End of training
Browse files
README.md
CHANGED
|
@@ -3,6 +3,8 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: c14kevincardenas/beit-large-patch16-384-limb-person-crop
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
@@ -16,10 +18,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# finetuned-beit-limb-seq-t4-8heads-1layers-1e-4lr
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [c14kevincardenas/beit-large-patch16-384-limb-person-crop](https://huggingface.co/c14kevincardenas/beit-large-patch16-384-limb-person-crop) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Accuracy: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: c14kevincardenas/beit-large-patch16-384-limb-person-crop
|
| 5 |
tags:
|
| 6 |
+
- image-sequence-classification
|
| 7 |
+
- vision
|
| 8 |
- generated_from_trainer
|
| 9 |
metrics:
|
| 10 |
- accuracy
|
|
|
|
| 18 |
|
| 19 |
# finetuned-beit-limb-seq-t4-8heads-1layers-1e-4lr
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [c14kevincardenas/beit-large-patch16-384-limb-person-crop](https://huggingface.co/c14kevincardenas/beit-large-patch16-384-limb-person-crop) on the c14kevincardenas/beta_caller_284_person_crop_seq_withlimb_4 dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.3927
|
| 24 |
+
- Accuracy: 0.9174
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.9173652694610779,
|
| 4 |
+
"eval_loss": 0.3927362263202667,
|
| 5 |
+
"eval_runtime": 59.0745,
|
| 6 |
+
"eval_samples_per_second": 14.135,
|
| 7 |
+
"eval_steps_per_second": 0.457,
|
| 8 |
+
"total_flos": 0.0,
|
| 9 |
+
"train_loss": 0.43191270973231344,
|
| 10 |
+
"train_runtime": 11059.7547,
|
| 11 |
+
"train_samples_per_second": 8.552,
|
| 12 |
+
"train_steps_per_second": 0.268
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.9173652694610779,
|
| 4 |
+
"eval_loss": 0.3927362263202667,
|
| 5 |
+
"eval_runtime": 59.0745,
|
| 6 |
+
"eval_samples_per_second": 14.135,
|
| 7 |
+
"eval_steps_per_second": 0.457
|
| 8 |
+
}
|
runs/Feb05_18-47-22_galactica.ad.cirange.net/events.out.tfevents.1738792601.galactica.ad.cirange.net.1303711.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:608b5795b3127874b09fee1a0e8f87e2754d1ed32cb7ca1ea340d33ef4165d5e
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.43191270973231344,
|
| 5 |
+
"train_runtime": 11059.7547,
|
| 6 |
+
"train_samples_per_second": 8.552,
|
| 7 |
+
"train_steps_per_second": 0.268
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1048 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.3927362263202667,
|
| 3 |
+
"best_model_checkpoint": "limb_classification_person_crop_seq/t4_8heads_1layers_1e-4lr/checkpoint-1776",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2960,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.16891891891891891,
|
| 13 |
+
"grad_norm": 358243.25,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 1.3942,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.33783783783783783,
|
| 20 |
+
"grad_norm": 251778.359375,
|
| 21 |
+
"learning_rate": 2e-05,
|
| 22 |
+
"loss": 1.3603,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.5067567567567568,
|
| 27 |
+
"grad_norm": 132885.078125,
|
| 28 |
+
"learning_rate": 3e-05,
|
| 29 |
+
"loss": 1.2986,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.6756756756756757,
|
| 34 |
+
"grad_norm": 586557.4375,
|
| 35 |
+
"learning_rate": 4e-05,
|
| 36 |
+
"loss": 1.1933,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.8445945945945946,
|
| 41 |
+
"grad_norm": 172983.421875,
|
| 42 |
+
"learning_rate": 5e-05,
|
| 43 |
+
"loss": 1.0563,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.0,
|
| 48 |
+
"eval_accuracy": 0.8754491017964072,
|
| 49 |
+
"eval_loss": 0.5863036513328552,
|
| 50 |
+
"eval_runtime": 60.375,
|
| 51 |
+
"eval_samples_per_second": 13.83,
|
| 52 |
+
"eval_steps_per_second": 0.447,
|
| 53 |
+
"step": 148
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.0135135135135136,
|
| 57 |
+
"grad_norm": 176986.609375,
|
| 58 |
+
"learning_rate": 6e-05,
|
| 59 |
+
"loss": 0.7536,
|
| 60 |
+
"step": 150
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.1824324324324325,
|
| 64 |
+
"grad_norm": 141175.546875,
|
| 65 |
+
"learning_rate": 7e-05,
|
| 66 |
+
"loss": 0.5917,
|
| 67 |
+
"step": 175
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.3513513513513513,
|
| 71 |
+
"grad_norm": 171763.8125,
|
| 72 |
+
"learning_rate": 8e-05,
|
| 73 |
+
"loss": 0.5229,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.5202702702702702,
|
| 78 |
+
"grad_norm": 235346.625,
|
| 79 |
+
"learning_rate": 9e-05,
|
| 80 |
+
"loss": 0.4497,
|
| 81 |
+
"step": 225
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.689189189189189,
|
| 85 |
+
"grad_norm": 91142.6953125,
|
| 86 |
+
"learning_rate": 0.0001,
|
| 87 |
+
"loss": 0.443,
|
| 88 |
+
"step": 250
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.8581081081081081,
|
| 92 |
+
"grad_norm": 174055.0,
|
| 93 |
+
"learning_rate": 9.907749077490776e-05,
|
| 94 |
+
"loss": 0.4235,
|
| 95 |
+
"step": 275
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0,
|
| 99 |
+
"eval_accuracy": 0.8982035928143712,
|
| 100 |
+
"eval_loss": 0.4248776435852051,
|
| 101 |
+
"eval_runtime": 60.0285,
|
| 102 |
+
"eval_samples_per_second": 13.91,
|
| 103 |
+
"eval_steps_per_second": 0.45,
|
| 104 |
+
"step": 296
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.027027027027027,
|
| 108 |
+
"grad_norm": 457668.3125,
|
| 109 |
+
"learning_rate": 9.81549815498155e-05,
|
| 110 |
+
"loss": 0.4702,
|
| 111 |
+
"step": 300
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 2.195945945945946,
|
| 115 |
+
"grad_norm": 132193.171875,
|
| 116 |
+
"learning_rate": 9.723247232472326e-05,
|
| 117 |
+
"loss": 0.4457,
|
| 118 |
+
"step": 325
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.364864864864865,
|
| 122 |
+
"grad_norm": 96091.71875,
|
| 123 |
+
"learning_rate": 9.6309963099631e-05,
|
| 124 |
+
"loss": 0.4403,
|
| 125 |
+
"step": 350
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.5337837837837838,
|
| 129 |
+
"grad_norm": 161375.140625,
|
| 130 |
+
"learning_rate": 9.538745387453874e-05,
|
| 131 |
+
"loss": 0.423,
|
| 132 |
+
"step": 375
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 2.7027027027027026,
|
| 136 |
+
"grad_norm": 130612.0078125,
|
| 137 |
+
"learning_rate": 9.44649446494465e-05,
|
| 138 |
+
"loss": 0.43,
|
| 139 |
+
"step": 400
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.8716216216216215,
|
| 143 |
+
"grad_norm": 238536.953125,
|
| 144 |
+
"learning_rate": 9.354243542435425e-05,
|
| 145 |
+
"loss": 0.4494,
|
| 146 |
+
"step": 425
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.0,
|
| 150 |
+
"eval_accuracy": 0.9089820359281438,
|
| 151 |
+
"eval_loss": 0.42482611536979675,
|
| 152 |
+
"eval_runtime": 59.7234,
|
| 153 |
+
"eval_samples_per_second": 13.981,
|
| 154 |
+
"eval_steps_per_second": 0.452,
|
| 155 |
+
"step": 444
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.0405405405405403,
|
| 159 |
+
"grad_norm": 60061.19921875,
|
| 160 |
+
"learning_rate": 9.2619926199262e-05,
|
| 161 |
+
"loss": 0.4362,
|
| 162 |
+
"step": 450
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.2094594594594597,
|
| 166 |
+
"grad_norm": 96977.171875,
|
| 167 |
+
"learning_rate": 9.169741697416975e-05,
|
| 168 |
+
"loss": 0.4302,
|
| 169 |
+
"step": 475
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.3783783783783785,
|
| 173 |
+
"grad_norm": 98531.53125,
|
| 174 |
+
"learning_rate": 9.077490774907749e-05,
|
| 175 |
+
"loss": 0.409,
|
| 176 |
+
"step": 500
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.5472972972972974,
|
| 180 |
+
"grad_norm": 101922.609375,
|
| 181 |
+
"learning_rate": 8.985239852398525e-05,
|
| 182 |
+
"loss": 0.3979,
|
| 183 |
+
"step": 525
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.7162162162162162,
|
| 187 |
+
"grad_norm": 108008.8671875,
|
| 188 |
+
"learning_rate": 8.892988929889299e-05,
|
| 189 |
+
"loss": 0.4353,
|
| 190 |
+
"step": 550
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.885135135135135,
|
| 194 |
+
"grad_norm": 93226.5390625,
|
| 195 |
+
"learning_rate": 8.800738007380073e-05,
|
| 196 |
+
"loss": 0.4183,
|
| 197 |
+
"step": 575
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 4.0,
|
| 201 |
+
"eval_accuracy": 0.8934131736526946,
|
| 202 |
+
"eval_loss": 0.43285271525382996,
|
| 203 |
+
"eval_runtime": 60.1768,
|
| 204 |
+
"eval_samples_per_second": 13.876,
|
| 205 |
+
"eval_steps_per_second": 0.449,
|
| 206 |
+
"step": 592
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.054054054054054,
|
| 210 |
+
"grad_norm": 104521.6328125,
|
| 211 |
+
"learning_rate": 8.708487084870849e-05,
|
| 212 |
+
"loss": 0.4753,
|
| 213 |
+
"step": 600
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.222972972972973,
|
| 217 |
+
"grad_norm": 67537.2265625,
|
| 218 |
+
"learning_rate": 8.616236162361624e-05,
|
| 219 |
+
"loss": 0.3917,
|
| 220 |
+
"step": 625
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 4.391891891891892,
|
| 224 |
+
"grad_norm": 150912.1875,
|
| 225 |
+
"learning_rate": 8.523985239852399e-05,
|
| 226 |
+
"loss": 0.4119,
|
| 227 |
+
"step": 650
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 4.5608108108108105,
|
| 231 |
+
"grad_norm": 73587.5078125,
|
| 232 |
+
"learning_rate": 8.431734317343174e-05,
|
| 233 |
+
"loss": 0.4247,
|
| 234 |
+
"step": 675
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.72972972972973,
|
| 238 |
+
"grad_norm": 115034.75,
|
| 239 |
+
"learning_rate": 8.339483394833948e-05,
|
| 240 |
+
"loss": 0.4285,
|
| 241 |
+
"step": 700
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 4.898648648648649,
|
| 245 |
+
"grad_norm": 99327.8828125,
|
| 246 |
+
"learning_rate": 8.247232472324724e-05,
|
| 247 |
+
"loss": 0.4203,
|
| 248 |
+
"step": 725
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 5.0,
|
| 252 |
+
"eval_accuracy": 0.918562874251497,
|
| 253 |
+
"eval_loss": 0.39791449904441833,
|
| 254 |
+
"eval_runtime": 60.3518,
|
| 255 |
+
"eval_samples_per_second": 13.836,
|
| 256 |
+
"eval_steps_per_second": 0.447,
|
| 257 |
+
"step": 740
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 5.0675675675675675,
|
| 261 |
+
"grad_norm": 198293.546875,
|
| 262 |
+
"learning_rate": 8.154981549815498e-05,
|
| 263 |
+
"loss": 0.4155,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 5.236486486486487,
|
| 268 |
+
"grad_norm": 82787.34375,
|
| 269 |
+
"learning_rate": 8.062730627306274e-05,
|
| 270 |
+
"loss": 0.3911,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 5.405405405405405,
|
| 275 |
+
"grad_norm": 114961.9765625,
|
| 276 |
+
"learning_rate": 7.970479704797048e-05,
|
| 277 |
+
"loss": 0.4102,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 5.574324324324325,
|
| 282 |
+
"grad_norm": 153899.546875,
|
| 283 |
+
"learning_rate": 7.878228782287823e-05,
|
| 284 |
+
"loss": 0.419,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 5.743243243243243,
|
| 289 |
+
"grad_norm": 108072.7109375,
|
| 290 |
+
"learning_rate": 7.785977859778598e-05,
|
| 291 |
+
"loss": 0.4064,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.912162162162162,
|
| 296 |
+
"grad_norm": 73436.2421875,
|
| 297 |
+
"learning_rate": 7.693726937269373e-05,
|
| 298 |
+
"loss": 0.4073,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 6.0,
|
| 303 |
+
"eval_accuracy": 0.9149700598802395,
|
| 304 |
+
"eval_loss": 0.39525091648101807,
|
| 305 |
+
"eval_runtime": 59.9209,
|
| 306 |
+
"eval_samples_per_second": 13.935,
|
| 307 |
+
"eval_steps_per_second": 0.451,
|
| 308 |
+
"step": 888
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 6.081081081081081,
|
| 312 |
+
"grad_norm": 109643.8359375,
|
| 313 |
+
"learning_rate": 7.601476014760149e-05,
|
| 314 |
+
"loss": 0.4227,
|
| 315 |
+
"step": 900
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 6.25,
|
| 319 |
+
"grad_norm": 103467.765625,
|
| 320 |
+
"learning_rate": 7.509225092250923e-05,
|
| 321 |
+
"loss": 0.3968,
|
| 322 |
+
"step": 925
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 6.418918918918919,
|
| 326 |
+
"grad_norm": 114479.0859375,
|
| 327 |
+
"learning_rate": 7.416974169741697e-05,
|
| 328 |
+
"loss": 0.4118,
|
| 329 |
+
"step": 950
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 6.587837837837838,
|
| 333 |
+
"grad_norm": 171495.734375,
|
| 334 |
+
"learning_rate": 7.324723247232473e-05,
|
| 335 |
+
"loss": 0.4134,
|
| 336 |
+
"step": 975
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 6.756756756756757,
|
| 340 |
+
"grad_norm": 166568.453125,
|
| 341 |
+
"learning_rate": 7.232472324723247e-05,
|
| 342 |
+
"loss": 0.3839,
|
| 343 |
+
"step": 1000
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 6.925675675675675,
|
| 347 |
+
"grad_norm": 68976.8828125,
|
| 348 |
+
"learning_rate": 7.140221402214023e-05,
|
| 349 |
+
"loss": 0.3927,
|
| 350 |
+
"step": 1025
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 7.0,
|
| 354 |
+
"eval_accuracy": 0.9209580838323354,
|
| 355 |
+
"eval_loss": 0.39880236983299255,
|
| 356 |
+
"eval_runtime": 60.3856,
|
| 357 |
+
"eval_samples_per_second": 13.828,
|
| 358 |
+
"eval_steps_per_second": 0.447,
|
| 359 |
+
"step": 1036
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 7.094594594594595,
|
| 363 |
+
"grad_norm": 101431.9921875,
|
| 364 |
+
"learning_rate": 7.047970479704797e-05,
|
| 365 |
+
"loss": 0.4122,
|
| 366 |
+
"step": 1050
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 7.263513513513513,
|
| 370 |
+
"grad_norm": 44332.38671875,
|
| 371 |
+
"learning_rate": 6.955719557195572e-05,
|
| 372 |
+
"loss": 0.4072,
|
| 373 |
+
"step": 1075
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 7.4324324324324325,
|
| 377 |
+
"grad_norm": 120717.484375,
|
| 378 |
+
"learning_rate": 6.863468634686348e-05,
|
| 379 |
+
"loss": 0.4207,
|
| 380 |
+
"step": 1100
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 7.601351351351351,
|
| 384 |
+
"grad_norm": 141817.53125,
|
| 385 |
+
"learning_rate": 6.771217712177122e-05,
|
| 386 |
+
"loss": 0.3748,
|
| 387 |
+
"step": 1125
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 7.77027027027027,
|
| 391 |
+
"grad_norm": 108992.875,
|
| 392 |
+
"learning_rate": 6.678966789667896e-05,
|
| 393 |
+
"loss": 0.3973,
|
| 394 |
+
"step": 1150
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 7.9391891891891895,
|
| 398 |
+
"grad_norm": 125266.5390625,
|
| 399 |
+
"learning_rate": 6.586715867158672e-05,
|
| 400 |
+
"loss": 0.3931,
|
| 401 |
+
"step": 1175
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 8.0,
|
| 405 |
+
"eval_accuracy": 0.9173652694610779,
|
| 406 |
+
"eval_loss": 0.4065897464752197,
|
| 407 |
+
"eval_runtime": 59.0702,
|
| 408 |
+
"eval_samples_per_second": 14.136,
|
| 409 |
+
"eval_steps_per_second": 0.457,
|
| 410 |
+
"step": 1184
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 8.108108108108109,
|
| 414 |
+
"grad_norm": 79784.984375,
|
| 415 |
+
"learning_rate": 6.494464944649446e-05,
|
| 416 |
+
"loss": 0.3869,
|
| 417 |
+
"step": 1200
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 8.277027027027026,
|
| 421 |
+
"grad_norm": 58690.61328125,
|
| 422 |
+
"learning_rate": 6.402214022140222e-05,
|
| 423 |
+
"loss": 0.3796,
|
| 424 |
+
"step": 1225
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 8.445945945945946,
|
| 428 |
+
"grad_norm": 79699.40625,
|
| 429 |
+
"learning_rate": 6.309963099630997e-05,
|
| 430 |
+
"loss": 0.4066,
|
| 431 |
+
"step": 1250
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 8.614864864864865,
|
| 435 |
+
"grad_norm": 116672.609375,
|
| 436 |
+
"learning_rate": 6.217712177121771e-05,
|
| 437 |
+
"loss": 0.3723,
|
| 438 |
+
"step": 1275
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.783783783783784,
|
| 442 |
+
"grad_norm": 84205.125,
|
| 443 |
+
"learning_rate": 6.125461254612547e-05,
|
| 444 |
+
"loss": 0.419,
|
| 445 |
+
"step": 1300
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 8.952702702702704,
|
| 449 |
+
"grad_norm": 194455.796875,
|
| 450 |
+
"learning_rate": 6.033210332103322e-05,
|
| 451 |
+
"loss": 0.4225,
|
| 452 |
+
"step": 1325
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 9.0,
|
| 456 |
+
"eval_accuracy": 0.9221556886227545,
|
| 457 |
+
"eval_loss": 0.39390280842781067,
|
| 458 |
+
"eval_runtime": 59.2665,
|
| 459 |
+
"eval_samples_per_second": 14.089,
|
| 460 |
+
"eval_steps_per_second": 0.456,
|
| 461 |
+
"step": 1332
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 9.121621621621621,
|
| 465 |
+
"grad_norm": 124840.765625,
|
| 466 |
+
"learning_rate": 5.940959409594096e-05,
|
| 467 |
+
"loss": 0.3988,
|
| 468 |
+
"step": 1350
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 9.29054054054054,
|
| 472 |
+
"grad_norm": 166515.8125,
|
| 473 |
+
"learning_rate": 5.848708487084871e-05,
|
| 474 |
+
"loss": 0.4045,
|
| 475 |
+
"step": 1375
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 9.45945945945946,
|
| 479 |
+
"grad_norm": 69126.875,
|
| 480 |
+
"learning_rate": 5.756457564575646e-05,
|
| 481 |
+
"loss": 0.3878,
|
| 482 |
+
"step": 1400
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 9.628378378378379,
|
| 486 |
+
"grad_norm": 93407.296875,
|
| 487 |
+
"learning_rate": 5.664206642066421e-05,
|
| 488 |
+
"loss": 0.3846,
|
| 489 |
+
"step": 1425
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 9.797297297297296,
|
| 493 |
+
"grad_norm": 70069.578125,
|
| 494 |
+
"learning_rate": 5.5719557195571956e-05,
|
| 495 |
+
"loss": 0.4031,
|
| 496 |
+
"step": 1450
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.966216216216216,
|
| 500 |
+
"grad_norm": 89726.5859375,
|
| 501 |
+
"learning_rate": 5.479704797047971e-05,
|
| 502 |
+
"loss": 0.3768,
|
| 503 |
+
"step": 1475
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 10.0,
|
| 507 |
+
"eval_accuracy": 0.911377245508982,
|
| 508 |
+
"eval_loss": 0.404273122549057,
|
| 509 |
+
"eval_runtime": 59.2046,
|
| 510 |
+
"eval_samples_per_second": 14.104,
|
| 511 |
+
"eval_steps_per_second": 0.456,
|
| 512 |
+
"step": 1480
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 10.135135135135135,
|
| 516 |
+
"grad_norm": 50900.421875,
|
| 517 |
+
"learning_rate": 5.387453874538746e-05,
|
| 518 |
+
"loss": 0.3971,
|
| 519 |
+
"step": 1500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 10.304054054054054,
|
| 523 |
+
"grad_norm": 195171.375,
|
| 524 |
+
"learning_rate": 5.295202952029521e-05,
|
| 525 |
+
"loss": 0.4006,
|
| 526 |
+
"step": 1525
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 10.472972972972974,
|
| 530 |
+
"grad_norm": 133812.78125,
|
| 531 |
+
"learning_rate": 5.202952029520295e-05,
|
| 532 |
+
"loss": 0.3527,
|
| 533 |
+
"step": 1550
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 10.641891891891891,
|
| 537 |
+
"grad_norm": 77933.1328125,
|
| 538 |
+
"learning_rate": 5.11070110701107e-05,
|
| 539 |
+
"loss": 0.377,
|
| 540 |
+
"step": 1575
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 10.81081081081081,
|
| 544 |
+
"grad_norm": 141872.90625,
|
| 545 |
+
"learning_rate": 5.018450184501845e-05,
|
| 546 |
+
"loss": 0.4182,
|
| 547 |
+
"step": 1600
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 10.97972972972973,
|
| 551 |
+
"grad_norm": 105366.984375,
|
| 552 |
+
"learning_rate": 4.92619926199262e-05,
|
| 553 |
+
"loss": 0.3918,
|
| 554 |
+
"step": 1625
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 11.0,
|
| 558 |
+
"eval_accuracy": 0.9101796407185628,
|
| 559 |
+
"eval_loss": 0.4043692350387573,
|
| 560 |
+
"eval_runtime": 58.8135,
|
| 561 |
+
"eval_samples_per_second": 14.197,
|
| 562 |
+
"eval_steps_per_second": 0.459,
|
| 563 |
+
"step": 1628
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 11.14864864864865,
|
| 567 |
+
"grad_norm": 102680.625,
|
| 568 |
+
"learning_rate": 4.833948339483395e-05,
|
| 569 |
+
"loss": 0.4068,
|
| 570 |
+
"step": 1650
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 11.317567567567568,
|
| 574 |
+
"grad_norm": 54038.0625,
|
| 575 |
+
"learning_rate": 4.74169741697417e-05,
|
| 576 |
+
"loss": 0.3948,
|
| 577 |
+
"step": 1675
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 11.486486486486486,
|
| 581 |
+
"grad_norm": 104827.5078125,
|
| 582 |
+
"learning_rate": 4.6494464944649444e-05,
|
| 583 |
+
"loss": 0.3606,
|
| 584 |
+
"step": 1700
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 11.655405405405405,
|
| 588 |
+
"grad_norm": 120384.34375,
|
| 589 |
+
"learning_rate": 4.55719557195572e-05,
|
| 590 |
+
"loss": 0.3815,
|
| 591 |
+
"step": 1725
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 11.824324324324325,
|
| 595 |
+
"grad_norm": 236821.96875,
|
| 596 |
+
"learning_rate": 4.464944649446495e-05,
|
| 597 |
+
"loss": 0.3613,
|
| 598 |
+
"step": 1750
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 11.993243243243244,
|
| 602 |
+
"grad_norm": 83007.3359375,
|
| 603 |
+
"learning_rate": 4.37269372693727e-05,
|
| 604 |
+
"loss": 0.398,
|
| 605 |
+
"step": 1775
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 12.0,
|
| 609 |
+
"eval_accuracy": 0.9173652694610779,
|
| 610 |
+
"eval_loss": 0.3927362263202667,
|
| 611 |
+
"eval_runtime": 59.8526,
|
| 612 |
+
"eval_samples_per_second": 13.951,
|
| 613 |
+
"eval_steps_per_second": 0.451,
|
| 614 |
+
"step": 1776
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 12.162162162162161,
|
| 618 |
+
"grad_norm": 114086.703125,
|
| 619 |
+
"learning_rate": 4.280442804428044e-05,
|
| 620 |
+
"loss": 0.3695,
|
| 621 |
+
"step": 1800
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 12.33108108108108,
|
| 625 |
+
"grad_norm": 189440.21875,
|
| 626 |
+
"learning_rate": 4.1881918819188195e-05,
|
| 627 |
+
"loss": 0.3917,
|
| 628 |
+
"step": 1825
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 12.5,
|
| 632 |
+
"grad_norm": 77842.9921875,
|
| 633 |
+
"learning_rate": 4.0959409594095944e-05,
|
| 634 |
+
"loss": 0.3908,
|
| 635 |
+
"step": 1850
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 12.66891891891892,
|
| 639 |
+
"grad_norm": 106204.4609375,
|
| 640 |
+
"learning_rate": 4.003690036900369e-05,
|
| 641 |
+
"loss": 0.4054,
|
| 642 |
+
"step": 1875
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 12.837837837837839,
|
| 646 |
+
"grad_norm": 79211.3125,
|
| 647 |
+
"learning_rate": 3.911439114391144e-05,
|
| 648 |
+
"loss": 0.381,
|
| 649 |
+
"step": 1900
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 13.0,
|
| 653 |
+
"eval_accuracy": 0.9173652694610779,
|
| 654 |
+
"eval_loss": 0.39785751700401306,
|
| 655 |
+
"eval_runtime": 59.3643,
|
| 656 |
+
"eval_samples_per_second": 14.066,
|
| 657 |
+
"eval_steps_per_second": 0.455,
|
| 658 |
+
"step": 1924
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 13.006756756756756,
|
| 662 |
+
"grad_norm": 96323.5703125,
|
| 663 |
+
"learning_rate": 3.819188191881919e-05,
|
| 664 |
+
"loss": 0.3419,
|
| 665 |
+
"step": 1925
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 13.175675675675675,
|
| 669 |
+
"grad_norm": 179315.421875,
|
| 670 |
+
"learning_rate": 3.726937269372694e-05,
|
| 671 |
+
"loss": 0.3987,
|
| 672 |
+
"step": 1950
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 13.344594594594595,
|
| 676 |
+
"grad_norm": 99669.9609375,
|
| 677 |
+
"learning_rate": 3.634686346863469e-05,
|
| 678 |
+
"loss": 0.3626,
|
| 679 |
+
"step": 1975
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 13.513513513513514,
|
| 683 |
+
"grad_norm": 90959.078125,
|
| 684 |
+
"learning_rate": 3.542435424354244e-05,
|
| 685 |
+
"loss": 0.3744,
|
| 686 |
+
"step": 2000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 13.682432432432432,
|
| 690 |
+
"grad_norm": 123356.265625,
|
| 691 |
+
"learning_rate": 3.4501845018450186e-05,
|
| 692 |
+
"loss": 0.3613,
|
| 693 |
+
"step": 2025
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 13.85135135135135,
|
| 697 |
+
"grad_norm": 62648.0234375,
|
| 698 |
+
"learning_rate": 3.3579335793357934e-05,
|
| 699 |
+
"loss": 0.3817,
|
| 700 |
+
"step": 2050
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 14.0,
|
| 704 |
+
"eval_accuracy": 0.9137724550898204,
|
| 705 |
+
"eval_loss": 0.4134121239185333,
|
| 706 |
+
"eval_runtime": 60.0106,
|
| 707 |
+
"eval_samples_per_second": 13.914,
|
| 708 |
+
"eval_steps_per_second": 0.45,
|
| 709 |
+
"step": 2072
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 14.02027027027027,
|
| 713 |
+
"grad_norm": 104165.796875,
|
| 714 |
+
"learning_rate": 3.265682656826568e-05,
|
| 715 |
+
"loss": 0.3751,
|
| 716 |
+
"step": 2075
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 14.18918918918919,
|
| 720 |
+
"grad_norm": 69535.90625,
|
| 721 |
+
"learning_rate": 3.173431734317343e-05,
|
| 722 |
+
"loss": 0.3675,
|
| 723 |
+
"step": 2100
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 14.358108108108109,
|
| 727 |
+
"grad_norm": 103158.5,
|
| 728 |
+
"learning_rate": 3.081180811808118e-05,
|
| 729 |
+
"loss": 0.39,
|
| 730 |
+
"step": 2125
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 14.527027027027026,
|
| 734 |
+
"grad_norm": 68135.5859375,
|
| 735 |
+
"learning_rate": 2.9889298892988933e-05,
|
| 736 |
+
"loss": 0.3658,
|
| 737 |
+
"step": 2150
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 14.695945945945946,
|
| 741 |
+
"grad_norm": 167864.78125,
|
| 742 |
+
"learning_rate": 2.8966789667896682e-05,
|
| 743 |
+
"loss": 0.3857,
|
| 744 |
+
"step": 2175
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 14.864864864864865,
|
| 748 |
+
"grad_norm": 98201.515625,
|
| 749 |
+
"learning_rate": 2.8044280442804427e-05,
|
| 750 |
+
"loss": 0.3815,
|
| 751 |
+
"step": 2200
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 15.0,
|
| 755 |
+
"eval_accuracy": 0.9173652694610779,
|
| 756 |
+
"eval_loss": 0.40676751732826233,
|
| 757 |
+
"eval_runtime": 58.4707,
|
| 758 |
+
"eval_samples_per_second": 14.281,
|
| 759 |
+
"eval_steps_per_second": 0.462,
|
| 760 |
+
"step": 2220
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 15.033783783783784,
|
| 764 |
+
"grad_norm": 186007.90625,
|
| 765 |
+
"learning_rate": 2.7121771217712176e-05,
|
| 766 |
+
"loss": 0.3717,
|
| 767 |
+
"step": 2225
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 15.202702702702704,
|
| 771 |
+
"grad_norm": 157017.15625,
|
| 772 |
+
"learning_rate": 2.619926199261993e-05,
|
| 773 |
+
"loss": 0.3717,
|
| 774 |
+
"step": 2250
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 15.371621621621621,
|
| 778 |
+
"grad_norm": 74874.7109375,
|
| 779 |
+
"learning_rate": 2.5276752767527677e-05,
|
| 780 |
+
"loss": 0.3537,
|
| 781 |
+
"step": 2275
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 15.54054054054054,
|
| 785 |
+
"grad_norm": 92439.0078125,
|
| 786 |
+
"learning_rate": 2.4354243542435426e-05,
|
| 787 |
+
"loss": 0.3535,
|
| 788 |
+
"step": 2300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 15.70945945945946,
|
| 792 |
+
"grad_norm": 164555.34375,
|
| 793 |
+
"learning_rate": 2.3431734317343175e-05,
|
| 794 |
+
"loss": 0.3645,
|
| 795 |
+
"step": 2325
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 15.878378378378379,
|
| 799 |
+
"grad_norm": 91601.59375,
|
| 800 |
+
"learning_rate": 2.2509225092250924e-05,
|
| 801 |
+
"loss": 0.3774,
|
| 802 |
+
"step": 2350
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 16.0,
|
| 806 |
+
"eval_accuracy": 0.9233532934131736,
|
| 807 |
+
"eval_loss": 0.39439886808395386,
|
| 808 |
+
"eval_runtime": 58.9596,
|
| 809 |
+
"eval_samples_per_second": 14.162,
|
| 810 |
+
"eval_steps_per_second": 0.458,
|
| 811 |
+
"step": 2368
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 16.0472972972973,
|
| 815 |
+
"grad_norm": 82947.125,
|
| 816 |
+
"learning_rate": 2.1586715867158673e-05,
|
| 817 |
+
"loss": 0.3644,
|
| 818 |
+
"step": 2375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 16.216216216216218,
|
| 822 |
+
"grad_norm": 77509.625,
|
| 823 |
+
"learning_rate": 2.066420664206642e-05,
|
| 824 |
+
"loss": 0.3418,
|
| 825 |
+
"step": 2400
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 16.385135135135137,
|
| 829 |
+
"grad_norm": 63176.84765625,
|
| 830 |
+
"learning_rate": 1.974169741697417e-05,
|
| 831 |
+
"loss": 0.3502,
|
| 832 |
+
"step": 2425
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 16.554054054054053,
|
| 836 |
+
"grad_norm": 194086.15625,
|
| 837 |
+
"learning_rate": 1.881918819188192e-05,
|
| 838 |
+
"loss": 0.3788,
|
| 839 |
+
"step": 2450
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 16.722972972972972,
|
| 843 |
+
"grad_norm": 81391.1171875,
|
| 844 |
+
"learning_rate": 1.7896678966789668e-05,
|
| 845 |
+
"loss": 0.347,
|
| 846 |
+
"step": 2475
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 16.89189189189189,
|
| 850 |
+
"grad_norm": 34088.36328125,
|
| 851 |
+
"learning_rate": 1.6974169741697417e-05,
|
| 852 |
+
"loss": 0.3499,
|
| 853 |
+
"step": 2500
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 17.0,
|
| 857 |
+
"eval_accuracy": 0.9209580838323354,
|
| 858 |
+
"eval_loss": 0.3971590995788574,
|
| 859 |
+
"eval_runtime": 59.2527,
|
| 860 |
+
"eval_samples_per_second": 14.092,
|
| 861 |
+
"eval_steps_per_second": 0.456,
|
| 862 |
+
"step": 2516
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 17.06081081081081,
|
| 866 |
+
"grad_norm": 151931.296875,
|
| 867 |
+
"learning_rate": 1.6051660516605166e-05,
|
| 868 |
+
"loss": 0.3927,
|
| 869 |
+
"step": 2525
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 17.22972972972973,
|
| 873 |
+
"grad_norm": 76409.84375,
|
| 874 |
+
"learning_rate": 1.5129151291512916e-05,
|
| 875 |
+
"loss": 0.3497,
|
| 876 |
+
"step": 2550
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 17.39864864864865,
|
| 880 |
+
"grad_norm": 65936.53125,
|
| 881 |
+
"learning_rate": 1.4206642066420663e-05,
|
| 882 |
+
"loss": 0.363,
|
| 883 |
+
"step": 2575
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 17.56756756756757,
|
| 887 |
+
"grad_norm": 50398.3203125,
|
| 888 |
+
"learning_rate": 1.3284132841328414e-05,
|
| 889 |
+
"loss": 0.3499,
|
| 890 |
+
"step": 2600
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 17.736486486486488,
|
| 894 |
+
"grad_norm": 109634.046875,
|
| 895 |
+
"learning_rate": 1.2361623616236164e-05,
|
| 896 |
+
"loss": 0.337,
|
| 897 |
+
"step": 2625
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 17.905405405405407,
|
| 901 |
+
"grad_norm": 40496.69921875,
|
| 902 |
+
"learning_rate": 1.1439114391143913e-05,
|
| 903 |
+
"loss": 0.3825,
|
| 904 |
+
"step": 2650
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 18.0,
|
| 908 |
+
"eval_accuracy": 0.9197604790419162,
|
| 909 |
+
"eval_loss": 0.4014655351638794,
|
| 910 |
+
"eval_runtime": 59.3898,
|
| 911 |
+
"eval_samples_per_second": 14.06,
|
| 912 |
+
"eval_steps_per_second": 0.455,
|
| 913 |
+
"step": 2664
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 18.074324324324323,
|
| 917 |
+
"grad_norm": 117493.03125,
|
| 918 |
+
"learning_rate": 1.0516605166051662e-05,
|
| 919 |
+
"loss": 0.3519,
|
| 920 |
+
"step": 2675
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 18.243243243243242,
|
| 924 |
+
"grad_norm": 46973.1875,
|
| 925 |
+
"learning_rate": 9.59409594095941e-06,
|
| 926 |
+
"loss": 0.3237,
|
| 927 |
+
"step": 2700
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 18.41216216216216,
|
| 931 |
+
"grad_norm": 114745.4609375,
|
| 932 |
+
"learning_rate": 8.67158671586716e-06,
|
| 933 |
+
"loss": 0.3714,
|
| 934 |
+
"step": 2725
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 18.58108108108108,
|
| 938 |
+
"grad_norm": 38971.9296875,
|
| 939 |
+
"learning_rate": 7.749077490774908e-06,
|
| 940 |
+
"loss": 0.3231,
|
| 941 |
+
"step": 2750
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 18.75,
|
| 945 |
+
"grad_norm": 156186.84375,
|
| 946 |
+
"learning_rate": 6.826568265682657e-06,
|
| 947 |
+
"loss": 0.3416,
|
| 948 |
+
"step": 2775
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 18.91891891891892,
|
| 952 |
+
"grad_norm": 76312.9921875,
|
| 953 |
+
"learning_rate": 5.904059040590406e-06,
|
| 954 |
+
"loss": 0.3731,
|
| 955 |
+
"step": 2800
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 19.0,
|
| 959 |
+
"eval_accuracy": 0.918562874251497,
|
| 960 |
+
"eval_loss": 0.39593929052352905,
|
| 961 |
+
"eval_runtime": 58.8099,
|
| 962 |
+
"eval_samples_per_second": 14.198,
|
| 963 |
+
"eval_steps_per_second": 0.459,
|
| 964 |
+
"step": 2812
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 19.08783783783784,
|
| 968 |
+
"grad_norm": 91183.2109375,
|
| 969 |
+
"learning_rate": 4.981549815498155e-06,
|
| 970 |
+
"loss": 0.3723,
|
| 971 |
+
"step": 2825
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 19.256756756756758,
|
| 975 |
+
"grad_norm": 80640.0234375,
|
| 976 |
+
"learning_rate": 4.059040590405904e-06,
|
| 977 |
+
"loss": 0.3349,
|
| 978 |
+
"step": 2850
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 19.425675675675677,
|
| 982 |
+
"grad_norm": 89848.125,
|
| 983 |
+
"learning_rate": 3.136531365313653e-06,
|
| 984 |
+
"loss": 0.3679,
|
| 985 |
+
"step": 2875
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 19.594594594594593,
|
| 989 |
+
"grad_norm": 107155.53125,
|
| 990 |
+
"learning_rate": 2.2140221402214023e-06,
|
| 991 |
+
"loss": 0.3463,
|
| 992 |
+
"step": 2900
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 19.763513513513512,
|
| 996 |
+
"grad_norm": 128204.28125,
|
| 997 |
+
"learning_rate": 1.2915129151291513e-06,
|
| 998 |
+
"loss": 0.3549,
|
| 999 |
+
"step": 2925
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 19.93243243243243,
|
| 1003 |
+
"grad_norm": 91720.7734375,
|
| 1004 |
+
"learning_rate": 3.690036900369004e-07,
|
| 1005 |
+
"loss": 0.323,
|
| 1006 |
+
"step": 2950
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 20.0,
|
| 1010 |
+
"eval_accuracy": 0.9209580838323354,
|
| 1011 |
+
"eval_loss": 0.3986169397830963,
|
| 1012 |
+
"eval_runtime": 58.2229,
|
| 1013 |
+
"eval_samples_per_second": 14.341,
|
| 1014 |
+
"eval_steps_per_second": 0.464,
|
| 1015 |
+
"step": 2960
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 20.0,
|
| 1019 |
+
"step": 2960,
|
| 1020 |
+
"total_flos": 0.0,
|
| 1021 |
+
"train_loss": 0.43191270973231344,
|
| 1022 |
+
"train_runtime": 11059.7547,
|
| 1023 |
+
"train_samples_per_second": 8.552,
|
| 1024 |
+
"train_steps_per_second": 0.268
|
| 1025 |
+
}
|
| 1026 |
+
],
|
| 1027 |
+
"logging_steps": 25,
|
| 1028 |
+
"max_steps": 2960,
|
| 1029 |
+
"num_input_tokens_seen": 0,
|
| 1030 |
+
"num_train_epochs": 20,
|
| 1031 |
+
"save_steps": 500,
|
| 1032 |
+
"stateful_callbacks": {
|
| 1033 |
+
"TrainerControl": {
|
| 1034 |
+
"args": {
|
| 1035 |
+
"should_epoch_stop": false,
|
| 1036 |
+
"should_evaluate": false,
|
| 1037 |
+
"should_log": false,
|
| 1038 |
+
"should_save": true,
|
| 1039 |
+
"should_training_stop": true
|
| 1040 |
+
},
|
| 1041 |
+
"attributes": {}
|
| 1042 |
+
}
|
| 1043 |
+
},
|
| 1044 |
+
"total_flos": 0.0,
|
| 1045 |
+
"train_batch_size": 32,
|
| 1046 |
+
"trial_name": null,
|
| 1047 |
+
"trial_params": null
|
| 1048 |
+
}
|