": 0.0428,
|
| 625 |
"step": 790
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 3.2,
|
| 629 |
+
"grad_norm": 1.8969135284423828,
|
| 630 |
"learning_rate": 4e-05,
|
| 631 |
+
"loss": 0.1346,
|
| 632 |
"step": 800
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 3.2,
|
| 636 |
+
"eval_accuracy": 0.969,
|
| 637 |
+
"eval_loss": 0.11203579604625702,
|
| 638 |
+
"eval_runtime": 14.1593,
|
| 639 |
+
"eval_samples_per_second": 70.625,
|
| 640 |
+
"eval_steps_per_second": 8.828,
|
| 641 |
"step": 800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 3.24,
|
| 645 |
+
"grad_norm": 0.32526883482933044,
|
| 646 |
"learning_rate": 3.8e-05,
|
| 647 |
+
"loss": 0.0576,
|
| 648 |
"step": 810
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 3.2800000000000002,
|
| 652 |
+
"grad_norm": 3.220393419265747,
|
| 653 |
"learning_rate": 3.6e-05,
|
| 654 |
+
"loss": 0.0532,
|
| 655 |
"step": 820
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 3.32,
|
| 659 |
+
"grad_norm": 0.09663262218236923,
|
| 660 |
"learning_rate": 3.4000000000000007e-05,
|
| 661 |
+
"loss": 0.0832,
|
| 662 |
"step": 830
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 3.36,
|
| 666 |
+
"grad_norm": 0.07815458625555038,
|
| 667 |
"learning_rate": 3.2000000000000005e-05,
|
| 668 |
+
"loss": 0.0653,
|
| 669 |
"step": 840
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 3.4,
|
| 673 |
+
"grad_norm": 0.09688442200422287,
|
| 674 |
"learning_rate": 3e-05,
|
| 675 |
+
"loss": 0.0763,
|
| 676 |
"step": 850
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 3.44,
|
| 680 |
+
"grad_norm": 0.07274580001831055,
|
| 681 |
"learning_rate": 2.8000000000000003e-05,
|
| 682 |
+
"loss": 0.04,
|
| 683 |
"step": 860
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 3.48,
|
| 687 |
+
"grad_norm": 0.06821909546852112,
|
| 688 |
"learning_rate": 2.6000000000000002e-05,
|
| 689 |
+
"loss": 0.0778,
|
| 690 |
"step": 870
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 3.52,
|
| 694 |
+
"grad_norm": 1.8587623834609985,
|
| 695 |
"learning_rate": 2.4e-05,
|
| 696 |
+
"loss": 0.0563,
|
| 697 |
"step": 880
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 3.56,
|
| 701 |
+
"grad_norm": 0.09389644116163254,
|
| 702 |
"learning_rate": 2.2000000000000003e-05,
|
| 703 |
+
"loss": 0.1213,
|
| 704 |
"step": 890
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 3.6,
|
| 708 |
+
"grad_norm": 6.0940117835998535,
|
| 709 |
"learning_rate": 2e-05,
|
| 710 |
+
"loss": 0.0638,
|
| 711 |
"step": 900
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 3.6,
|
| 715 |
+
"eval_accuracy": 0.978,
|
| 716 |
+
"eval_loss": 0.10251828283071518,
|
| 717 |
+
"eval_runtime": 14.4118,
|
| 718 |
+
"eval_samples_per_second": 69.388,
|
| 719 |
+
"eval_steps_per_second": 8.673,
|
| 720 |
"step": 900
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 3.64,
|
| 724 |
+
"grad_norm": 0.1492830514907837,
|
| 725 |
"learning_rate": 1.8e-05,
|
| 726 |
+
"loss": 0.0821,
|
| 727 |
"step": 910
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 3.68,
|
| 731 |
+
"grad_norm": 0.07156217098236084,
|
| 732 |
"learning_rate": 1.6000000000000003e-05,
|
| 733 |
+
"loss": 0.023,
|
| 734 |
"step": 920
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 3.7199999999999998,
|
| 738 |
+
"grad_norm": 0.07530596107244492,
|
| 739 |
"learning_rate": 1.4000000000000001e-05,
|
| 740 |
+
"loss": 0.0821,
|
| 741 |
"step": 930
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 3.76,
|
| 745 |
+
"grad_norm": 0.061225246638059616,
|
| 746 |
"learning_rate": 1.2e-05,
|
| 747 |
+
"loss": 0.0518,
|
| 748 |
"step": 940
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 3.8,
|
| 752 |
+
"grad_norm": 0.15823562443256378,
|
| 753 |
"learning_rate": 1e-05,
|
| 754 |
+
"loss": 0.0286,
|
| 755 |
"step": 950
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 3.84,
|
| 759 |
+
"grad_norm": 0.09221441298723221,
|
| 760 |
"learning_rate": 8.000000000000001e-06,
|
| 761 |
+
"loss": 0.0364,
|
| 762 |
"step": 960
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 3.88,
|
| 766 |
+
"grad_norm": 0.05617209151387215,
|
| 767 |
"learning_rate": 6e-06,
|
| 768 |
+
"loss": 0.0498,
|
| 769 |
"step": 970
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 3.92,
|
| 773 |
+
"grad_norm": 0.08119833469390869,
|
| 774 |
"learning_rate": 4.000000000000001e-06,
|
| 775 |
+
"loss": 0.049,
|
| 776 |
"step": 980
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 3.96,
|
| 780 |
+
"grad_norm": 0.08950098603963852,
|
| 781 |
"learning_rate": 2.0000000000000003e-06,
|
| 782 |
+
"loss": 0.0642,
|
| 783 |
"step": 990
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 4.0,
|
| 787 |
+
"grad_norm": 0.07529381662607193,
|
| 788 |
"learning_rate": 0.0,
|
| 789 |
+
"loss": 0.0574,
|
| 790 |
"step": 1000
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 4.0,
|
| 794 |
+
"eval_accuracy": 0.981,
|
| 795 |
+
"eval_loss": 0.09177211672067642,
|
| 796 |
+
"eval_runtime": 15.0707,
|
| 797 |
+
"eval_samples_per_second": 66.354,
|
| 798 |
+
"eval_steps_per_second": 8.294,
|
| 799 |
"step": 1000
|
| 800 |
},
|
| 801 |
{
|
| 802 |
"epoch": 4.0,
|
| 803 |
"step": 1000,
|
| 804 |
"total_flos": 1.239905171570688e+18,
|
| 805 |
+
"train_loss": 0.3459155881404877,
|
| 806 |
+
"train_runtime": 562.8746,
|
| 807 |
+
"train_samples_per_second": 28.426,
|
| 808 |
+
"train_steps_per_second": 1.777
|
| 809 |
}
|
| 810 |
],
|
| 811 |
"logging_steps": 10,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5112
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c40dc00985c0a75ad675d49b33b6ad4495940384eabfc0ead05db15735440fc4
|
| 3 |
size 5112
|