Commit
•
96d0021
1
Parent(s):
2722c1e
doodle-dash
Browse files- README.md +22 -3
- all_results.json +10 -10
- model.safetensors +1 -1
- preprocessor_config.json +3 -3
- test_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +765 -17
- training_args.bin +1 -1
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 [apple/mobilevit-small](https://huggingface.co/apple/mobilevit-small) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss:
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -42,12 +42,31 @@ The following hyperparameters were used during training:
|
|
| 42 |
- eval_batch_size: 256
|
| 43 |
- seed: 42
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
-
- lr_scheduler_type:
|
| 46 |
- num_epochs: 5
|
| 47 |
- mixed_precision_training: Native AMP
|
| 48 |
|
| 49 |
### Training results
|
| 50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
|
| 52 |
|
| 53 |
### Framework versions
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [apple/mobilevit-small](https://huggingface.co/apple/mobilevit-small) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.9890
|
| 21 |
+
- Accuracy: 0.7494
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 42 |
- eval_batch_size: 256
|
| 43 |
- seed: 42
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
+
- lr_scheduler_type: cosine
|
| 46 |
- num_epochs: 5
|
| 47 |
- mixed_precision_training: Native AMP
|
| 48 |
|
| 49 |
### Training results
|
| 50 |
|
| 51 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 52 |
+
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
|
| 53 |
+
| 1.4668 | 0.28 | 5000 | 1.4146 | 0.6488 |
|
| 54 |
+
| 1.3282 | 0.57 | 10000 | 1.2907 | 0.6766 |
|
| 55 |
+
| 1.2617 | 0.85 | 15000 | 1.2270 | 0.6905 |
|
| 56 |
+
| 1.196 | 1.14 | 20000 | 1.1758 | 0.7035 |
|
| 57 |
+
| 1.1664 | 1.42 | 25000 | 1.1527 | 0.7093 |
|
| 58 |
+
| 1.1504 | 1.71 | 30000 | 1.1152 | 0.7170 |
|
| 59 |
+
| 1.1234 | 1.99 | 35000 | 1.0903 | 0.7241 |
|
| 60 |
+
| 1.0819 | 2.28 | 40000 | 1.0728 | 0.7283 |
|
| 61 |
+
| 1.0707 | 2.56 | 45000 | 1.0533 | 0.7334 |
|
| 62 |
+
| 1.049 | 2.84 | 50000 | 1.0399 | 0.7369 |
|
| 63 |
+
| 1.0017 | 3.13 | 55000 | 1.0253 | 0.7405 |
|
| 64 |
+
| 0.995 | 3.41 | 60000 | 1.0120 | 0.7438 |
|
| 65 |
+
| 0.9829 | 3.7 | 65000 | 0.9977 | 0.7468 |
|
| 66 |
+
| 0.9807 | 3.98 | 70000 | 0.9908 | 0.7487 |
|
| 67 |
+
| 0.9452 | 4.27 | 75000 | 0.9875 | 0.7498 |
|
| 68 |
+
| 0.949 | 4.55 | 80000 | 0.9845 | 0.7507 |
|
| 69 |
+
| 0.9509 | 4.84 | 85000 | 0.9841 | 0.7509 |
|
| 70 |
|
| 71 |
|
| 72 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 1.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 1.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.749444,
|
| 4 |
+
"eval_loss": 0.9889541864395142,
|
| 5 |
+
"eval_runtime": 480.919,
|
| 6 |
+
"eval_samples_per_second": 519.838,
|
| 7 |
+
"eval_steps_per_second": 2.032,
|
| 8 |
+
"total_flos": 5.4597447576e+17,
|
| 9 |
+
"train_loss": 1.1248133655402486,
|
| 10 |
+
"train_runtime": 51144.106,
|
| 11 |
+
"train_samples_per_second": 439.933,
|
| 12 |
+
"train_steps_per_second": 1.719
|
| 13 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 20730036
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3cdbd0fe9d0acb014744296fad548e9c5803a702dc65c640d5a141db652944dd
|
| 3 |
size 20730036
|
preprocessor_config.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"crop_size": {
|
| 3 |
-
"height":
|
| 4 |
-
"width":
|
| 5 |
},
|
| 6 |
"do_center_crop": true,
|
| 7 |
"do_convert_rgb": false,
|
|
@@ -12,6 +12,6 @@
|
|
| 12 |
"resample": 2,
|
| 13 |
"rescale_factor": 0.00392156862745098,
|
| 14 |
"size": {
|
| 15 |
-
"shortest_edge":
|
| 16 |
}
|
| 17 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"crop_size": {
|
| 3 |
+
"height": 28,
|
| 4 |
+
"width": 28
|
| 5 |
},
|
| 6 |
"do_center_crop": true,
|
| 7 |
"do_convert_rgb": false,
|
|
|
|
| 12 |
"resample": 2,
|
| 13 |
"rescale_factor": 0.00392156862745098,
|
| 14 |
"size": {
|
| 15 |
+
"shortest_edge": 28
|
| 16 |
}
|
| 17 |
}
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.749444,
|
| 4 |
+
"eval_loss": 0.9889541864395142,
|
| 5 |
+
"eval_runtime": 480.919,
|
| 6 |
+
"eval_samples_per_second": 519.838,
|
| 7 |
+
"eval_steps_per_second": 2.032
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 1.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"total_flos": 5.4597447576e+17,
|
| 4 |
+
"train_loss": 1.1248133655402486,
|
| 5 |
+
"train_runtime": 51144.106,
|
| 6 |
+
"train_samples_per_second": 439.933,
|
| 7 |
+
"train_steps_per_second": 1.719
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -3,41 +3,789 @@
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 5.0,
|
| 5 |
"eval_steps": 5000,
|
| 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":
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 0.
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 1000
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch":
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 0.
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 2000
|
| 24 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
{
|
| 26 |
"epoch": 5.0,
|
| 27 |
-
"step":
|
| 28 |
-
"total_flos":
|
| 29 |
-
"train_loss": 1.
|
| 30 |
-
"train_runtime":
|
| 31 |
-
"train_samples_per_second":
|
| 32 |
-
"train_steps_per_second": 1.
|
| 33 |
}
|
| 34 |
],
|
| 35 |
"logging_steps": 1000,
|
| 36 |
-
"max_steps":
|
| 37 |
"num_input_tokens_seen": 0,
|
| 38 |
"num_train_epochs": 5,
|
| 39 |
"save_steps": 5000,
|
| 40 |
-
"total_flos":
|
| 41 |
"train_batch_size": 256,
|
| 42 |
"trial_name": null,
|
| 43 |
"trial_params": null
|
|
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 5.0,
|
| 5 |
"eval_steps": 5000,
|
| 6 |
+
"global_step": 87895,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.06,
|
| 13 |
+
"grad_norm": 2.239067792892456,
|
| 14 |
+
"learning_rate": 0.0007997460514380593,
|
| 15 |
+
"loss": 2.6519,
|
| 16 |
"step": 1000
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.11,
|
| 20 |
+
"grad_norm": 1.9583632946014404,
|
| 21 |
+
"learning_rate": 0.0007989814716192582,
|
| 22 |
+
"loss": 1.7524,
|
| 23 |
"step": 2000
|
| 24 |
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.17,
|
| 27 |
+
"grad_norm": 1.9535893201828003,
|
| 28 |
+
"learning_rate": 0.0007977072349143816,
|
| 29 |
+
"loss": 1.6001,
|
| 30 |
+
"step": 3000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.23,
|
| 34 |
+
"grad_norm": 2.063387632369995,
|
| 35 |
+
"learning_rate": 0.0007959249690268624,
|
| 36 |
+
"loss": 1.5217,
|
| 37 |
+
"step": 4000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.28,
|
| 41 |
+
"grad_norm": 1.9860559701919556,
|
| 42 |
+
"learning_rate": 0.0007936369506140068,
|
| 43 |
+
"loss": 1.4668,
|
| 44 |
+
"step": 5000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.28,
|
| 48 |
+
"eval_accuracy": 0.648804,
|
| 49 |
+
"eval_loss": 1.4146429300308228,
|
| 50 |
+
"eval_runtime": 478.1117,
|
| 51 |
+
"eval_samples_per_second": 522.89,
|
| 52 |
+
"eval_steps_per_second": 2.043,
|
| 53 |
+
"step": 5000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.34,
|
| 57 |
+
"grad_norm": 2.0301945209503174,
|
| 58 |
+
"learning_rate": 0.0007908461023788039,
|
| 59 |
+
"loss": 1.4298,
|
| 60 |
+
"step": 6000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.4,
|
| 64 |
+
"grad_norm": 1.5042238235473633,
|
| 65 |
+
"learning_rate": 0.0007875595274863339,
|
| 66 |
+
"loss": 1.3942,
|
| 67 |
+
"step": 7000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.46,
|
| 71 |
+
"grad_norm": 1.6014552116394043,
|
| 72 |
+
"learning_rate": 0.0007837748451621974,
|
| 73 |
+
"loss": 1.3678,
|
| 74 |
+
"step": 8000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.51,
|
| 78 |
+
"grad_norm": 1.509202241897583,
|
| 79 |
+
"learning_rate": 0.0007795044488381889,
|
| 80 |
+
"loss": 1.3487,
|
| 81 |
+
"step": 9000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.57,
|
| 85 |
+
"grad_norm": 1.4688481092453003,
|
| 86 |
+
"learning_rate": 0.0007747452450657864,
|
| 87 |
+
"loss": 1.3282,
|
| 88 |
+
"step": 10000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.57,
|
| 92 |
+
"eval_accuracy": 0.676572,
|
| 93 |
+
"eval_loss": 1.2906984090805054,
|
| 94 |
+
"eval_runtime": 477.0964,
|
| 95 |
+
"eval_samples_per_second": 524.003,
|
| 96 |
+
"eval_steps_per_second": 2.048,
|
| 97 |
+
"step": 10000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.63,
|
| 101 |
+
"grad_norm": 1.4521937370300293,
|
| 102 |
+
"learning_rate": 0.0007695128184733147,
|
| 103 |
+
"loss": 1.3244,
|
| 104 |
+
"step": 11000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.68,
|
| 108 |
+
"grad_norm": 1.4638675451278687,
|
| 109 |
+
"learning_rate": 0.000763803378723512,
|
| 110 |
+
"loss": 1.2934,
|
| 111 |
+
"step": 12000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.74,
|
| 115 |
+
"grad_norm": 1.614662528038025,
|
| 116 |
+
"learning_rate": 0.0007576292184031235,
|
| 117 |
+
"loss": 1.2847,
|
| 118 |
+
"step": 13000
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.8,
|
| 122 |
+
"grad_norm": 1.3933509588241577,
|
| 123 |
+
"learning_rate": 0.0007510050807642281,
|
| 124 |
+
"loss": 1.2779,
|
| 125 |
+
"step": 14000
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.85,
|
| 129 |
+
"grad_norm": 1.424271583557129,
|
| 130 |
+
"learning_rate": 0.0007439261673227427,
|
| 131 |
+
"loss": 1.2617,
|
| 132 |
+
"step": 15000
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.85,
|
| 136 |
+
"eval_accuracy": 0.690468,
|
| 137 |
+
"eval_loss": 1.227000117301941,
|
| 138 |
+
"eval_runtime": 474.9613,
|
| 139 |
+
"eval_samples_per_second": 526.359,
|
| 140 |
+
"eval_steps_per_second": 2.057,
|
| 141 |
+
"step": 15000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.91,
|
| 145 |
+
"grad_norm": 1.5942885875701904,
|
| 146 |
+
"learning_rate": 0.0007364156588963428,
|
| 147 |
+
"loss": 1.2525,
|
| 148 |
+
"step": 16000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.97,
|
| 152 |
+
"grad_norm": 1.342366099357605,
|
| 153 |
+
"learning_rate": 0.0007284681149194631,
|
| 154 |
+
"loss": 1.2437,
|
| 155 |
+
"step": 17000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 1.02,
|
| 159 |
+
"grad_norm": 1.3556225299835205,
|
| 160 |
+
"learning_rate": 0.0007201095606528187,
|
| 161 |
+
"loss": 1.2252,
|
| 162 |
+
"step": 18000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.08,
|
| 166 |
+
"grad_norm": 1.18772554397583,
|
| 167 |
+
"learning_rate": 0.0007113339412117152,
|
| 168 |
+
"loss": 1.1982,
|
| 169 |
+
"step": 19000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.14,
|
| 173 |
+
"grad_norm": 1.2163102626800537,
|
| 174 |
+
"learning_rate": 0.0007021606253904673,
|
| 175 |
+
"loss": 1.196,
|
| 176 |
+
"step": 20000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.14,
|
| 180 |
+
"eval_accuracy": 0.703472,
|
| 181 |
+
"eval_loss": 1.1758021116256714,
|
| 182 |
+
"eval_runtime": 477.2497,
|
| 183 |
+
"eval_samples_per_second": 523.835,
|
| 184 |
+
"eval_steps_per_second": 2.047,
|
| 185 |
+
"step": 20000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.19,
|
| 189 |
+
"grad_norm": 1.4264676570892334,
|
| 190 |
+
"learning_rate": 0.0006926013311356693,
|
| 191 |
+
"loss": 1.1909,
|
| 192 |
+
"step": 21000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.25,
|
| 196 |
+
"grad_norm": 1.4394527673721313,
|
| 197 |
+
"learning_rate": 0.0006826783850151473,
|
| 198 |
+
"loss": 1.1906,
|
| 199 |
+
"step": 22000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.31,
|
| 203 |
+
"grad_norm": 1.6702680587768555,
|
| 204 |
+
"learning_rate": 0.0006723845989084832,
|
| 205 |
+
"loss": 1.1836,
|
| 206 |
+
"step": 23000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.37,
|
| 210 |
+
"grad_norm": 1.4739456176757812,
|
| 211 |
+
"learning_rate": 0.0006617536811497173,
|
| 212 |
+
"loss": 1.1826,
|
| 213 |
+
"step": 24000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.42,
|
| 217 |
+
"grad_norm": 1.8557323217391968,
|
| 218 |
+
"learning_rate": 0.0006507779307681826,
|
| 219 |
+
"loss": 1.1664,
|
| 220 |
+
"step": 25000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.42,
|
| 224 |
+
"eval_accuracy": 0.709336,
|
| 225 |
+
"eval_loss": 1.1526598930358887,
|
| 226 |
+
"eval_runtime": 478.812,
|
| 227 |
+
"eval_samples_per_second": 522.126,
|
| 228 |
+
"eval_steps_per_second": 2.04,
|
| 229 |
+
"step": 25000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 1.48,
|
| 233 |
+
"grad_norm": 1.0533519983291626,
|
| 234 |
+
"learning_rate": 0.0006394818379437445,
|
| 235 |
+
"loss": 1.1641,
|
| 236 |
+
"step": 26000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.54,
|
| 240 |
+
"grad_norm": 1.3917081356048584,
|
| 241 |
+
"learning_rate": 0.0006278798322474558,
|
| 242 |
+
"loss": 1.1606,
|
| 243 |
+
"step": 27000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.59,
|
| 247 |
+
"grad_norm": 1.281437873840332,
|
| 248 |
+
"learning_rate": 0.000616010800690931,
|
| 249 |
+
"loss": 1.1553,
|
| 250 |
+
"step": 28000
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.65,
|
| 254 |
+
"grad_norm": 1.2246198654174805,
|
| 255 |
+
"learning_rate": 0.0006038423385201748,
|
| 256 |
+
"loss": 1.149,
|
| 257 |
+
"step": 29000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.71,
|
| 261 |
+
"grad_norm": 1.2770333290100098,
|
| 262 |
+
"learning_rate": 0.0005914134891913911,
|
| 263 |
+
"loss": 1.1504,
|
| 264 |
+
"step": 30000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.71,
|
| 268 |
+
"eval_accuracy": 0.717008,
|
| 269 |
+
"eval_loss": 1.115194320678711,
|
| 270 |
+
"eval_runtime": 479.3599,
|
| 271 |
+
"eval_samples_per_second": 521.529,
|
| 272 |
+
"eval_steps_per_second": 2.038,
|
| 273 |
+
"step": 30000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.76,
|
| 277 |
+
"grad_norm": 1.062719702720642,
|
| 278 |
+
"learning_rate": 0.0005787401292529838,
|
| 279 |
+
"loss": 1.1433,
|
| 280 |
+
"step": 31000
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.82,
|
| 284 |
+
"grad_norm": 1.4529552459716797,
|
| 285 |
+
"learning_rate": 0.0005658514578562903,
|
| 286 |
+
"loss": 1.1445,
|
| 287 |
+
"step": 32000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.88,
|
| 291 |
+
"grad_norm": 1.107438564300537,
|
| 292 |
+
"learning_rate": 0.000552738138527959,
|
| 293 |
+
"loss": 1.1345,
|
| 294 |
+
"step": 33000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.93,
|
| 298 |
+
"grad_norm": 1.309801697731018,
|
| 299 |
+
"learning_rate": 0.0005394431125282525,
|
| 300 |
+
"loss": 1.1255,
|
| 301 |
+
"step": 34000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.99,
|
| 305 |
+
"grad_norm": 1.2301312685012817,
|
| 306 |
+
"learning_rate": 0.0005259567490230731,
|
| 307 |
+
"loss": 1.1234,
|
| 308 |
+
"step": 35000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.99,
|
| 312 |
+
"eval_accuracy": 0.724084,
|
| 313 |
+
"eval_loss": 1.0902520418167114,
|
| 314 |
+
"eval_runtime": 480.0595,
|
| 315 |
+
"eval_samples_per_second": 520.769,
|
| 316 |
+
"eval_steps_per_second": 2.035,
|
| 317 |
+
"step": 35000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 2.05,
|
| 321 |
+
"grad_norm": 1.4310553073883057,
|
| 322 |
+
"learning_rate": 0.0005123232108540917,
|
| 323 |
+
"loss": 1.0933,
|
| 324 |
+
"step": 36000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.1,
|
| 328 |
+
"grad_norm": 1.1885666847229004,
|
| 329 |
+
"learning_rate": 0.0004985326219711018,
|
| 330 |
+
"loss": 1.0855,
|
| 331 |
+
"step": 37000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.16,
|
| 335 |
+
"grad_norm": 1.2444497346878052,
|
| 336 |
+
"learning_rate": 0.0004846161680220303,
|
| 337 |
+
"loss": 1.0785,
|
| 338 |
+
"step": 38000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.22,
|
| 342 |
+
"grad_norm": 1.5714720487594604,
|
| 343 |
+
"learning_rate": 0.0004705916258137954,
|
| 344 |
+
"loss": 1.0748,
|
| 345 |
+
"step": 39000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.28,
|
| 349 |
+
"grad_norm": 1.5065577030181885,
|
| 350 |
+
"learning_rate": 0.0004564910639893322,
|
| 351 |
+
"loss": 1.0819,
|
| 352 |
+
"step": 40000
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.28,
|
| 356 |
+
"eval_accuracy": 0.728284,
|
| 357 |
+
"eval_loss": 1.072808027267456,
|
| 358 |
+
"eval_runtime": 478.1415,
|
| 359 |
+
"eval_samples_per_second": 522.858,
|
| 360 |
+
"eval_steps_per_second": 2.043,
|
| 361 |
+
"step": 40000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.33,
|
| 365 |
+
"grad_norm": 1.459184169769287,
|
| 366 |
+
"learning_rate": 0.00044230426818940436,
|
| 367 |
+
"loss": 1.077,
|
| 368 |
+
"step": 41000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.39,
|
| 372 |
+
"grad_norm": 1.0881271362304688,
|
| 373 |
+
"learning_rate": 0.0004280776949118281,
|
| 374 |
+
"loss": 1.0763,
|
| 375 |
+
"step": 42000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.45,
|
| 379 |
+
"grad_norm": 1.257857322692871,
|
| 380 |
+
"learning_rate": 0.000413801038515455,
|
| 381 |
+
"loss": 1.0654,
|
| 382 |
+
"step": 43000
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.5,
|
| 386 |
+
"grad_norm": 1.3307342529296875,
|
| 387 |
+
"learning_rate": 0.000399521049758247,
|
| 388 |
+
"loss": 1.0652,
|
| 389 |
+
"step": 44000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.56,
|
| 393 |
+
"grad_norm": 1.1706323623657227,
|
| 394 |
+
"learning_rate": 0.00038522738430521474,
|
| 395 |
+
"loss": 1.0707,
|
| 396 |
+
"step": 45000
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.56,
|
| 400 |
+
"eval_accuracy": 0.733432,
|
| 401 |
+
"eval_loss": 1.0532697439193726,
|
| 402 |
+
"eval_runtime": 479.4432,
|
| 403 |
+
"eval_samples_per_second": 521.438,
|
| 404 |
+
"eval_steps_per_second": 2.038,
|
| 405 |
+
"step": 45000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.62,
|
| 409 |
+
"grad_norm": 1.3150368928909302,
|
| 410 |
+
"learning_rate": 0.00037096684861267625,
|
| 411 |
+
"loss": 1.0657,
|
| 412 |
+
"step": 46000
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.67,
|
| 416 |
+
"grad_norm": 1.3128031492233276,
|
| 417 |
+
"learning_rate": 0.00035672911243732087,
|
| 418 |
+
"loss": 1.0601,
|
| 419 |
+
"step": 47000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.73,
|
| 423 |
+
"grad_norm": 1.3551691770553589,
|
| 424 |
+
"learning_rate": 0.0003425607990878131,
|
| 425 |
+
"loss": 1.0575,
|
| 426 |
+
"step": 48000
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.79,
|
| 430 |
+
"grad_norm": 1.2833973169326782,
|
| 431 |
+
"learning_rate": 0.0003284516451261337,
|
| 432 |
+
"loss": 1.059,
|
| 433 |
+
"step": 49000
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.84,
|
| 437 |
+
"grad_norm": 1.4435667991638184,
|
| 438 |
+
"learning_rate": 0.000314447852801857,
|
| 439 |
+
"loss": 1.049,
|
| 440 |
+
"step": 50000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.84,
|
| 444 |
+
"eval_accuracy": 0.736904,
|
| 445 |
+
"eval_loss": 1.039907693862915,
|
| 446 |
+
"eval_runtime": 482.5464,
|
| 447 |
+
"eval_samples_per_second": 518.085,
|
| 448 |
+
"eval_steps_per_second": 2.025,
|
| 449 |
+
"step": 50000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 2.9,
|
| 453 |
+
"grad_norm": 1.2984023094177246,
|
| 454 |
+
"learning_rate": 0.0003005392778549707,
|
| 455 |
+
"loss": 1.0418,
|
| 456 |
+
"step": 51000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 2.96,
|
| 460 |
+
"grad_norm": 1.109937310218811,
|
| 461 |
+
"learning_rate": 0.0002867714657131215,
|
| 462 |
+
"loss": 1.048,
|
| 463 |
+
"step": 52000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.01,
|
| 467 |
+
"grad_norm": 1.1215381622314453,
|
| 468 |
+
"learning_rate": 0.00027313444305887276,
|
| 469 |
+
"loss": 1.0303,
|
| 470 |
+
"step": 53000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.07,
|
| 474 |
+
"grad_norm": 1.3672548532485962,
|
| 475 |
+
"learning_rate": 0.00025967286608676553,
|
| 476 |
+
"loss": 1.0047,
|
| 477 |
+
"step": 54000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.13,
|
| 481 |
+
"grad_norm": 1.4645944833755493,
|
| 482 |
+
"learning_rate": 0.0002463769833142144,
|
| 483 |
+
"loss": 1.0017,
|
| 484 |
+
"step": 55000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.13,
|
| 488 |
+
"eval_accuracy": 0.740544,
|
| 489 |
+
"eval_loss": 1.0252685546875,
|
| 490 |
+
"eval_runtime": 480.6997,
|
| 491 |
+
"eval_samples_per_second": 520.075,
|
| 492 |
+
"eval_steps_per_second": 2.032,
|
| 493 |
+
"step": 55000
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 3.19,
|
| 497 |
+
"grad_norm": 1.3108264207839966,
|
| 498 |
+
"learning_rate": 0.00023329033382909358,
|
| 499 |
+
"loss": 1.0012,
|
| 500 |
+
"step": 56000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.24,
|
| 504 |
+
"grad_norm": 1.4289699792861938,
|
| 505 |
+
"learning_rate": 0.00022040343774395584,
|
| 506 |
+
"loss": 0.9987,
|
| 507 |
+
"step": 57000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.3,
|
| 511 |
+
"grad_norm": 1.2855439186096191,
|
| 512 |
+
"learning_rate": 0.00020775849486686778,
|
| 513 |
+
"loss": 0.9988,
|
| 514 |
+
"step": 58000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 3.36,
|
| 518 |
+
"grad_norm": 1.162169098854065,
|
| 519 |
+
"learning_rate": 0.00019534634527990013,
|
| 520 |
+
"loss": 1.0006,
|
| 521 |
+
"step": 59000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 3.41,
|
| 525 |
+
"grad_norm": 1.2256358861923218,
|
| 526 |
+
"learning_rate": 0.00018320763417230612,
|
| 527 |
+
"loss": 0.995,
|
| 528 |
+
"step": 60000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 3.41,
|
| 532 |
+
"eval_accuracy": 0.743764,
|
| 533 |
+
"eval_loss": 1.0119822025299072,
|
| 534 |
+
"eval_runtime": 480.4029,
|
| 535 |
+
"eval_samples_per_second": 520.396,
|
| 536 |
+
"eval_steps_per_second": 2.034,
|
| 537 |
+
"step": 60000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 3.47,
|
| 541 |
+
"grad_norm": 1.2014998197555542,
|
| 542 |
+
"learning_rate": 0.00017133356833730004,
|
| 543 |
+
"loss": 0.9976,
|
| 544 |
+
"step": 61000
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 3.53,
|
| 548 |
+
"grad_norm": 1.1588609218597412,
|
| 549 |
+
"learning_rate": 0.00015975159982961664,
|
| 550 |
+
"loss": 1.001,
|
| 551 |
+
"step": 62000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 3.58,
|
| 555 |
+
"grad_norm": 1.2534708976745605,
|
| 556 |
+
"learning_rate": 0.00014847652339644127,
|
| 557 |
+
"loss": 0.9967,
|
| 558 |
+
"step": 63000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 3.64,
|
| 562 |
+
"grad_norm": 1.2062395811080933,
|
| 563 |
+
"learning_rate": 0.00013753353036528426,
|
| 564 |
+
"loss": 0.9972,
|
| 565 |
+
"step": 64000
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 3.7,
|
| 569 |
+
"grad_norm": 1.280290126800537,
|
| 570 |
+
"learning_rate": 0.00012691469370009338,
|
| 571 |
+
"loss": 0.9829,
|
| 572 |
+
"step": 65000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.7,
|
| 576 |
+
"eval_accuracy": 0.746796,
|
| 577 |
+
"eval_loss": 0.9977088570594788,
|
| 578 |
+
"eval_runtime": 476.9935,
|
| 579 |
+
"eval_samples_per_second": 524.116,
|
| 580 |
+
"eval_steps_per_second": 2.048,
|
| 581 |
+
"step": 65000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 3.75,
|
| 585 |
+
"grad_norm": 1.2085012197494507,
|
| 586 |
+
"learning_rate": 0.00011665478615500634,
|
| 587 |
+
"loss": 0.9863,
|
| 588 |
+
"step": 66000
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 3.81,
|
| 592 |
+
"grad_norm": 1.7737034559249878,
|
| 593 |
+
"learning_rate": 0.00010674637551684559,
|
| 594 |
+
"loss": 0.9777,
|
| 595 |
+
"step": 67000
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 3.87,
|
| 599 |
+
"grad_norm": 1.1058114767074585,
|
| 600 |
+
"learning_rate": 9.722190815783432e-05,
|
| 601 |
+
"loss": 0.9903,
|
| 602 |
+
"step": 68000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.93,
|
| 606 |
+
"grad_norm": 1.285079836845398,
|
| 607 |
+
"learning_rate": 8.807448461308951e-05,
|
| 608 |
+
"loss": 0.9814,
|
| 609 |
+
"step": 69000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 3.98,
|
| 613 |
+
"grad_norm": 1.4150310754776,
|
| 614 |
+
"learning_rate": 7.933405930416787e-05,
|
| 615 |
+
"loss": 0.9807,
|
| 616 |
+
"step": 70000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 3.98,
|
| 620 |
+
"eval_accuracy": 0.7487,
|
| 621 |
+
"eval_loss": 0.9908215999603271,
|
| 622 |
+
"eval_runtime": 478.9031,
|
| 623 |
+
"eval_samples_per_second": 522.026,
|
| 624 |
+
"eval_steps_per_second": 2.04,
|
| 625 |
+
"step": 70000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 4.04,
|
| 629 |
+
"grad_norm": 1.0839868783950806,
|
| 630 |
+
"learning_rate": 7.100243228624242e-05,
|
| 631 |
+
"loss": 0.963,
|
| 632 |
+
"step": 71000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 4.1,
|
| 636 |
+
"grad_norm": 1.2299129962921143,
|
| 637 |
+
"learning_rate": 6.30825184828638e-05,
|
| 638 |
+
"loss": 0.955,
|
| 639 |
+
"step": 72000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 4.15,
|
| 643 |
+
"grad_norm": 1.3780415058135986,
|
| 644 |
+
"learning_rate": 5.5592981342836236e-05,
|
| 645 |
+
"loss": 0.9508,
|
| 646 |
+
"step": 73000
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 4.21,
|
| 650 |
+
"grad_norm": 1.1320440769195557,
|
| 651 |
+
"learning_rate": 4.8543387962500266e-05,
|
| 652 |
+
"loss": 0.9493,
|
| 653 |
+
"step": 74000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 4.27,
|
| 657 |
+
"grad_norm": 1.120684027671814,
|
| 658 |
+
"learning_rate": 4.1949116996191016e-05,
|
| 659 |
+
"loss": 0.9452,
|
| 660 |
+
"step": 75000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 4.27,
|
| 664 |
+
"eval_accuracy": 0.749816,
|
| 665 |
+
"eval_loss": 0.9875096082687378,
|
| 666 |
+
"eval_runtime": 477.8512,
|
| 667 |
+
"eval_samples_per_second": 523.175,
|
| 668 |
+
"eval_steps_per_second": 2.045,
|
| 669 |
+
"step": 75000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 4.32,
|
| 673 |
+
"grad_norm": 1.4629569053649902,
|
| 674 |
+
"learning_rate": 3.5805391595057494e-05,
|
| 675 |
+
"loss": 0.9484,
|
| 676 |
+
"step": 76000
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 4.38,
|
| 680 |
+
"grad_norm": 1.2522917985916138,
|
| 681 |
+
"learning_rate": 3.013233018257653e-05,
|
| 682 |
+
"loss": 0.9567,
|
| 683 |
+
"step": 77000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 4.44,
|
| 687 |
+
"grad_norm": 1.331236720085144,
|
| 688 |
+
"learning_rate": 2.492582322836503e-05,
|
| 689 |
+
"loss": 0.9494,
|
| 690 |
+
"step": 78000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 4.49,
|
| 694 |
+
"grad_norm": 1.312769889831543,
|
| 695 |
+
"learning_rate": 2.0202920112872213e-05,
|
| 696 |
+
"loss": 0.9494,
|
| 697 |
+
"step": 79000
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 4.55,
|
| 701 |
+
"grad_norm": 1.0436272621154785,
|
| 702 |
+
"learning_rate": 1.5960199594472392e-05,
|
| 703 |
+
"loss": 0.949,
|
| 704 |
+
"step": 80000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 4.55,
|
| 708 |
+
"eval_accuracy": 0.750728,
|
| 709 |
+
"eval_loss": 0.9844790697097778,
|
| 710 |
+
"eval_runtime": 480.1647,
|
| 711 |
+
"eval_samples_per_second": 520.655,
|
| 712 |
+
"eval_steps_per_second": 2.035,
|
| 713 |
+
"step": 80000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 4.61,
|
| 717 |
+
"grad_norm": 1.7521427869796753,
|
| 718 |
+
"learning_rate": 1.221155501027127e-05,
|
| 719 |
+
"loss": 0.9459,
|
| 720 |
+
"step": 81000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 4.66,
|
| 724 |
+
"grad_norm": 1.4406906366348267,
|
| 725 |
+
"learning_rate": 8.954270865460369e-06,
|
| 726 |
+
"loss": 0.9507,
|
| 727 |
+
"step": 82000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 4.72,
|
| 731 |
+
"grad_norm": 1.2117027044296265,
|
| 732 |
+
"learning_rate": 6.196506529965529e-06,
|
| 733 |
+
"loss": 0.9471,
|
| 734 |
+
"step": 83000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.78,
|
| 738 |
+
"grad_norm": 1.1171038150787354,
|
| 739 |
+
"learning_rate": 3.943787197483806e-06,
|
| 740 |
+
"loss": 0.951,
|
| 741 |
+
"step": 84000
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 4.84,
|
| 745 |
+
"grad_norm": 1.388543963432312,
|
| 746 |
+
"learning_rate": 2.194481012149785e-06,
|
| 747 |
+
"loss": 0.9509,
|
| 748 |
+
"step": 85000
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 4.84,
|
| 752 |
+
"eval_accuracy": 0.750944,
|
| 753 |
+
"eval_loss": 0.984071671962738,
|
| 754 |
+
"eval_runtime": 479.0227,
|
| 755 |
+
"eval_samples_per_second": 521.896,
|
| 756 |
+
"eval_steps_per_second": 2.04,
|
| 757 |
+
"step": 85000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 4.89,
|
| 761 |
+
"grad_norm": 1.2518051862716675,
|
| 762 |
+
"learning_rate": 9.543163033286728e-07,
|
| 763 |
+
"loss": 0.955,
|
| 764 |
+
"step": 86000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 4.95,
|
| 768 |
+
"grad_norm": 1.1738619804382324,
|
| 769 |
+
"learning_rate": 2.2239470342309e-07,
|
| 770 |
+
"loss": 0.9482,
|
| 771 |
+
"step": 87000
|
| 772 |
+
},
|
| 773 |
{
|
| 774 |
"epoch": 5.0,
|
| 775 |
+
"step": 87895,
|
| 776 |
+
"total_flos": 5.4597447576e+17,
|
| 777 |
+
"train_loss": 1.1248133655402486,
|
| 778 |
+
"train_runtime": 51144.106,
|
| 779 |
+
"train_samples_per_second": 439.933,
|
| 780 |
+
"train_steps_per_second": 1.719
|
| 781 |
}
|
| 782 |
],
|
| 783 |
"logging_steps": 1000,
|
| 784 |
+
"max_steps": 87895,
|
| 785 |
"num_input_tokens_seen": 0,
|
| 786 |
"num_train_epochs": 5,
|
| 787 |
"save_steps": 5000,
|
| 788 |
+
"total_flos": 5.4597447576e+17,
|
| 789 |
"train_batch_size": 256,
|
| 790 |
"trial_name": null,
|
| 791 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4856
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d9a6f1a8233545c2d422a032a79a3101f63ca1427a7c955c87c0125bf8df20e
|
| 3 |
size 4856
|