End of training
Browse files- README.md +3 -3
- all_results.json +4 -4
- confusion_matrix_test_fold_2.png +0 -0
- confusion_matrix_valid_fold_2.png +0 -0
- test_results.json +4 -4
- trainer_state.json +224 -126
README.md
CHANGED
|
@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [facebook/timesformer-base-finetuned-k400](https://huggingface.co/facebook/timesformer-base-finetuned-k400) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
- Accuracy: 0.7782
|
| 23 |
|
| 24 |
## Model description
|
|
@@ -45,13 +45,13 @@ The following hyperparameters were used during training:
|
|
| 45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
- lr_scheduler_type: linear
|
| 47 |
- lr_scheduler_warmup_ratio: 0.1
|
| 48 |
-
- training_steps:
|
| 49 |
|
| 50 |
### Training results
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 54 |
-
| 0.
|
| 55 |
|
| 56 |
|
| 57 |
### Framework versions
|
|
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [facebook/timesformer-base-finetuned-k400](https://huggingface.co/facebook/timesformer-base-finetuned-k400) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.8434
|
| 22 |
- Accuracy: 0.7782
|
| 23 |
|
| 24 |
## Model description
|
|
|
|
| 45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
- lr_scheduler_type: linear
|
| 47 |
- lr_scheduler_warmup_ratio: 0.1
|
| 48 |
+
- training_steps: 4015
|
| 49 |
|
| 50 |
### Training results
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 54 |
+
| 0.0403 | 1.0 | 4015 | 0.0018 | 0.9997 |
|
| 55 |
|
| 56 |
|
| 57 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
"eval_accuracy": 0.7782131661442007,
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 2.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
"eval_accuracy": 0.7782131661442007,
|
| 4 |
+
"eval_loss": 0.843409538269043,
|
| 5 |
+
"eval_runtime": 509.169,
|
| 6 |
+
"eval_samples_per_second": 2.506,
|
| 7 |
+
"eval_steps_per_second": 1.253
|
| 8 |
}
|
confusion_matrix_test_fold_2.png
ADDED
|
confusion_matrix_valid_fold_2.png
ADDED
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
"eval_accuracy": 0.7782131661442007,
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 2.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
"eval_accuracy": 0.7782131661442007,
|
| 4 |
+
"eval_loss": 0.843409538269043,
|
| 5 |
+
"eval_runtime": 509.169,
|
| 6 |
+
"eval_samples_per_second": 2.506,
|
| 7 |
+
"eval_steps_per_second": 1.253
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,225 +1,323 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/Timesformer_default_fold_10_10epoch_noAug_batch4/checkpoint-
|
| 4 |
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 1.
|
| 15 |
-
"loss":
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 4.
|
| 36 |
-
"loss":
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 4.
|
| 43 |
-
"loss":
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 4.
|
| 50 |
-
"loss":
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 4.
|
| 57 |
-
"loss":
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 3.
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
-
"learning_rate":
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate":
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate": 1.
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 2600
|
| 192 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 193 |
{
|
| 194 |
"epoch": 1.0,
|
| 195 |
-
"eval_accuracy": 0.
|
| 196 |
-
"eval_loss": 0.
|
| 197 |
-
"eval_runtime":
|
| 198 |
-
"eval_samples_per_second": 2.
|
| 199 |
-
"eval_steps_per_second": 1.
|
| 200 |
-
"step":
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.0,
|
| 204 |
-
"step":
|
| 205 |
-
"total_flos":
|
| 206 |
-
"train_loss": 0.
|
| 207 |
-
"train_runtime":
|
| 208 |
-
"train_samples_per_second": 1.
|
| 209 |
-
"train_steps_per_second": 0.
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 1.0,
|
| 213 |
"eval_accuracy": 0.7782131661442007,
|
| 214 |
-
"eval_loss": 0.
|
| 215 |
-
"eval_runtime":
|
| 216 |
-
"eval_samples_per_second": 2.
|
| 217 |
-
"eval_steps_per_second": 1.
|
| 218 |
-
"step":
|
| 219 |
}
|
| 220 |
],
|
| 221 |
"logging_steps": 100,
|
| 222 |
-
"max_steps":
|
| 223 |
"num_input_tokens_seen": 0,
|
| 224 |
"num_train_epochs": 9223372036854775807,
|
| 225 |
"save_steps": 500,
|
|
@@ -235,7 +333,7 @@
|
|
| 235 |
"attributes": {}
|
| 236 |
}
|
| 237 |
},
|
| 238 |
-
"total_flos":
|
| 239 |
"train_batch_size": 2,
|
| 240 |
"trial_name": null,
|
| 241 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9997126436781609,
|
| 3 |
+
"best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/Timesformer_default_fold_10_10epoch_noAug_batch4/checkpoint-4015",
|
| 4 |
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 4015,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.024906600249066,
|
| 13 |
+
"grad_norm": 0.0679977759718895,
|
| 14 |
+
"learning_rate": 1.2437810945273633e-05,
|
| 15 |
+
"loss": 0.0161,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.049813200498132,
|
| 20 |
+
"grad_norm": 1.2939980030059814,
|
| 21 |
+
"learning_rate": 2.4875621890547266e-05,
|
| 22 |
+
"loss": 0.0335,
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.074719800747198,
|
| 27 |
+
"grad_norm": 23.8648738861084,
|
| 28 |
+
"learning_rate": 3.73134328358209e-05,
|
| 29 |
+
"loss": 0.0463,
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.099626400996264,
|
| 34 |
+
"grad_norm": 0.10109841078519821,
|
| 35 |
+
"learning_rate": 4.975124378109453e-05,
|
| 36 |
+
"loss": 0.0968,
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.12453300124533001,
|
| 41 |
+
"grad_norm": 0.40682753920555115,
|
| 42 |
+
"learning_rate": 4.864378632715195e-05,
|
| 43 |
+
"loss": 0.1198,
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.149439601494396,
|
| 48 |
+
"grad_norm": 3.5996460914611816,
|
| 49 |
+
"learning_rate": 4.7259894824245784e-05,
|
| 50 |
+
"loss": 0.2129,
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.17434620174346202,
|
| 55 |
+
"grad_norm": 0.060298651456832886,
|
| 56 |
+
"learning_rate": 4.587600332133961e-05,
|
| 57 |
+
"loss": 0.1732,
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.199252801992528,
|
| 62 |
+
"grad_norm": 0.05418834835290909,
|
| 63 |
+
"learning_rate": 4.449211181843343e-05,
|
| 64 |
+
"loss": 0.1067,
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.22415940224159403,
|
| 69 |
+
"grad_norm": 0.07002612948417664,
|
| 70 |
+
"learning_rate": 4.3108220315527264e-05,
|
| 71 |
+
"loss": 0.0959,
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.24906600249066002,
|
| 76 |
+
"grad_norm": 0.09562501311302185,
|
| 77 |
+
"learning_rate": 4.1724328812621095e-05,
|
| 78 |
+
"loss": 0.114,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.273972602739726,
|
| 83 |
+
"grad_norm": 2.8778584003448486,
|
| 84 |
+
"learning_rate": 4.034043730971492e-05,
|
| 85 |
+
"loss": 0.1913,
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.298879202988792,
|
| 90 |
+
"grad_norm": 0.15859746932983398,
|
| 91 |
+
"learning_rate": 3.895654580680875e-05,
|
| 92 |
+
"loss": 0.0471,
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.32378580323785805,
|
| 97 |
+
"grad_norm": 0.038741584867239,
|
| 98 |
+
"learning_rate": 3.7572654303902575e-05,
|
| 99 |
+
"loss": 0.1192,
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.34869240348692404,
|
| 104 |
+
"grad_norm": 3.3708555698394775,
|
| 105 |
+
"learning_rate": 3.6188762800996406e-05,
|
| 106 |
+
"loss": 0.0563,
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.37359900373599003,
|
| 111 |
+
"grad_norm": 0.10332904011011124,
|
| 112 |
+
"learning_rate": 3.480487129809023e-05,
|
| 113 |
+
"loss": 0.0653,
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.398505603985056,
|
| 118 |
+
"grad_norm": 0.05150986090302467,
|
| 119 |
+
"learning_rate": 3.342097979518406e-05,
|
| 120 |
+
"loss": 0.0765,
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.42341220423412207,
|
| 125 |
+
"grad_norm": 4.3742194175720215,
|
| 126 |
+
"learning_rate": 3.2037088292277886e-05,
|
| 127 |
+
"loss": 0.0732,
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.44831880448318806,
|
| 132 |
+
"grad_norm": 0.17955519258975983,
|
| 133 |
+
"learning_rate": 3.065319678937171e-05,
|
| 134 |
+
"loss": 0.0852,
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.47322540473225405,
|
| 139 |
+
"grad_norm": 0.017491327598690987,
|
| 140 |
+
"learning_rate": 2.926930528646554e-05,
|
| 141 |
+
"loss": 0.0321,
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.49813200498132004,
|
| 146 |
+
"grad_norm": 0.062625452876091,
|
| 147 |
+
"learning_rate": 2.7885413783559373e-05,
|
| 148 |
+
"loss": 0.0639,
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.523038605230386,
|
| 153 |
+
"grad_norm": 0.01796746626496315,
|
| 154 |
+
"learning_rate": 2.6501522280653197e-05,
|
| 155 |
+
"loss": 0.0204,
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.547945205479452,
|
| 160 |
+
"grad_norm": 0.006298160646110773,
|
| 161 |
+
"learning_rate": 2.5117630777747025e-05,
|
| 162 |
+
"loss": 0.0376,
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.572851805728518,
|
| 167 |
+
"grad_norm": 0.11808665096759796,
|
| 168 |
+
"learning_rate": 2.3733739274840856e-05,
|
| 169 |
+
"loss": 0.0786,
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.597758405977584,
|
| 174 |
+
"grad_norm": 0.026163829490542412,
|
| 175 |
+
"learning_rate": 2.234984777193468e-05,
|
| 176 |
+
"loss": 0.0087,
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.6226650062266501,
|
| 181 |
+
"grad_norm": 0.051311738789081573,
|
| 182 |
+
"learning_rate": 2.0965956269028508e-05,
|
| 183 |
+
"loss": 0.0345,
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.6475716064757161,
|
| 188 |
+
"grad_norm": 0.004274987615644932,
|
| 189 |
+
"learning_rate": 1.958206476612234e-05,
|
| 190 |
+
"loss": 0.0484,
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6724782067247821,
|
| 195 |
+
"grad_norm": 0.008581398986279964,
|
| 196 |
+
"learning_rate": 1.8198173263216163e-05,
|
| 197 |
+
"loss": 0.0112,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6973848069738481,
|
| 202 |
+
"grad_norm": 0.08034462481737137,
|
| 203 |
+
"learning_rate": 1.681428176030999e-05,
|
| 204 |
+
"loss": 0.0118,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7222914072229141,
|
| 209 |
+
"grad_norm": 0.0067258900962769985,
|
| 210 |
+
"learning_rate": 1.543039025740382e-05,
|
| 211 |
+
"loss": 0.0251,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7471980074719801,
|
| 216 |
+
"grad_norm": 0.004430562723428011,
|
| 217 |
+
"learning_rate": 1.4046498754497647e-05,
|
| 218 |
+
"loss": 0.0085,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.772104607721046,
|
| 223 |
+
"grad_norm": 0.0027492486406117678,
|
| 224 |
+
"learning_rate": 1.2662607251591476e-05,
|
| 225 |
+
"loss": 0.0089,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.797011207970112,
|
| 230 |
+
"grad_norm": 0.003413543803617358,
|
| 231 |
+
"learning_rate": 1.1278715748685304e-05,
|
| 232 |
+
"loss": 0.0069,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.821917808219178,
|
| 237 |
+
"grad_norm": 0.00517587224021554,
|
| 238 |
+
"learning_rate": 9.894824245779132e-06,
|
| 239 |
+
"loss": 0.001,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8468244084682441,
|
| 244 |
+
"grad_norm": 0.0020790928974747658,
|
| 245 |
+
"learning_rate": 8.51093274287296e-06,
|
| 246 |
+
"loss": 0.002,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8717310087173101,
|
| 251 |
+
"grad_norm": 0.010764083825051785,
|
| 252 |
+
"learning_rate": 7.127041239966786e-06,
|
| 253 |
+
"loss": 0.0009,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.8966376089663761,
|
| 258 |
+
"grad_norm": 0.005092688836157322,
|
| 259 |
+
"learning_rate": 5.743149737060614e-06,
|
| 260 |
+
"loss": 0.0003,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9215442092154421,
|
| 265 |
+
"grad_norm": 0.0029076840728521347,
|
| 266 |
+
"learning_rate": 4.359258234154443e-06,
|
| 267 |
+
"loss": 0.0012,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9464508094645081,
|
| 272 |
+
"grad_norm": 0.0030453051440417767,
|
| 273 |
+
"learning_rate": 2.97536673124827e-06,
|
| 274 |
+
"loss": 0.0032,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9713574097135741,
|
| 279 |
+
"grad_norm": 0.002343358937650919,
|
| 280 |
+
"learning_rate": 1.5914752283420978e-06,
|
| 281 |
+
"loss": 0.0028,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.9962640099626401,
|
| 286 |
+
"grad_norm": 0.007323646917939186,
|
| 287 |
+
"learning_rate": 2.0758372543592582e-07,
|
| 288 |
+
"loss": 0.0403,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
+
"eval_accuracy": 0.9997126436781609,
|
| 294 |
+
"eval_loss": 0.0017972222995012999,
|
| 295 |
+
"eval_runtime": 1501.9344,
|
| 296 |
+
"eval_samples_per_second": 2.317,
|
| 297 |
+
"eval_steps_per_second": 1.159,
|
| 298 |
+
"step": 4015
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 1.0,
|
| 302 |
+
"step": 4015,
|
| 303 |
+
"total_flos": 7.038050675870269e+18,
|
| 304 |
+
"train_loss": 0.054263464616586085,
|
| 305 |
+
"train_runtime": 5551.0777,
|
| 306 |
+
"train_samples_per_second": 1.447,
|
| 307 |
+
"train_steps_per_second": 0.723
|
| 308 |
},
|
| 309 |
{
|
| 310 |
"epoch": 1.0,
|
| 311 |
"eval_accuracy": 0.7782131661442007,
|
| 312 |
+
"eval_loss": 0.843409538269043,
|
| 313 |
+
"eval_runtime": 509.169,
|
| 314 |
+
"eval_samples_per_second": 2.506,
|
| 315 |
+
"eval_steps_per_second": 1.253,
|
| 316 |
+
"step": 4015
|
| 317 |
}
|
| 318 |
],
|
| 319 |
"logging_steps": 100,
|
| 320 |
+
"max_steps": 4015,
|
| 321 |
"num_input_tokens_seen": 0,
|
| 322 |
"num_train_epochs": 9223372036854775807,
|
| 323 |
"save_steps": 500,
|
|
|
|
| 333 |
"attributes": {}
|
| 334 |
}
|
| 335 |
},
|
| 336 |
+
"total_flos": 7.038050675870269e+18,
|
| 337 |
"train_batch_size": 2,
|
| 338 |
"trial_name": null,
|
| 339 |
"trial_params": null
|