Model save
Browse files- README.md +10 -6
- all_results.json +6 -6
- test_results.json +6 -6
- trainer_state.json +666 -112
- val_results.json +6 -6
README.md
CHANGED
|
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
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.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -44,13 +44,17 @@ The following hyperparameters were used during training:
|
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
- lr_scheduler_warmup_ratio: 0.1
|
| 47 |
-
- training_steps:
|
| 48 |
|
| 49 |
### Training results
|
| 50 |
|
| 51 |
-
| Training Loss | Epoch
|
| 52 |
-
|
| 53 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
|
| 55 |
|
| 56 |
### Framework versions
|
|
|
|
| 17 |
|
| 18 |
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.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.5500
|
| 21 |
+
- Accuracy: 0.8548
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
- lr_scheduler_warmup_ratio: 0.1
|
| 47 |
+
- training_steps: 925
|
| 48 |
|
| 49 |
### Training results
|
| 50 |
|
| 51 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 52 |
+
|:-------------:|:------:|:----:|:---------------:|:--------:|
|
| 53 |
+
| 0.4309 | 0.2011 | 186 | 0.6522 | 0.7220 |
|
| 54 |
+
| 0.6819 | 1.2011 | 372 | 0.7019 | 0.7552 |
|
| 55 |
+
| 0.5011 | 2.2011 | 558 | 0.5440 | 0.8216 |
|
| 56 |
+
| 0.3274 | 3.2011 | 744 | 0.6834 | 0.8257 |
|
| 57 |
+
| 0.0083 | 4.1957 | 925 | 0.5500 | 0.8548 |
|
| 58 |
|
| 59 |
|
| 60 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 0.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.195675675675676,
|
| 3 |
+
"eval_accuracy": 0.8547717842323651,
|
| 4 |
+
"eval_loss": 0.5500420928001404,
|
| 5 |
+
"eval_runtime": 434.3892,
|
| 6 |
+
"eval_samples_per_second": 0.555,
|
| 7 |
+
"eval_steps_per_second": 0.14
|
| 8 |
}
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 0.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.195675675675676,
|
| 3 |
+
"eval_accuracy": 0.8269662921348314,
|
| 4 |
+
"eval_loss": 0.7350317239761353,
|
| 5 |
+
"eval_runtime": 796.0622,
|
| 6 |
+
"eval_samples_per_second": 0.559,
|
| 7 |
+
"eval_steps_per_second": 0.141
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,187 +1,741 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "videomae-timesformer-surf-analytics\\checkpoint-
|
| 4 |
-
"epoch":
|
| 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":
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 3.
|
| 50 |
-
"loss":
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 3.
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm": 11.
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss":
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm": 11.
|
| 126 |
-
"learning_rate": 4.
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
-
"eval_accuracy": 0.
|
| 140 |
-
"eval_loss": 0.
|
| 141 |
-
"eval_runtime":
|
| 142 |
-
"eval_samples_per_second": 0.
|
| 143 |
-
"eval_steps_per_second": 0.
|
| 144 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 145 |
},
|
| 146 |
{
|
| 147 |
-
"epoch":
|
| 148 |
-
"
|
| 149 |
-
"
|
| 150 |
-
"
|
| 151 |
-
"
|
| 152 |
-
"
|
| 153 |
-
"
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch":
|
| 157 |
-
"
|
| 158 |
-
"
|
| 159 |
-
"
|
| 160 |
-
"
|
| 161 |
-
"eval_steps_per_second": 0.113,
|
| 162 |
-
"step": 185
|
| 163 |
},
|
| 164 |
{
|
| 165 |
-
"epoch":
|
| 166 |
-
"
|
| 167 |
-
"
|
| 168 |
-
"
|
| 169 |
-
"
|
| 170 |
-
"eval_steps_per_second": 0.125,
|
| 171 |
-
"step": 185
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"
|
| 176 |
-
"
|
| 177 |
-
"
|
| 178 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 179 |
"eval_steps_per_second": 0.127,
|
| 180 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 181 |
}
|
| 182 |
],
|
| 183 |
"logging_steps": 10,
|
| 184 |
-
"max_steps":
|
| 185 |
"num_input_tokens_seen": 0,
|
| 186 |
"num_train_epochs": 9223372036854775807,
|
| 187 |
"save_steps": 500,
|
|
@@ -197,7 +751,7 @@
|
|
| 197 |
"attributes": {}
|
| 198 |
}
|
| 199 |
},
|
| 200 |
-
"total_flos":
|
| 201 |
"train_batch_size": 4,
|
| 202 |
"trial_name": null,
|
| 203 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8547717842323651,
|
| 3 |
+
"best_model_checkpoint": "videomae-timesformer-surf-analytics\\checkpoint-925",
|
| 4 |
+
"epoch": 4.195675675675676,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 925,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.010810810810810811,
|
| 13 |
+
"grad_norm": 10.205573081970215,
|
| 14 |
+
"learning_rate": 5.376344086021506e-06,
|
| 15 |
+
"loss": 1.5305,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.021621621621621623,
|
| 20 |
+
"grad_norm": 11.725934028625488,
|
| 21 |
+
"learning_rate": 1.0752688172043012e-05,
|
| 22 |
+
"loss": 1.2199,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.032432432432432434,
|
| 27 |
+
"grad_norm": 6.974359035491943,
|
| 28 |
+
"learning_rate": 1.6129032258064517e-05,
|
| 29 |
+
"loss": 1.281,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.043243243243243246,
|
| 34 |
+
"grad_norm": 7.546658515930176,
|
| 35 |
+
"learning_rate": 2.1505376344086024e-05,
|
| 36 |
+
"loss": 1.0878,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.05405405405405406,
|
| 41 |
+
"grad_norm": 9.739876747131348,
|
| 42 |
+
"learning_rate": 2.6881720430107527e-05,
|
| 43 |
+
"loss": 1.133,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.06486486486486487,
|
| 48 |
+
"grad_norm": 14.410557746887207,
|
| 49 |
+
"learning_rate": 3.2258064516129034e-05,
|
| 50 |
+
"loss": 1.049,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.07567567567567568,
|
| 55 |
+
"grad_norm": 8.305644989013672,
|
| 56 |
+
"learning_rate": 3.763440860215054e-05,
|
| 57 |
+
"loss": 1.0185,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.08648648648648649,
|
| 62 |
+
"grad_norm": 5.283512115478516,
|
| 63 |
+
"learning_rate": 4.301075268817205e-05,
|
| 64 |
+
"loss": 0.6385,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.0972972972972973,
|
| 69 |
+
"grad_norm": 9.341029167175293,
|
| 70 |
+
"learning_rate": 4.8387096774193554e-05,
|
| 71 |
+
"loss": 0.7407,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.10810810810810811,
|
| 76 |
+
"grad_norm": 6.474923610687256,
|
| 77 |
+
"learning_rate": 4.957932692307692e-05,
|
| 78 |
+
"loss": 0.5965,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.11891891891891893,
|
| 83 |
+
"grad_norm": 7.116113662719727,
|
| 84 |
+
"learning_rate": 4.897836538461539e-05,
|
| 85 |
+
"loss": 0.9715,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.12972972972972974,
|
| 90 |
+
"grad_norm": 10.062265396118164,
|
| 91 |
+
"learning_rate": 4.8377403846153846e-05,
|
| 92 |
+
"loss": 0.5473,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.14054054054054055,
|
| 97 |
+
"grad_norm": 32.85737228393555,
|
| 98 |
+
"learning_rate": 4.777644230769231e-05,
|
| 99 |
+
"loss": 0.8372,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.15135135135135136,
|
| 104 |
+
"grad_norm": 16.32477378845215,
|
| 105 |
+
"learning_rate": 4.717548076923077e-05,
|
| 106 |
+
"loss": 0.653,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.16216216216216217,
|
| 111 |
+
"grad_norm": 11.066850662231445,
|
| 112 |
+
"learning_rate": 4.657451923076923e-05,
|
| 113 |
+
"loss": 0.6233,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.17297297297297298,
|
| 118 |
+
"grad_norm": 4.975943565368652,
|
| 119 |
+
"learning_rate": 4.5973557692307694e-05,
|
| 120 |
+
"loss": 1.0554,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.1837837837837838,
|
| 125 |
+
"grad_norm": 11.884200096130371,
|
| 126 |
+
"learning_rate": 4.5372596153846156e-05,
|
| 127 |
+
"loss": 0.6109,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.1945945945945946,
|
| 132 |
+
"grad_norm": 11.737774848937988,
|
| 133 |
+
"learning_rate": 4.477163461538462e-05,
|
| 134 |
+
"loss": 0.4309,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.20108108108108108,
|
| 139 |
+
"eval_accuracy": 0.7219917012448133,
|
| 140 |
+
"eval_loss": 0.652153491973877,
|
| 141 |
+
"eval_runtime": 361.5654,
|
| 142 |
+
"eval_samples_per_second": 0.667,
|
| 143 |
+
"eval_steps_per_second": 0.169,
|
| 144 |
+
"step": 186
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.0043243243243243,
|
| 148 |
+
"grad_norm": 4.409055709838867,
|
| 149 |
+
"learning_rate": 4.417067307692308e-05,
|
| 150 |
+
"loss": 0.5466,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.0151351351351352,
|
| 155 |
+
"grad_norm": 5.313009262084961,
|
| 156 |
+
"learning_rate": 4.3569711538461535e-05,
|
| 157 |
+
"loss": 0.0821,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.025945945945946,
|
| 162 |
+
"grad_norm": 1.3797563314437866,
|
| 163 |
+
"learning_rate": 4.2968750000000004e-05,
|
| 164 |
+
"loss": 0.2795,
|
| 165 |
+
"step": 210
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.0367567567567568,
|
| 169 |
+
"grad_norm": 1.0455204248428345,
|
| 170 |
+
"learning_rate": 4.2367788461538466e-05,
|
| 171 |
+
"loss": 0.798,
|
| 172 |
+
"step": 220
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.0475675675675675,
|
| 176 |
+
"grad_norm": 17.128549575805664,
|
| 177 |
+
"learning_rate": 4.176682692307692e-05,
|
| 178 |
+
"loss": 0.3279,
|
| 179 |
+
"step": 230
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.0583783783783784,
|
| 183 |
+
"grad_norm": 18.71080207824707,
|
| 184 |
+
"learning_rate": 4.116586538461539e-05,
|
| 185 |
+
"loss": 0.5378,
|
| 186 |
+
"step": 240
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.0691891891891891,
|
| 190 |
+
"grad_norm": 15.340435028076172,
|
| 191 |
+
"learning_rate": 4.0564903846153846e-05,
|
| 192 |
+
"loss": 0.3799,
|
| 193 |
+
"step": 250
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.08,
|
| 197 |
+
"grad_norm": 3.928271532058716,
|
| 198 |
+
"learning_rate": 3.996394230769231e-05,
|
| 199 |
+
"loss": 0.3517,
|
| 200 |
+
"step": 260
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.0908108108108108,
|
| 204 |
+
"grad_norm": 14.76327133178711,
|
| 205 |
+
"learning_rate": 3.936298076923077e-05,
|
| 206 |
+
"loss": 0.5122,
|
| 207 |
+
"step": 270
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.1016216216216217,
|
| 211 |
+
"grad_norm": 18.79231071472168,
|
| 212 |
+
"learning_rate": 3.876201923076923e-05,
|
| 213 |
+
"loss": 0.4962,
|
| 214 |
+
"step": 280
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.1124324324324324,
|
| 218 |
+
"grad_norm": 19.632802963256836,
|
| 219 |
+
"learning_rate": 3.8161057692307694e-05,
|
| 220 |
+
"loss": 0.5597,
|
| 221 |
+
"step": 290
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.1232432432432433,
|
| 225 |
+
"grad_norm": 19.008041381835938,
|
| 226 |
+
"learning_rate": 3.7560096153846156e-05,
|
| 227 |
+
"loss": 0.3686,
|
| 228 |
+
"step": 300
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.134054054054054,
|
| 232 |
+
"grad_norm": 19.428789138793945,
|
| 233 |
+
"learning_rate": 3.695913461538462e-05,
|
| 234 |
+
"loss": 0.4503,
|
| 235 |
+
"step": 310
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.144864864864865,
|
| 239 |
+
"grad_norm": 14.225759506225586,
|
| 240 |
+
"learning_rate": 3.635817307692308e-05,
|
| 241 |
+
"loss": 0.4095,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.1556756756756756,
|
| 246 |
+
"grad_norm": 4.436385631561279,
|
| 247 |
+
"learning_rate": 3.5757211538461535e-05,
|
| 248 |
+
"loss": 0.3878,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.1664864864864866,
|
| 253 |
+
"grad_norm": 0.6838646531105042,
|
| 254 |
+
"learning_rate": 3.5156250000000004e-05,
|
| 255 |
+
"loss": 0.3264,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.1772972972972973,
|
| 260 |
+
"grad_norm": 0.13564491271972656,
|
| 261 |
+
"learning_rate": 3.4555288461538466e-05,
|
| 262 |
+
"loss": 0.4799,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.1881081081081082,
|
| 267 |
+
"grad_norm": 7.611846923828125,
|
| 268 |
+
"learning_rate": 3.395432692307692e-05,
|
| 269 |
+
"loss": 0.3353,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.1989189189189189,
|
| 274 |
+
"grad_norm": 19.25046730041504,
|
| 275 |
+
"learning_rate": 3.335336538461539e-05,
|
| 276 |
+
"loss": 0.6819,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.201081081081081,
|
| 281 |
+
"eval_accuracy": 0.7551867219917012,
|
| 282 |
+
"eval_loss": 0.7018738985061646,
|
| 283 |
+
"eval_runtime": 356.8316,
|
| 284 |
+
"eval_samples_per_second": 0.675,
|
| 285 |
+
"eval_steps_per_second": 0.171,
|
| 286 |
+
"step": 372
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 2.0086486486486486,
|
| 290 |
+
"grad_norm": 0.20281045138835907,
|
| 291 |
+
"learning_rate": 3.2752403846153846e-05,
|
| 292 |
+
"loss": 0.1651,
|
| 293 |
+
"step": 380
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 2.0194594594594593,
|
| 297 |
+
"grad_norm": 28.839569091796875,
|
| 298 |
+
"learning_rate": 3.215144230769231e-05,
|
| 299 |
+
"loss": 0.3506,
|
| 300 |
+
"step": 390
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.0302702702702704,
|
| 304 |
+
"grad_norm": 22.747913360595703,
|
| 305 |
+
"learning_rate": 3.155048076923077e-05,
|
| 306 |
+
"loss": 0.6242,
|
| 307 |
+
"step": 400
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.041081081081081,
|
| 311 |
+
"grad_norm": 29.88939094543457,
|
| 312 |
+
"learning_rate": 3.094951923076923e-05,
|
| 313 |
+
"loss": 0.4168,
|
| 314 |
+
"step": 410
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.051891891891892,
|
| 318 |
+
"grad_norm": 0.13965100049972534,
|
| 319 |
+
"learning_rate": 3.0348557692307694e-05,
|
| 320 |
+
"loss": 0.0609,
|
| 321 |
+
"step": 420
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.0627027027027025,
|
| 325 |
+
"grad_norm": 7.307075500488281,
|
| 326 |
+
"learning_rate": 2.974759615384616e-05,
|
| 327 |
+
"loss": 0.4065,
|
| 328 |
+
"step": 430
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.0735135135135137,
|
| 332 |
+
"grad_norm": 0.021403660997748375,
|
| 333 |
+
"learning_rate": 2.9146634615384614e-05,
|
| 334 |
+
"loss": 0.3352,
|
| 335 |
+
"step": 440
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 2.0843243243243244,
|
| 339 |
+
"grad_norm": 17.655433654785156,
|
| 340 |
+
"learning_rate": 2.854567307692308e-05,
|
| 341 |
+
"loss": 0.1129,
|
| 342 |
+
"step": 450
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.095135135135135,
|
| 346 |
+
"grad_norm": 0.23121196031570435,
|
| 347 |
+
"learning_rate": 2.794471153846154e-05,
|
| 348 |
+
"loss": 0.1074,
|
| 349 |
+
"step": 460
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.1059459459459458,
|
| 353 |
+
"grad_norm": 0.016081752255558968,
|
| 354 |
+
"learning_rate": 2.734375e-05,
|
| 355 |
+
"loss": 0.5961,
|
| 356 |
+
"step": 470
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.116756756756757,
|
| 360 |
+
"grad_norm": 0.013322114944458008,
|
| 361 |
+
"learning_rate": 2.6742788461538466e-05,
|
| 362 |
+
"loss": 0.2839,
|
| 363 |
+
"step": 480
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.1275675675675676,
|
| 367 |
+
"grad_norm": 4.522771835327148,
|
| 368 |
+
"learning_rate": 2.6141826923076925e-05,
|
| 369 |
+
"loss": 0.0814,
|
| 370 |
+
"step": 490
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 2.1383783783783783,
|
| 374 |
+
"grad_norm": 0.02786172181367874,
|
| 375 |
+
"learning_rate": 2.5540865384615387e-05,
|
| 376 |
+
"loss": 0.0448,
|
| 377 |
+
"step": 500
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.149189189189189,
|
| 381 |
+
"grad_norm": 16.996612548828125,
|
| 382 |
+
"learning_rate": 2.493990384615385e-05,
|
| 383 |
+
"loss": 0.3222,
|
| 384 |
+
"step": 510
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.16,
|
| 388 |
+
"grad_norm": 1.9226462841033936,
|
| 389 |
+
"learning_rate": 2.4338942307692307e-05,
|
| 390 |
+
"loss": 0.1569,
|
| 391 |
+
"step": 520
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.170810810810811,
|
| 395 |
+
"grad_norm": 1.0920050144195557,
|
| 396 |
+
"learning_rate": 2.373798076923077e-05,
|
| 397 |
+
"loss": 0.1651,
|
| 398 |
+
"step": 530
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.1816216216216215,
|
| 402 |
+
"grad_norm": 0.230145663022995,
|
| 403 |
+
"learning_rate": 2.313701923076923e-05,
|
| 404 |
+
"loss": 0.2688,
|
| 405 |
+
"step": 540
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.1924324324324322,
|
| 409 |
+
"grad_norm": 5.873709201812744,
|
| 410 |
+
"learning_rate": 2.2536057692307694e-05,
|
| 411 |
+
"loss": 0.5011,
|
| 412 |
+
"step": 550
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.2010810810810812,
|
| 416 |
+
"eval_accuracy": 0.8215767634854771,
|
| 417 |
+
"eval_loss": 0.5440417528152466,
|
| 418 |
+
"eval_runtime": 588.0061,
|
| 419 |
+
"eval_samples_per_second": 0.41,
|
| 420 |
+
"eval_steps_per_second": 0.104,
|
| 421 |
+
"step": 558
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.002162162162162,
|
| 425 |
+
"grad_norm": 12.916172981262207,
|
| 426 |
+
"learning_rate": 2.1935096153846156e-05,
|
| 427 |
+
"loss": 0.6714,
|
| 428 |
+
"step": 560
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.012972972972973,
|
| 432 |
+
"grad_norm": 18.048948287963867,
|
| 433 |
+
"learning_rate": 2.1334134615384614e-05,
|
| 434 |
+
"loss": 0.1235,
|
| 435 |
+
"step": 570
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.023783783783784,
|
| 439 |
+
"grad_norm": 0.9115002751350403,
|
| 440 |
+
"learning_rate": 2.073317307692308e-05,
|
| 441 |
+
"loss": 0.1749,
|
| 442 |
+
"step": 580
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.0345945945945947,
|
| 446 |
+
"grad_norm": 18.67428207397461,
|
| 447 |
+
"learning_rate": 2.0132211538461542e-05,
|
| 448 |
+
"loss": 0.1547,
|
| 449 |
+
"step": 590
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.0454054054054054,
|
| 453 |
+
"grad_norm": 0.07008351385593414,
|
| 454 |
+
"learning_rate": 1.953125e-05,
|
| 455 |
+
"loss": 0.0246,
|
| 456 |
+
"step": 600
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.056216216216216,
|
| 460 |
+
"grad_norm": 0.0067512416280806065,
|
| 461 |
+
"learning_rate": 1.8930288461538462e-05,
|
| 462 |
+
"loss": 0.0041,
|
| 463 |
+
"step": 610
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.0670270270270272,
|
| 467 |
+
"grad_norm": 0.01588040590286255,
|
| 468 |
+
"learning_rate": 1.832932692307692e-05,
|
| 469 |
+
"loss": 0.1187,
|
| 470 |
+
"step": 620
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.077837837837838,
|
| 474 |
+
"grad_norm": 13.167031288146973,
|
| 475 |
+
"learning_rate": 1.7728365384615387e-05,
|
| 476 |
+
"loss": 0.0238,
|
| 477 |
+
"step": 630
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.0886486486486486,
|
| 481 |
+
"grad_norm": 0.06922053545713425,
|
| 482 |
+
"learning_rate": 1.712740384615385e-05,
|
| 483 |
+
"loss": 0.1063,
|
| 484 |
+
"step": 640
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.0994594594594593,
|
| 488 |
+
"grad_norm": 74.04322814941406,
|
| 489 |
+
"learning_rate": 1.6526442307692307e-05,
|
| 490 |
+
"loss": 0.376,
|
| 491 |
+
"step": 650
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.1102702702702705,
|
| 495 |
+
"grad_norm": 32.97873306274414,
|
| 496 |
+
"learning_rate": 1.592548076923077e-05,
|
| 497 |
+
"loss": 0.4791,
|
| 498 |
+
"step": 660
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.121081081081081,
|
| 502 |
+
"grad_norm": 0.11083012819290161,
|
| 503 |
+
"learning_rate": 1.532451923076923e-05,
|
| 504 |
+
"loss": 0.0166,
|
| 505 |
+
"step": 670
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.131891891891892,
|
| 509 |
+
"grad_norm": 0.01234583742916584,
|
| 510 |
+
"learning_rate": 1.4723557692307693e-05,
|
| 511 |
+
"loss": 0.0195,
|
| 512 |
+
"step": 680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.1427027027027026,
|
| 516 |
+
"grad_norm": 19.750049591064453,
|
| 517 |
+
"learning_rate": 1.4122596153846154e-05,
|
| 518 |
+
"loss": 0.1691,
|
| 519 |
+
"step": 690
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.1535135135135137,
|
| 523 |
+
"grad_norm": 0.6507856249809265,
|
| 524 |
+
"learning_rate": 1.3521634615384616e-05,
|
| 525 |
+
"loss": 0.2198,
|
| 526 |
+
"step": 700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.1643243243243244,
|
| 530 |
+
"grad_norm": 0.04525623098015785,
|
| 531 |
+
"learning_rate": 1.292067307692308e-05,
|
| 532 |
+
"loss": 0.3103,
|
| 533 |
+
"step": 710
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 3.175135135135135,
|
| 537 |
+
"grad_norm": 0.009144825860857964,
|
| 538 |
+
"learning_rate": 1.231971153846154e-05,
|
| 539 |
+
"loss": 0.0903,
|
| 540 |
+
"step": 720
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 3.185945945945946,
|
| 544 |
+
"grad_norm": 6.008044719696045,
|
| 545 |
+
"learning_rate": 1.171875e-05,
|
| 546 |
+
"loss": 0.2296,
|
| 547 |
+
"step": 730
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 3.1967567567567565,
|
| 551 |
+
"grad_norm": 0.15373368561267853,
|
| 552 |
+
"learning_rate": 1.111778846153846e-05,
|
| 553 |
+
"loss": 0.3274,
|
| 554 |
+
"step": 740
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 3.2010810810810812,
|
| 558 |
+
"eval_accuracy": 0.8257261410788381,
|
| 559 |
+
"eval_loss": 0.683351993560791,
|
| 560 |
+
"eval_runtime": 459.2726,
|
| 561 |
+
"eval_samples_per_second": 0.525,
|
| 562 |
+
"eval_steps_per_second": 0.133,
|
| 563 |
+
"step": 744
|
| 564 |
},
|
| 565 |
{
|
| 566 |
+
"epoch": 4.006486486486486,
|
| 567 |
+
"grad_norm": 0.008657727390527725,
|
| 568 |
+
"learning_rate": 1.0516826923076924e-05,
|
| 569 |
+
"loss": 0.1495,
|
| 570 |
+
"step": 750
|
|
|
|
|
|
|
| 571 |
},
|
| 572 |
{
|
| 573 |
+
"epoch": 4.017297297297297,
|
| 574 |
+
"grad_norm": 0.033678218722343445,
|
| 575 |
+
"learning_rate": 9.915865384615385e-06,
|
| 576 |
+
"loss": 0.0142,
|
| 577 |
+
"step": 760
|
|
|
|
|
|
|
| 578 |
},
|
| 579 |
{
|
| 580 |
+
"epoch": 4.028108108108108,
|
| 581 |
+
"grad_norm": 50.353580474853516,
|
| 582 |
+
"learning_rate": 9.314903846153847e-06,
|
| 583 |
+
"loss": 0.1612,
|
| 584 |
+
"step": 770
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.0389189189189185,
|
| 588 |
+
"grad_norm": 0.035760682076215744,
|
| 589 |
+
"learning_rate": 8.713942307692307e-06,
|
| 590 |
+
"loss": 0.2428,
|
| 591 |
+
"step": 780
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.04972972972973,
|
| 595 |
+
"grad_norm": 12.059436798095703,
|
| 596 |
+
"learning_rate": 8.112980769230769e-06,
|
| 597 |
+
"loss": 0.2061,
|
| 598 |
+
"step": 790
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.060540540540541,
|
| 602 |
+
"grad_norm": 0.006958181504160166,
|
| 603 |
+
"learning_rate": 7.512019230769231e-06,
|
| 604 |
+
"loss": 0.0098,
|
| 605 |
+
"step": 800
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 4.0713513513513515,
|
| 609 |
+
"grad_norm": 0.04333299770951271,
|
| 610 |
+
"learning_rate": 6.911057692307693e-06,
|
| 611 |
+
"loss": 0.2669,
|
| 612 |
+
"step": 810
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 4.082162162162162,
|
| 616 |
+
"grad_norm": 0.02775268815457821,
|
| 617 |
+
"learning_rate": 6.310096153846154e-06,
|
| 618 |
+
"loss": 0.0539,
|
| 619 |
+
"step": 820
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 4.092972972972973,
|
| 623 |
+
"grad_norm": 0.48154217004776,
|
| 624 |
+
"learning_rate": 5.709134615384616e-06,
|
| 625 |
+
"loss": 0.0897,
|
| 626 |
+
"step": 830
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.103783783783784,
|
| 630 |
+
"grad_norm": 0.012652811594307423,
|
| 631 |
+
"learning_rate": 5.108173076923077e-06,
|
| 632 |
+
"loss": 0.0778,
|
| 633 |
+
"step": 840
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.114594594594594,
|
| 637 |
+
"grad_norm": 0.006140949670225382,
|
| 638 |
+
"learning_rate": 4.507211538461539e-06,
|
| 639 |
+
"loss": 0.0017,
|
| 640 |
+
"step": 850
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 4.125405405405405,
|
| 644 |
+
"grad_norm": 11.695226669311523,
|
| 645 |
+
"learning_rate": 3.90625e-06,
|
| 646 |
+
"loss": 0.0203,
|
| 647 |
+
"step": 860
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 4.136216216216217,
|
| 651 |
+
"grad_norm": 0.015799295157194138,
|
| 652 |
+
"learning_rate": 3.3052884615384617e-06,
|
| 653 |
+
"loss": 0.0149,
|
| 654 |
+
"step": 870
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 4.147027027027027,
|
| 658 |
+
"grad_norm": 21.965198516845703,
|
| 659 |
+
"learning_rate": 2.7043269230769233e-06,
|
| 660 |
+
"loss": 0.3756,
|
| 661 |
+
"step": 880
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 4.157837837837838,
|
| 665 |
+
"grad_norm": 0.17932982742786407,
|
| 666 |
+
"learning_rate": 2.103365384615385e-06,
|
| 667 |
+
"loss": 0.1004,
|
| 668 |
+
"step": 890
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.168648648648649,
|
| 672 |
+
"grad_norm": 0.005346647929400206,
|
| 673 |
+
"learning_rate": 1.5024038461538464e-06,
|
| 674 |
+
"loss": 0.1684,
|
| 675 |
+
"step": 900
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 4.179459459459459,
|
| 679 |
+
"grad_norm": 6.568191051483154,
|
| 680 |
+
"learning_rate": 9.014423076923077e-07,
|
| 681 |
+
"loss": 0.0342,
|
| 682 |
+
"step": 910
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 4.19027027027027,
|
| 686 |
+
"grad_norm": 0.2238602489233017,
|
| 687 |
+
"learning_rate": 3.0048076923076924e-07,
|
| 688 |
+
"loss": 0.0083,
|
| 689 |
+
"step": 920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 4.195675675675676,
|
| 693 |
+
"eval_accuracy": 0.8547717842323651,
|
| 694 |
+
"eval_loss": 0.5500421524047852,
|
| 695 |
+
"eval_runtime": 452.3168,
|
| 696 |
+
"eval_samples_per_second": 0.533,
|
| 697 |
+
"eval_steps_per_second": 0.135,
|
| 698 |
+
"step": 925
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.195675675675676,
|
| 702 |
+
"step": 925,
|
| 703 |
+
"total_flos": 3.2312722580082524e+18,
|
| 704 |
+
"train_loss": 0.3789604688234426,
|
| 705 |
+
"train_runtime": 19695.7134,
|
| 706 |
+
"train_samples_per_second": 0.188,
|
| 707 |
+
"train_steps_per_second": 0.047
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 4.195675675675676,
|
| 711 |
+
"eval_accuracy": 0.9676113360323887,
|
| 712 |
+
"eval_loss": 0.1311430186033249,
|
| 713 |
+
"eval_runtime": 1467.8713,
|
| 714 |
+
"eval_samples_per_second": 0.505,
|
| 715 |
"eval_steps_per_second": 0.127,
|
| 716 |
+
"step": 925
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.195675675675676,
|
| 720 |
+
"eval_accuracy": 0.8269662921348314,
|
| 721 |
+
"eval_loss": 0.7350317239761353,
|
| 722 |
+
"eval_runtime": 796.0622,
|
| 723 |
+
"eval_samples_per_second": 0.559,
|
| 724 |
+
"eval_steps_per_second": 0.141,
|
| 725 |
+
"step": 925
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 4.195675675675676,
|
| 729 |
+
"eval_accuracy": 0.8547717842323651,
|
| 730 |
+
"eval_loss": 0.5500420928001404,
|
| 731 |
+
"eval_runtime": 434.3892,
|
| 732 |
+
"eval_samples_per_second": 0.555,
|
| 733 |
+
"eval_steps_per_second": 0.14,
|
| 734 |
+
"step": 925
|
| 735 |
}
|
| 736 |
],
|
| 737 |
"logging_steps": 10,
|
| 738 |
+
"max_steps": 925,
|
| 739 |
"num_input_tokens_seen": 0,
|
| 740 |
"num_train_epochs": 9223372036854775807,
|
| 741 |
"save_steps": 500,
|
|
|
|
| 751 |
"attributes": {}
|
| 752 |
}
|
| 753 |
},
|
| 754 |
+
"total_flos": 3.2312722580082524e+18,
|
| 755 |
"train_batch_size": 4,
|
| 756 |
"trial_name": null,
|
| 757 |
"trial_params": null
|
val_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 0.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.195675675675676,
|
| 3 |
+
"eval_accuracy": 0.8547717842323651,
|
| 4 |
+
"eval_loss": 0.5500420928001404,
|
| 5 |
+
"eval_runtime": 434.3892,
|
| 6 |
+
"eval_samples_per_second": 0.555,
|
| 7 |
+
"eval_steps_per_second": 0.14
|
| 8 |
}
|