🍻 cheers
Browse files- README.md +5 -2
- all_results.json +13 -0
- eval_results.json +8 -0
- runs/May23_15-41-21_7dfef7dbbba9/events.out.tfevents.1716482773.7dfef7dbbba9.72877.4 +3 -0
- train_results.json +8 -0
- trainer_state.json +2061 -0
README.md
CHANGED
|
@@ -1,5 +1,8 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
| 2 |
tags:
|
|
|
|
| 3 |
- generated_from_trainer
|
| 4 |
datasets:
|
| 5 |
- imagefolder
|
|
@@ -12,7 +15,7 @@ model-index:
|
|
| 12 |
name: Image Classification
|
| 13 |
type: image-classification
|
| 14 |
dataset:
|
| 15 |
-
name:
|
| 16 |
type: imagefolder
|
| 17 |
config: default
|
| 18 |
split: train
|
|
@@ -28,7 +31,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 28 |
|
| 29 |
# vit_epochs5_batch64_lr5e-05_size224_tiles1_seed1_classic_image_classification
|
| 30 |
|
| 31 |
-
This model
|
| 32 |
It achieves the following results on the evaluation set:
|
| 33 |
- Loss: 0.0066
|
| 34 |
- Accuracy: 0.9990
|
|
|
|
| 1 |
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- imagefolder
|
|
|
|
| 15 |
name: Image Classification
|
| 16 |
type: image-classification
|
| 17 |
dataset:
|
| 18 |
+
name: Dogs_vs_Cats
|
| 19 |
type: imagefolder
|
| 20 |
config: default
|
| 21 |
split: train
|
|
|
|
| 31 |
|
| 32 |
# vit_epochs5_batch64_lr5e-05_size224_tiles1_seed1_classic_image_classification
|
| 33 |
|
| 34 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the Dogs_vs_Cats dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
- Loss: 0.0066
|
| 37 |
- Accuracy: 0.9990
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9989655681986109,
|
| 4 |
+
"eval_loss": 0.006617749575525522,
|
| 5 |
+
"eval_runtime": 99.5517,
|
| 6 |
+
"eval_samples_per_second": 67.975,
|
| 7 |
+
"eval_steps_per_second": 1.065,
|
| 8 |
+
"total_flos": 6.98086588443052e+18,
|
| 9 |
+
"train_loss": 0.042465827911299596,
|
| 10 |
+
"train_runtime": 3748.8226,
|
| 11 |
+
"train_samples_per_second": 24.03,
|
| 12 |
+
"train_steps_per_second": 0.376
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9989655681986109,
|
| 4 |
+
"eval_loss": 0.006617749575525522,
|
| 5 |
+
"eval_runtime": 99.5517,
|
| 6 |
+
"eval_samples_per_second": 67.975,
|
| 7 |
+
"eval_steps_per_second": 1.065
|
| 8 |
+
}
|
runs/May23_15-41-21_7dfef7dbbba9/events.out.tfevents.1716482773.7dfef7dbbba9.72877.4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc81315485aa54ba5028dc4ce214429a250b1046eb4eec0d2a07923176474a31
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 6.98086588443052e+18,
|
| 4 |
+
"train_loss": 0.042465827911299596,
|
| 5 |
+
"train_runtime": 3748.8226,
|
| 6 |
+
"train_samples_per_second": 24.03,
|
| 7 |
+
"train_steps_per_second": 0.376
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2061 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.006617749575525522,
|
| 3 |
+
"best_model_checkpoint": "vit_epochs5_batch64_lr5e-05_size224_tiles1_seed1_classic_image_classification/checkpoint-1410",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1410,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01773049645390071,
|
| 13 |
+
"grad_norm": 1.1057369709014893,
|
| 14 |
+
"learning_rate": 4.9822695035461e-05,
|
| 15 |
+
"loss": 0.6106,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03546099290780142,
|
| 20 |
+
"grad_norm": 1.056671142578125,
|
| 21 |
+
"learning_rate": 4.964539007092199e-05,
|
| 22 |
+
"loss": 0.462,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05319148936170213,
|
| 27 |
+
"grad_norm": 0.9491887092590332,
|
| 28 |
+
"learning_rate": 4.946808510638298e-05,
|
| 29 |
+
"loss": 0.31,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07092198581560284,
|
| 34 |
+
"grad_norm": 1.2302058935165405,
|
| 35 |
+
"learning_rate": 4.929078014184397e-05,
|
| 36 |
+
"loss": 0.2197,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08865248226950355,
|
| 41 |
+
"grad_norm": 1.1420009136199951,
|
| 42 |
+
"learning_rate": 4.911347517730497e-05,
|
| 43 |
+
"loss": 0.1468,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10638297872340426,
|
| 48 |
+
"grad_norm": 1.3249993324279785,
|
| 49 |
+
"learning_rate": 4.893617021276596e-05,
|
| 50 |
+
"loss": 0.1098,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12411347517730496,
|
| 55 |
+
"grad_norm": 0.8271305561065674,
|
| 56 |
+
"learning_rate": 4.875886524822695e-05,
|
| 57 |
+
"loss": 0.0994,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14184397163120568,
|
| 62 |
+
"grad_norm": 1.1268686056137085,
|
| 63 |
+
"learning_rate": 4.858156028368794e-05,
|
| 64 |
+
"loss": 0.0718,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1595744680851064,
|
| 69 |
+
"grad_norm": 0.8098769187927246,
|
| 70 |
+
"learning_rate": 4.840425531914894e-05,
|
| 71 |
+
"loss": 0.0861,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1773049645390071,
|
| 76 |
+
"grad_norm": 1.5605119466781616,
|
| 77 |
+
"learning_rate": 4.822695035460993e-05,
|
| 78 |
+
"loss": 0.0882,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1950354609929078,
|
| 83 |
+
"grad_norm": 0.6594899892807007,
|
| 84 |
+
"learning_rate": 4.804964539007092e-05,
|
| 85 |
+
"loss": 0.0567,
|
| 86 |
+
"step": 55
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2127659574468085,
|
| 90 |
+
"grad_norm": 0.3054638206958771,
|
| 91 |
+
"learning_rate": 4.787234042553192e-05,
|
| 92 |
+
"loss": 0.043,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23049645390070922,
|
| 97 |
+
"grad_norm": 1.3071125745773315,
|
| 98 |
+
"learning_rate": 4.769503546099291e-05,
|
| 99 |
+
"loss": 0.0675,
|
| 100 |
+
"step": 65
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24822695035460993,
|
| 104 |
+
"grad_norm": 1.4442698955535889,
|
| 105 |
+
"learning_rate": 4.751773049645391e-05,
|
| 106 |
+
"loss": 0.0818,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.26595744680851063,
|
| 111 |
+
"grad_norm": 1.443562388420105,
|
| 112 |
+
"learning_rate": 4.734042553191489e-05,
|
| 113 |
+
"loss": 0.09,
|
| 114 |
+
"step": 75
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28368794326241137,
|
| 118 |
+
"grad_norm": 1.0713045597076416,
|
| 119 |
+
"learning_rate": 4.716312056737589e-05,
|
| 120 |
+
"loss": 0.0766,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.30141843971631205,
|
| 125 |
+
"grad_norm": 0.24775153398513794,
|
| 126 |
+
"learning_rate": 4.698581560283688e-05,
|
| 127 |
+
"loss": 0.067,
|
| 128 |
+
"step": 85
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3191489361702128,
|
| 132 |
+
"grad_norm": 1.436522126197815,
|
| 133 |
+
"learning_rate": 4.680851063829788e-05,
|
| 134 |
+
"loss": 0.0792,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.33687943262411346,
|
| 139 |
+
"grad_norm": 0.49839749932289124,
|
| 140 |
+
"learning_rate": 4.663120567375887e-05,
|
| 141 |
+
"loss": 0.0595,
|
| 142 |
+
"step": 95
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3546099290780142,
|
| 146 |
+
"grad_norm": 0.6618617177009583,
|
| 147 |
+
"learning_rate": 4.645390070921986e-05,
|
| 148 |
+
"loss": 0.036,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3723404255319149,
|
| 153 |
+
"grad_norm": 0.8074954748153687,
|
| 154 |
+
"learning_rate": 4.627659574468085e-05,
|
| 155 |
+
"loss": 0.0589,
|
| 156 |
+
"step": 105
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3900709219858156,
|
| 160 |
+
"grad_norm": 1.512774109840393,
|
| 161 |
+
"learning_rate": 4.609929078014185e-05,
|
| 162 |
+
"loss": 0.0477,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4078014184397163,
|
| 167 |
+
"grad_norm": 1.3695027828216553,
|
| 168 |
+
"learning_rate": 4.592198581560284e-05,
|
| 169 |
+
"loss": 0.1018,
|
| 170 |
+
"step": 115
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.425531914893617,
|
| 174 |
+
"grad_norm": 0.7358906865119934,
|
| 175 |
+
"learning_rate": 4.574468085106383e-05,
|
| 176 |
+
"loss": 0.0689,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4432624113475177,
|
| 181 |
+
"grad_norm": 0.2363295555114746,
|
| 182 |
+
"learning_rate": 4.556737588652483e-05,
|
| 183 |
+
"loss": 0.0644,
|
| 184 |
+
"step": 125
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.46099290780141844,
|
| 188 |
+
"grad_norm": 2.4520492553710938,
|
| 189 |
+
"learning_rate": 4.539007092198582e-05,
|
| 190 |
+
"loss": 0.0489,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4787234042553192,
|
| 195 |
+
"grad_norm": 1.8659327030181885,
|
| 196 |
+
"learning_rate": 4.5212765957446815e-05,
|
| 197 |
+
"loss": 0.0544,
|
| 198 |
+
"step": 135
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.49645390070921985,
|
| 202 |
+
"grad_norm": 0.5137922167778015,
|
| 203 |
+
"learning_rate": 4.50354609929078e-05,
|
| 204 |
+
"loss": 0.0733,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5141843971631206,
|
| 209 |
+
"grad_norm": 0.2559826076030731,
|
| 210 |
+
"learning_rate": 4.48581560283688e-05,
|
| 211 |
+
"loss": 0.0349,
|
| 212 |
+
"step": 145
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5319148936170213,
|
| 216 |
+
"grad_norm": 1.2222139835357666,
|
| 217 |
+
"learning_rate": 4.468085106382979e-05,
|
| 218 |
+
"loss": 0.061,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.549645390070922,
|
| 223 |
+
"grad_norm": 0.19817987084388733,
|
| 224 |
+
"learning_rate": 4.4503546099290785e-05,
|
| 225 |
+
"loss": 0.0548,
|
| 226 |
+
"step": 155
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5673758865248227,
|
| 230 |
+
"grad_norm": 0.6621349453926086,
|
| 231 |
+
"learning_rate": 4.432624113475177e-05,
|
| 232 |
+
"loss": 0.0409,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5851063829787234,
|
| 237 |
+
"grad_norm": 0.8852491974830627,
|
| 238 |
+
"learning_rate": 4.414893617021277e-05,
|
| 239 |
+
"loss": 0.027,
|
| 240 |
+
"step": 165
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6028368794326241,
|
| 244 |
+
"grad_norm": 0.1130673736333847,
|
| 245 |
+
"learning_rate": 4.3971631205673764e-05,
|
| 246 |
+
"loss": 0.0271,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6205673758865248,
|
| 251 |
+
"grad_norm": 2.4764528274536133,
|
| 252 |
+
"learning_rate": 4.3794326241134755e-05,
|
| 253 |
+
"loss": 0.0578,
|
| 254 |
+
"step": 175
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6382978723404256,
|
| 258 |
+
"grad_norm": 1.4849659204483032,
|
| 259 |
+
"learning_rate": 4.3617021276595746e-05,
|
| 260 |
+
"loss": 0.0425,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6560283687943262,
|
| 265 |
+
"grad_norm": 0.5148060321807861,
|
| 266 |
+
"learning_rate": 4.343971631205674e-05,
|
| 267 |
+
"loss": 0.036,
|
| 268 |
+
"step": 185
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6737588652482269,
|
| 272 |
+
"grad_norm": 1.4764983654022217,
|
| 273 |
+
"learning_rate": 4.3262411347517734e-05,
|
| 274 |
+
"loss": 0.0815,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6914893617021277,
|
| 279 |
+
"grad_norm": 0.46336182951927185,
|
| 280 |
+
"learning_rate": 4.3085106382978725e-05,
|
| 281 |
+
"loss": 0.0225,
|
| 282 |
+
"step": 195
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7092198581560284,
|
| 286 |
+
"grad_norm": 0.6195293664932251,
|
| 287 |
+
"learning_rate": 4.2907801418439716e-05,
|
| 288 |
+
"loss": 0.0228,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7269503546099291,
|
| 293 |
+
"grad_norm": 0.27533644437789917,
|
| 294 |
+
"learning_rate": 4.273049645390071e-05,
|
| 295 |
+
"loss": 0.0709,
|
| 296 |
+
"step": 205
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7446808510638298,
|
| 300 |
+
"grad_norm": 0.6618344187736511,
|
| 301 |
+
"learning_rate": 4.2553191489361704e-05,
|
| 302 |
+
"loss": 0.0708,
|
| 303 |
+
"step": 210
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7624113475177305,
|
| 307 |
+
"grad_norm": 0.9032989740371704,
|
| 308 |
+
"learning_rate": 4.2375886524822695e-05,
|
| 309 |
+
"loss": 0.0396,
|
| 310 |
+
"step": 215
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7801418439716312,
|
| 314 |
+
"grad_norm": 2.292132616043091,
|
| 315 |
+
"learning_rate": 4.219858156028369e-05,
|
| 316 |
+
"loss": 0.0927,
|
| 317 |
+
"step": 220
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7978723404255319,
|
| 321 |
+
"grad_norm": 2.1669278144836426,
|
| 322 |
+
"learning_rate": 4.2021276595744684e-05,
|
| 323 |
+
"loss": 0.1068,
|
| 324 |
+
"step": 225
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8156028368794326,
|
| 328 |
+
"grad_norm": 1.218011498451233,
|
| 329 |
+
"learning_rate": 4.1843971631205674e-05,
|
| 330 |
+
"loss": 0.0444,
|
| 331 |
+
"step": 230
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8333333333333334,
|
| 335 |
+
"grad_norm": 1.6065305471420288,
|
| 336 |
+
"learning_rate": 4.166666666666667e-05,
|
| 337 |
+
"loss": 0.0443,
|
| 338 |
+
"step": 235
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.851063829787234,
|
| 342 |
+
"grad_norm": 1.1235682964324951,
|
| 343 |
+
"learning_rate": 4.148936170212766e-05,
|
| 344 |
+
"loss": 0.0683,
|
| 345 |
+
"step": 240
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8687943262411347,
|
| 349 |
+
"grad_norm": 1.7434436082839966,
|
| 350 |
+
"learning_rate": 4.1312056737588654e-05,
|
| 351 |
+
"loss": 0.0857,
|
| 352 |
+
"step": 245
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8865248226950354,
|
| 356 |
+
"grad_norm": 0.5483965277671814,
|
| 357 |
+
"learning_rate": 4.1134751773049644e-05,
|
| 358 |
+
"loss": 0.0437,
|
| 359 |
+
"step": 250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9042553191489362,
|
| 363 |
+
"grad_norm": 1.176393985748291,
|
| 364 |
+
"learning_rate": 4.095744680851064e-05,
|
| 365 |
+
"loss": 0.0771,
|
| 366 |
+
"step": 255
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9219858156028369,
|
| 370 |
+
"grad_norm": 1.270770788192749,
|
| 371 |
+
"learning_rate": 4.078014184397163e-05,
|
| 372 |
+
"loss": 0.0371,
|
| 373 |
+
"step": 260
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9397163120567376,
|
| 377 |
+
"grad_norm": 1.3304009437561035,
|
| 378 |
+
"learning_rate": 4.0602836879432624e-05,
|
| 379 |
+
"loss": 0.0815,
|
| 380 |
+
"step": 265
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9574468085106383,
|
| 384 |
+
"grad_norm": 0.5169529914855957,
|
| 385 |
+
"learning_rate": 4.0425531914893614e-05,
|
| 386 |
+
"loss": 0.0511,
|
| 387 |
+
"step": 270
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.975177304964539,
|
| 391 |
+
"grad_norm": 0.4900430142879486,
|
| 392 |
+
"learning_rate": 4.024822695035461e-05,
|
| 393 |
+
"loss": 0.0304,
|
| 394 |
+
"step": 275
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.9929078014184397,
|
| 398 |
+
"grad_norm": 1.8383599519729614,
|
| 399 |
+
"learning_rate": 4.007092198581561e-05,
|
| 400 |
+
"loss": 0.0449,
|
| 401 |
+
"step": 280
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0,
|
| 405 |
+
"eval_accuracy": 0.9957144968228167,
|
| 406 |
+
"eval_loss": 0.018303971737623215,
|
| 407 |
+
"eval_runtime": 99.5406,
|
| 408 |
+
"eval_samples_per_second": 67.982,
|
| 409 |
+
"eval_steps_per_second": 1.065,
|
| 410 |
+
"step": 282
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.0106382978723405,
|
| 414 |
+
"grad_norm": 0.5246375203132629,
|
| 415 |
+
"learning_rate": 3.9893617021276594e-05,
|
| 416 |
+
"loss": 0.0246,
|
| 417 |
+
"step": 285
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.0283687943262412,
|
| 421 |
+
"grad_norm": 1.0449841022491455,
|
| 422 |
+
"learning_rate": 3.971631205673759e-05,
|
| 423 |
+
"loss": 0.0212,
|
| 424 |
+
"step": 290
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.0460992907801419,
|
| 428 |
+
"grad_norm": 0.26453474164009094,
|
| 429 |
+
"learning_rate": 3.953900709219858e-05,
|
| 430 |
+
"loss": 0.0358,
|
| 431 |
+
"step": 295
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.0638297872340425,
|
| 435 |
+
"grad_norm": 0.6182503700256348,
|
| 436 |
+
"learning_rate": 3.936170212765958e-05,
|
| 437 |
+
"loss": 0.0387,
|
| 438 |
+
"step": 300
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.0815602836879432,
|
| 442 |
+
"grad_norm": 0.46771693229675293,
|
| 443 |
+
"learning_rate": 3.918439716312057e-05,
|
| 444 |
+
"loss": 0.0285,
|
| 445 |
+
"step": 305
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.099290780141844,
|
| 449 |
+
"grad_norm": 1.8450125455856323,
|
| 450 |
+
"learning_rate": 3.900709219858156e-05,
|
| 451 |
+
"loss": 0.0475,
|
| 452 |
+
"step": 310
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.1170212765957448,
|
| 456 |
+
"grad_norm": 0.7594661712646484,
|
| 457 |
+
"learning_rate": 3.882978723404255e-05,
|
| 458 |
+
"loss": 0.0156,
|
| 459 |
+
"step": 315
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.1347517730496455,
|
| 463 |
+
"grad_norm": 0.29846104979515076,
|
| 464 |
+
"learning_rate": 3.865248226950355e-05,
|
| 465 |
+
"loss": 0.0375,
|
| 466 |
+
"step": 320
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.1524822695035462,
|
| 470 |
+
"grad_norm": 1.215070128440857,
|
| 471 |
+
"learning_rate": 3.847517730496454e-05,
|
| 472 |
+
"loss": 0.0626,
|
| 473 |
+
"step": 325
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.1702127659574468,
|
| 477 |
+
"grad_norm": 0.27688533067703247,
|
| 478 |
+
"learning_rate": 3.829787234042553e-05,
|
| 479 |
+
"loss": 0.0327,
|
| 480 |
+
"step": 330
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.1879432624113475,
|
| 484 |
+
"grad_norm": 1.9927847385406494,
|
| 485 |
+
"learning_rate": 3.812056737588653e-05,
|
| 486 |
+
"loss": 0.0557,
|
| 487 |
+
"step": 335
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.2056737588652482,
|
| 491 |
+
"grad_norm": 4.12082576751709,
|
| 492 |
+
"learning_rate": 3.794326241134752e-05,
|
| 493 |
+
"loss": 0.0669,
|
| 494 |
+
"step": 340
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.2234042553191489,
|
| 498 |
+
"grad_norm": 1.2266093492507935,
|
| 499 |
+
"learning_rate": 3.776595744680852e-05,
|
| 500 |
+
"loss": 0.0575,
|
| 501 |
+
"step": 345
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.2411347517730495,
|
| 505 |
+
"grad_norm": 0.7239733934402466,
|
| 506 |
+
"learning_rate": 3.75886524822695e-05,
|
| 507 |
+
"loss": 0.0316,
|
| 508 |
+
"step": 350
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.2588652482269502,
|
| 512 |
+
"grad_norm": 1.3613947629928589,
|
| 513 |
+
"learning_rate": 3.74113475177305e-05,
|
| 514 |
+
"loss": 0.0512,
|
| 515 |
+
"step": 355
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.2765957446808511,
|
| 519 |
+
"grad_norm": 2.1938693523406982,
|
| 520 |
+
"learning_rate": 3.723404255319149e-05,
|
| 521 |
+
"loss": 0.0511,
|
| 522 |
+
"step": 360
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.2943262411347518,
|
| 526 |
+
"grad_norm": 0.5384643077850342,
|
| 527 |
+
"learning_rate": 3.705673758865249e-05,
|
| 528 |
+
"loss": 0.0479,
|
| 529 |
+
"step": 365
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.3120567375886525,
|
| 533 |
+
"grad_norm": 0.4367295801639557,
|
| 534 |
+
"learning_rate": 3.687943262411347e-05,
|
| 535 |
+
"loss": 0.0595,
|
| 536 |
+
"step": 370
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.3297872340425532,
|
| 540 |
+
"grad_norm": 1.2368284463882446,
|
| 541 |
+
"learning_rate": 3.670212765957447e-05,
|
| 542 |
+
"loss": 0.0332,
|
| 543 |
+
"step": 375
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.3475177304964538,
|
| 547 |
+
"grad_norm": 1.422599196434021,
|
| 548 |
+
"learning_rate": 3.6524822695035466e-05,
|
| 549 |
+
"loss": 0.0433,
|
| 550 |
+
"step": 380
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.3652482269503547,
|
| 554 |
+
"grad_norm": 1.9355744123458862,
|
| 555 |
+
"learning_rate": 3.634751773049646e-05,
|
| 556 |
+
"loss": 0.0401,
|
| 557 |
+
"step": 385
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.3829787234042552,
|
| 561 |
+
"grad_norm": 0.11844053119421005,
|
| 562 |
+
"learning_rate": 3.617021276595745e-05,
|
| 563 |
+
"loss": 0.0308,
|
| 564 |
+
"step": 390
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.400709219858156,
|
| 568 |
+
"grad_norm": 1.5698013305664062,
|
| 569 |
+
"learning_rate": 3.599290780141844e-05,
|
| 570 |
+
"loss": 0.024,
|
| 571 |
+
"step": 395
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.4184397163120568,
|
| 575 |
+
"grad_norm": 0.24424462020397186,
|
| 576 |
+
"learning_rate": 3.5815602836879437e-05,
|
| 577 |
+
"loss": 0.0284,
|
| 578 |
+
"step": 400
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.4361702127659575,
|
| 582 |
+
"grad_norm": 0.8079735040664673,
|
| 583 |
+
"learning_rate": 3.563829787234043e-05,
|
| 584 |
+
"loss": 0.0289,
|
| 585 |
+
"step": 405
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.4539007092198581,
|
| 589 |
+
"grad_norm": 0.5713645815849304,
|
| 590 |
+
"learning_rate": 3.546099290780142e-05,
|
| 591 |
+
"loss": 0.0489,
|
| 592 |
+
"step": 410
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.4716312056737588,
|
| 596 |
+
"grad_norm": 1.418959140777588,
|
| 597 |
+
"learning_rate": 3.528368794326241e-05,
|
| 598 |
+
"loss": 0.0494,
|
| 599 |
+
"step": 415
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.4893617021276595,
|
| 603 |
+
"grad_norm": 1.3852524757385254,
|
| 604 |
+
"learning_rate": 3.5106382978723407e-05,
|
| 605 |
+
"loss": 0.0416,
|
| 606 |
+
"step": 420
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.5070921985815602,
|
| 610 |
+
"grad_norm": 1.9135431051254272,
|
| 611 |
+
"learning_rate": 3.49290780141844e-05,
|
| 612 |
+
"loss": 0.0244,
|
| 613 |
+
"step": 425
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.524822695035461,
|
| 617 |
+
"grad_norm": 0.7146717309951782,
|
| 618 |
+
"learning_rate": 3.4751773049645395e-05,
|
| 619 |
+
"loss": 0.0512,
|
| 620 |
+
"step": 430
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.5425531914893615,
|
| 624 |
+
"grad_norm": 0.1944599747657776,
|
| 625 |
+
"learning_rate": 3.4574468085106386e-05,
|
| 626 |
+
"loss": 0.0191,
|
| 627 |
+
"step": 435
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.5602836879432624,
|
| 631 |
+
"grad_norm": 0.32298266887664795,
|
| 632 |
+
"learning_rate": 3.4397163120567377e-05,
|
| 633 |
+
"loss": 0.0221,
|
| 634 |
+
"step": 440
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.5780141843971631,
|
| 638 |
+
"grad_norm": 0.7677968740463257,
|
| 639 |
+
"learning_rate": 3.4219858156028374e-05,
|
| 640 |
+
"loss": 0.023,
|
| 641 |
+
"step": 445
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.5957446808510638,
|
| 645 |
+
"grad_norm": 0.19664119184017181,
|
| 646 |
+
"learning_rate": 3.4042553191489365e-05,
|
| 647 |
+
"loss": 0.0402,
|
| 648 |
+
"step": 450
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.6134751773049647,
|
| 652 |
+
"grad_norm": 0.8602023124694824,
|
| 653 |
+
"learning_rate": 3.3865248226950356e-05,
|
| 654 |
+
"loss": 0.0343,
|
| 655 |
+
"step": 455
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.6312056737588652,
|
| 659 |
+
"grad_norm": 2.0019989013671875,
|
| 660 |
+
"learning_rate": 3.3687943262411347e-05,
|
| 661 |
+
"loss": 0.0521,
|
| 662 |
+
"step": 460
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.648936170212766,
|
| 666 |
+
"grad_norm": 0.42928043007850647,
|
| 667 |
+
"learning_rate": 3.3510638297872344e-05,
|
| 668 |
+
"loss": 0.0434,
|
| 669 |
+
"step": 465
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.6666666666666665,
|
| 673 |
+
"grad_norm": 0.31418949365615845,
|
| 674 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 675 |
+
"loss": 0.0125,
|
| 676 |
+
"step": 470
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.6843971631205674,
|
| 680 |
+
"grad_norm": 0.6069945096969604,
|
| 681 |
+
"learning_rate": 3.3156028368794326e-05,
|
| 682 |
+
"loss": 0.055,
|
| 683 |
+
"step": 475
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.702127659574468,
|
| 687 |
+
"grad_norm": 2.3644161224365234,
|
| 688 |
+
"learning_rate": 3.2978723404255317e-05,
|
| 689 |
+
"loss": 0.0652,
|
| 690 |
+
"step": 480
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.7198581560283688,
|
| 694 |
+
"grad_norm": 0.054500628262758255,
|
| 695 |
+
"learning_rate": 3.2801418439716314e-05,
|
| 696 |
+
"loss": 0.0398,
|
| 697 |
+
"step": 485
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.7375886524822695,
|
| 701 |
+
"grad_norm": 2.016355037689209,
|
| 702 |
+
"learning_rate": 3.262411347517731e-05,
|
| 703 |
+
"loss": 0.0788,
|
| 704 |
+
"step": 490
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.7553191489361701,
|
| 708 |
+
"grad_norm": 0.6187741160392761,
|
| 709 |
+
"learning_rate": 3.2446808510638296e-05,
|
| 710 |
+
"loss": 0.0579,
|
| 711 |
+
"step": 495
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.773049645390071,
|
| 715 |
+
"grad_norm": 1.820343255996704,
|
| 716 |
+
"learning_rate": 3.226950354609929e-05,
|
| 717 |
+
"loss": 0.0632,
|
| 718 |
+
"step": 500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.7907801418439715,
|
| 722 |
+
"grad_norm": 0.4283825755119324,
|
| 723 |
+
"learning_rate": 3.2092198581560284e-05,
|
| 724 |
+
"loss": 0.0449,
|
| 725 |
+
"step": 505
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.8085106382978724,
|
| 729 |
+
"grad_norm": 1.2107046842575073,
|
| 730 |
+
"learning_rate": 3.191489361702128e-05,
|
| 731 |
+
"loss": 0.0293,
|
| 732 |
+
"step": 510
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.826241134751773,
|
| 736 |
+
"grad_norm": 0.751215398311615,
|
| 737 |
+
"learning_rate": 3.173758865248227e-05,
|
| 738 |
+
"loss": 0.0436,
|
| 739 |
+
"step": 515
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.8439716312056738,
|
| 743 |
+
"grad_norm": 0.8796170353889465,
|
| 744 |
+
"learning_rate": 3.156028368794326e-05,
|
| 745 |
+
"loss": 0.052,
|
| 746 |
+
"step": 520
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.8617021276595744,
|
| 750 |
+
"grad_norm": 0.14203386008739471,
|
| 751 |
+
"learning_rate": 3.1382978723404254e-05,
|
| 752 |
+
"loss": 0.0103,
|
| 753 |
+
"step": 525
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.8794326241134751,
|
| 757 |
+
"grad_norm": 1.044580340385437,
|
| 758 |
+
"learning_rate": 3.120567375886525e-05,
|
| 759 |
+
"loss": 0.0269,
|
| 760 |
+
"step": 530
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.897163120567376,
|
| 764 |
+
"grad_norm": 2.8254494667053223,
|
| 765 |
+
"learning_rate": 3.102836879432624e-05,
|
| 766 |
+
"loss": 0.0662,
|
| 767 |
+
"step": 535
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.9148936170212765,
|
| 771 |
+
"grad_norm": 0.6130526065826416,
|
| 772 |
+
"learning_rate": 3.085106382978723e-05,
|
| 773 |
+
"loss": 0.0277,
|
| 774 |
+
"step": 540
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.9326241134751774,
|
| 778 |
+
"grad_norm": 0.12123987823724747,
|
| 779 |
+
"learning_rate": 3.067375886524823e-05,
|
| 780 |
+
"loss": 0.035,
|
| 781 |
+
"step": 545
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.950354609929078,
|
| 785 |
+
"grad_norm": 0.5822201371192932,
|
| 786 |
+
"learning_rate": 3.0496453900709222e-05,
|
| 787 |
+
"loss": 0.0294,
|
| 788 |
+
"step": 550
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.9680851063829787,
|
| 792 |
+
"grad_norm": 0.28792017698287964,
|
| 793 |
+
"learning_rate": 3.0319148936170216e-05,
|
| 794 |
+
"loss": 0.0543,
|
| 795 |
+
"step": 555
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.9858156028368794,
|
| 799 |
+
"grad_norm": 1.7587871551513672,
|
| 800 |
+
"learning_rate": 3.0141843971631207e-05,
|
| 801 |
+
"loss": 0.04,
|
| 802 |
+
"step": 560
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.0,
|
| 806 |
+
"eval_accuracy": 0.9980789123688488,
|
| 807 |
+
"eval_loss": 0.010071107186377048,
|
| 808 |
+
"eval_runtime": 99.7696,
|
| 809 |
+
"eval_samples_per_second": 67.826,
|
| 810 |
+
"eval_steps_per_second": 1.062,
|
| 811 |
+
"step": 564
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.00354609929078,
|
| 815 |
+
"grad_norm": 0.3479292690753937,
|
| 816 |
+
"learning_rate": 2.99645390070922e-05,
|
| 817 |
+
"loss": 0.0726,
|
| 818 |
+
"step": 565
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.021276595744681,
|
| 822 |
+
"grad_norm": 0.10121797770261765,
|
| 823 |
+
"learning_rate": 2.9787234042553192e-05,
|
| 824 |
+
"loss": 0.0185,
|
| 825 |
+
"step": 570
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.0390070921985815,
|
| 829 |
+
"grad_norm": 0.9174487590789795,
|
| 830 |
+
"learning_rate": 2.9609929078014186e-05,
|
| 831 |
+
"loss": 0.0637,
|
| 832 |
+
"step": 575
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.0567375886524824,
|
| 836 |
+
"grad_norm": 0.4720062017440796,
|
| 837 |
+
"learning_rate": 2.9432624113475177e-05,
|
| 838 |
+
"loss": 0.0246,
|
| 839 |
+
"step": 580
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.074468085106383,
|
| 843 |
+
"grad_norm": 0.3442896008491516,
|
| 844 |
+
"learning_rate": 2.925531914893617e-05,
|
| 845 |
+
"loss": 0.0159,
|
| 846 |
+
"step": 585
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.0921985815602837,
|
| 850 |
+
"grad_norm": 1.1636303663253784,
|
| 851 |
+
"learning_rate": 2.9078014184397162e-05,
|
| 852 |
+
"loss": 0.018,
|
| 853 |
+
"step": 590
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.1099290780141846,
|
| 857 |
+
"grad_norm": 0.7616550326347351,
|
| 858 |
+
"learning_rate": 2.8900709219858156e-05,
|
| 859 |
+
"loss": 0.0392,
|
| 860 |
+
"step": 595
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.127659574468085,
|
| 864 |
+
"grad_norm": 0.17084839940071106,
|
| 865 |
+
"learning_rate": 2.8723404255319154e-05,
|
| 866 |
+
"loss": 0.0172,
|
| 867 |
+
"step": 600
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.145390070921986,
|
| 871 |
+
"grad_norm": 0.4418025612831116,
|
| 872 |
+
"learning_rate": 2.8546099290780144e-05,
|
| 873 |
+
"loss": 0.0287,
|
| 874 |
+
"step": 605
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.1631205673758864,
|
| 878 |
+
"grad_norm": 0.05007610097527504,
|
| 879 |
+
"learning_rate": 2.836879432624114e-05,
|
| 880 |
+
"loss": 0.0311,
|
| 881 |
+
"step": 610
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.1808510638297873,
|
| 885 |
+
"grad_norm": 2.390089273452759,
|
| 886 |
+
"learning_rate": 2.819148936170213e-05,
|
| 887 |
+
"loss": 0.0393,
|
| 888 |
+
"step": 615
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.198581560283688,
|
| 892 |
+
"grad_norm": 1.8016951084136963,
|
| 893 |
+
"learning_rate": 2.8014184397163124e-05,
|
| 894 |
+
"loss": 0.0518,
|
| 895 |
+
"step": 620
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.2163120567375887,
|
| 899 |
+
"grad_norm": 0.39676979184150696,
|
| 900 |
+
"learning_rate": 2.7836879432624114e-05,
|
| 901 |
+
"loss": 0.0319,
|
| 902 |
+
"step": 625
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.2340425531914896,
|
| 906 |
+
"grad_norm": 0.2608576714992523,
|
| 907 |
+
"learning_rate": 2.765957446808511e-05,
|
| 908 |
+
"loss": 0.0148,
|
| 909 |
+
"step": 630
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.25177304964539,
|
| 913 |
+
"grad_norm": 1.5164140462875366,
|
| 914 |
+
"learning_rate": 2.74822695035461e-05,
|
| 915 |
+
"loss": 0.04,
|
| 916 |
+
"step": 635
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.269503546099291,
|
| 920 |
+
"grad_norm": 0.3103680908679962,
|
| 921 |
+
"learning_rate": 2.7304964539007094e-05,
|
| 922 |
+
"loss": 0.0234,
|
| 923 |
+
"step": 640
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.2872340425531914,
|
| 927 |
+
"grad_norm": 0.4016372561454773,
|
| 928 |
+
"learning_rate": 2.7127659574468084e-05,
|
| 929 |
+
"loss": 0.0171,
|
| 930 |
+
"step": 645
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.3049645390070923,
|
| 934 |
+
"grad_norm": 0.5997406840324402,
|
| 935 |
+
"learning_rate": 2.695035460992908e-05,
|
| 936 |
+
"loss": 0.0387,
|
| 937 |
+
"step": 650
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.3226950354609928,
|
| 941 |
+
"grad_norm": 0.8444051742553711,
|
| 942 |
+
"learning_rate": 2.6773049645390076e-05,
|
| 943 |
+
"loss": 0.0539,
|
| 944 |
+
"step": 655
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.3404255319148937,
|
| 948 |
+
"grad_norm": 0.108007051050663,
|
| 949 |
+
"learning_rate": 2.6595744680851064e-05,
|
| 950 |
+
"loss": 0.0315,
|
| 951 |
+
"step": 660
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.3581560283687946,
|
| 955 |
+
"grad_norm": 0.21093517541885376,
|
| 956 |
+
"learning_rate": 2.641843971631206e-05,
|
| 957 |
+
"loss": 0.0175,
|
| 958 |
+
"step": 665
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.375886524822695,
|
| 962 |
+
"grad_norm": 0.4845324754714966,
|
| 963 |
+
"learning_rate": 2.624113475177305e-05,
|
| 964 |
+
"loss": 0.0185,
|
| 965 |
+
"step": 670
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.393617021276596,
|
| 969 |
+
"grad_norm": 2.6781203746795654,
|
| 970 |
+
"learning_rate": 2.6063829787234046e-05,
|
| 971 |
+
"loss": 0.0562,
|
| 972 |
+
"step": 675
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.4113475177304964,
|
| 976 |
+
"grad_norm": 0.301878958940506,
|
| 977 |
+
"learning_rate": 2.5886524822695034e-05,
|
| 978 |
+
"loss": 0.0298,
|
| 979 |
+
"step": 680
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.4290780141843973,
|
| 983 |
+
"grad_norm": 0.9099262952804565,
|
| 984 |
+
"learning_rate": 2.570921985815603e-05,
|
| 985 |
+
"loss": 0.0421,
|
| 986 |
+
"step": 685
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.4468085106382977,
|
| 990 |
+
"grad_norm": 0.11847200244665146,
|
| 991 |
+
"learning_rate": 2.5531914893617022e-05,
|
| 992 |
+
"loss": 0.0174,
|
| 993 |
+
"step": 690
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.4645390070921986,
|
| 997 |
+
"grad_norm": 0.12315840274095535,
|
| 998 |
+
"learning_rate": 2.5354609929078016e-05,
|
| 999 |
+
"loss": 0.0129,
|
| 1000 |
+
"step": 695
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.482269503546099,
|
| 1004 |
+
"grad_norm": 0.955663800239563,
|
| 1005 |
+
"learning_rate": 2.5177304964539007e-05,
|
| 1006 |
+
"loss": 0.0239,
|
| 1007 |
+
"step": 700
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.5,
|
| 1011 |
+
"grad_norm": 0.13527634739875793,
|
| 1012 |
+
"learning_rate": 2.5e-05,
|
| 1013 |
+
"loss": 0.0217,
|
| 1014 |
+
"step": 705
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.5177304964539005,
|
| 1018 |
+
"grad_norm": 0.24646982550621033,
|
| 1019 |
+
"learning_rate": 2.4822695035460995e-05,
|
| 1020 |
+
"loss": 0.0388,
|
| 1021 |
+
"step": 710
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.5354609929078014,
|
| 1025 |
+
"grad_norm": 0.09928718209266663,
|
| 1026 |
+
"learning_rate": 2.4645390070921986e-05,
|
| 1027 |
+
"loss": 0.0117,
|
| 1028 |
+
"step": 715
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.5531914893617023,
|
| 1032 |
+
"grad_norm": 2.1757736206054688,
|
| 1033 |
+
"learning_rate": 2.446808510638298e-05,
|
| 1034 |
+
"loss": 0.0205,
|
| 1035 |
+
"step": 720
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.5709219858156027,
|
| 1039 |
+
"grad_norm": 0.4307613968849182,
|
| 1040 |
+
"learning_rate": 2.429078014184397e-05,
|
| 1041 |
+
"loss": 0.0495,
|
| 1042 |
+
"step": 725
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.5886524822695036,
|
| 1046 |
+
"grad_norm": 0.07029764354228973,
|
| 1047 |
+
"learning_rate": 2.4113475177304965e-05,
|
| 1048 |
+
"loss": 0.0568,
|
| 1049 |
+
"step": 730
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.6063829787234045,
|
| 1053 |
+
"grad_norm": 1.3018437623977661,
|
| 1054 |
+
"learning_rate": 2.393617021276596e-05,
|
| 1055 |
+
"loss": 0.0389,
|
| 1056 |
+
"step": 735
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.624113475177305,
|
| 1060 |
+
"grad_norm": 0.2572399973869324,
|
| 1061 |
+
"learning_rate": 2.3758865248226954e-05,
|
| 1062 |
+
"loss": 0.0392,
|
| 1063 |
+
"step": 740
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.6418439716312054,
|
| 1067 |
+
"grad_norm": 1.5049391984939575,
|
| 1068 |
+
"learning_rate": 2.3581560283687945e-05,
|
| 1069 |
+
"loss": 0.0273,
|
| 1070 |
+
"step": 745
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.6595744680851063,
|
| 1074 |
+
"grad_norm": 1.8658090829849243,
|
| 1075 |
+
"learning_rate": 2.340425531914894e-05,
|
| 1076 |
+
"loss": 0.0503,
|
| 1077 |
+
"step": 750
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.6773049645390072,
|
| 1081 |
+
"grad_norm": 1.4984087944030762,
|
| 1082 |
+
"learning_rate": 2.322695035460993e-05,
|
| 1083 |
+
"loss": 0.0256,
|
| 1084 |
+
"step": 755
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.6950354609929077,
|
| 1088 |
+
"grad_norm": 1.7155365943908691,
|
| 1089 |
+
"learning_rate": 2.3049645390070924e-05,
|
| 1090 |
+
"loss": 0.0371,
|
| 1091 |
+
"step": 760
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.7127659574468086,
|
| 1095 |
+
"grad_norm": 2.706315517425537,
|
| 1096 |
+
"learning_rate": 2.2872340425531915e-05,
|
| 1097 |
+
"loss": 0.0289,
|
| 1098 |
+
"step": 765
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.7304964539007095,
|
| 1102 |
+
"grad_norm": 0.5071002244949341,
|
| 1103 |
+
"learning_rate": 2.269503546099291e-05,
|
| 1104 |
+
"loss": 0.0144,
|
| 1105 |
+
"step": 770
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.74822695035461,
|
| 1109 |
+
"grad_norm": 2.5877788066864014,
|
| 1110 |
+
"learning_rate": 2.25177304964539e-05,
|
| 1111 |
+
"loss": 0.0462,
|
| 1112 |
+
"step": 775
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.7659574468085104,
|
| 1116 |
+
"grad_norm": 0.13342903554439545,
|
| 1117 |
+
"learning_rate": 2.2340425531914894e-05,
|
| 1118 |
+
"loss": 0.0311,
|
| 1119 |
+
"step": 780
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.7836879432624113,
|
| 1123 |
+
"grad_norm": 0.5681129693984985,
|
| 1124 |
+
"learning_rate": 2.2163120567375885e-05,
|
| 1125 |
+
"loss": 0.0496,
|
| 1126 |
+
"step": 785
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.801418439716312,
|
| 1130 |
+
"grad_norm": 1.3924553394317627,
|
| 1131 |
+
"learning_rate": 2.1985815602836882e-05,
|
| 1132 |
+
"loss": 0.0302,
|
| 1133 |
+
"step": 790
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.8191489361702127,
|
| 1137 |
+
"grad_norm": 0.9623014330863953,
|
| 1138 |
+
"learning_rate": 2.1808510638297873e-05,
|
| 1139 |
+
"loss": 0.0166,
|
| 1140 |
+
"step": 795
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.8368794326241136,
|
| 1144 |
+
"grad_norm": 1.2445993423461914,
|
| 1145 |
+
"learning_rate": 2.1631205673758867e-05,
|
| 1146 |
+
"loss": 0.0237,
|
| 1147 |
+
"step": 800
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.854609929078014,
|
| 1151 |
+
"grad_norm": 0.8575385212898254,
|
| 1152 |
+
"learning_rate": 2.1453900709219858e-05,
|
| 1153 |
+
"loss": 0.0753,
|
| 1154 |
+
"step": 805
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.872340425531915,
|
| 1158 |
+
"grad_norm": 1.026553988456726,
|
| 1159 |
+
"learning_rate": 2.1276595744680852e-05,
|
| 1160 |
+
"loss": 0.0266,
|
| 1161 |
+
"step": 810
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.8900709219858154,
|
| 1165 |
+
"grad_norm": 1.974367380142212,
|
| 1166 |
+
"learning_rate": 2.1099290780141846e-05,
|
| 1167 |
+
"loss": 0.0302,
|
| 1168 |
+
"step": 815
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.9078014184397163,
|
| 1172 |
+
"grad_norm": 1.4900615215301514,
|
| 1173 |
+
"learning_rate": 2.0921985815602837e-05,
|
| 1174 |
+
"loss": 0.0499,
|
| 1175 |
+
"step": 820
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.925531914893617,
|
| 1179 |
+
"grad_norm": 2.0212316513061523,
|
| 1180 |
+
"learning_rate": 2.074468085106383e-05,
|
| 1181 |
+
"loss": 0.0411,
|
| 1182 |
+
"step": 825
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.9432624113475176,
|
| 1186 |
+
"grad_norm": 1.2716954946517944,
|
| 1187 |
+
"learning_rate": 2.0567375886524822e-05,
|
| 1188 |
+
"loss": 0.0538,
|
| 1189 |
+
"step": 830
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.9609929078014185,
|
| 1193 |
+
"grad_norm": 1.8260325193405151,
|
| 1194 |
+
"learning_rate": 2.0390070921985816e-05,
|
| 1195 |
+
"loss": 0.0405,
|
| 1196 |
+
"step": 835
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.978723404255319,
|
| 1200 |
+
"grad_norm": 0.23081009089946747,
|
| 1201 |
+
"learning_rate": 2.0212765957446807e-05,
|
| 1202 |
+
"loss": 0.0255,
|
| 1203 |
+
"step": 840
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.99645390070922,
|
| 1207 |
+
"grad_norm": 1.7825428247451782,
|
| 1208 |
+
"learning_rate": 2.0035460992907805e-05,
|
| 1209 |
+
"loss": 0.0303,
|
| 1210 |
+
"step": 845
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 3.0,
|
| 1214 |
+
"eval_accuracy": 0.9985222402837298,
|
| 1215 |
+
"eval_loss": 0.008113781921565533,
|
| 1216 |
+
"eval_runtime": 100.8534,
|
| 1217 |
+
"eval_samples_per_second": 67.097,
|
| 1218 |
+
"eval_steps_per_second": 1.051,
|
| 1219 |
+
"step": 846
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 3.0141843971631204,
|
| 1223 |
+
"grad_norm": 1.1453741788864136,
|
| 1224 |
+
"learning_rate": 1.9858156028368796e-05,
|
| 1225 |
+
"loss": 0.0207,
|
| 1226 |
+
"step": 850
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.0319148936170213,
|
| 1230 |
+
"grad_norm": 0.37254324555397034,
|
| 1231 |
+
"learning_rate": 1.968085106382979e-05,
|
| 1232 |
+
"loss": 0.0222,
|
| 1233 |
+
"step": 855
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 3.049645390070922,
|
| 1237 |
+
"grad_norm": 2.6893270015716553,
|
| 1238 |
+
"learning_rate": 1.950354609929078e-05,
|
| 1239 |
+
"loss": 0.0606,
|
| 1240 |
+
"step": 860
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 3.0673758865248226,
|
| 1244 |
+
"grad_norm": 2.4312515258789062,
|
| 1245 |
+
"learning_rate": 1.9326241134751775e-05,
|
| 1246 |
+
"loss": 0.0251,
|
| 1247 |
+
"step": 865
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 3.0851063829787235,
|
| 1251 |
+
"grad_norm": 1.2454279661178589,
|
| 1252 |
+
"learning_rate": 1.9148936170212766e-05,
|
| 1253 |
+
"loss": 0.0173,
|
| 1254 |
+
"step": 870
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 3.102836879432624,
|
| 1258 |
+
"grad_norm": 0.6530167460441589,
|
| 1259 |
+
"learning_rate": 1.897163120567376e-05,
|
| 1260 |
+
"loss": 0.0156,
|
| 1261 |
+
"step": 875
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 3.120567375886525,
|
| 1265 |
+
"grad_norm": 0.12946569919586182,
|
| 1266 |
+
"learning_rate": 1.879432624113475e-05,
|
| 1267 |
+
"loss": 0.0225,
|
| 1268 |
+
"step": 880
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.1382978723404253,
|
| 1272 |
+
"grad_norm": 0.357861191034317,
|
| 1273 |
+
"learning_rate": 1.8617021276595745e-05,
|
| 1274 |
+
"loss": 0.0121,
|
| 1275 |
+
"step": 885
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.1560283687943262,
|
| 1279 |
+
"grad_norm": 0.1260479986667633,
|
| 1280 |
+
"learning_rate": 1.8439716312056736e-05,
|
| 1281 |
+
"loss": 0.0496,
|
| 1282 |
+
"step": 890
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 3.173758865248227,
|
| 1286 |
+
"grad_norm": 0.7033432722091675,
|
| 1287 |
+
"learning_rate": 1.8262411347517733e-05,
|
| 1288 |
+
"loss": 0.0313,
|
| 1289 |
+
"step": 895
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 3.1914893617021276,
|
| 1293 |
+
"grad_norm": 1.0441181659698486,
|
| 1294 |
+
"learning_rate": 1.8085106382978724e-05,
|
| 1295 |
+
"loss": 0.0143,
|
| 1296 |
+
"step": 900
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 3.2092198581560285,
|
| 1300 |
+
"grad_norm": 0.032710377126932144,
|
| 1301 |
+
"learning_rate": 1.7907801418439718e-05,
|
| 1302 |
+
"loss": 0.0091,
|
| 1303 |
+
"step": 905
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 3.226950354609929,
|
| 1307 |
+
"grad_norm": 1.0562541484832764,
|
| 1308 |
+
"learning_rate": 1.773049645390071e-05,
|
| 1309 |
+
"loss": 0.0642,
|
| 1310 |
+
"step": 910
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 3.24468085106383,
|
| 1314 |
+
"grad_norm": 0.16099953651428223,
|
| 1315 |
+
"learning_rate": 1.7553191489361703e-05,
|
| 1316 |
+
"loss": 0.0195,
|
| 1317 |
+
"step": 915
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 3.2624113475177303,
|
| 1321 |
+
"grad_norm": 0.868291437625885,
|
| 1322 |
+
"learning_rate": 1.7375886524822697e-05,
|
| 1323 |
+
"loss": 0.0272,
|
| 1324 |
+
"step": 920
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 3.280141843971631,
|
| 1328 |
+
"grad_norm": 0.9113226532936096,
|
| 1329 |
+
"learning_rate": 1.7198581560283688e-05,
|
| 1330 |
+
"loss": 0.0317,
|
| 1331 |
+
"step": 925
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 3.297872340425532,
|
| 1335 |
+
"grad_norm": 1.1665312051773071,
|
| 1336 |
+
"learning_rate": 1.7021276595744682e-05,
|
| 1337 |
+
"loss": 0.0301,
|
| 1338 |
+
"step": 930
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 3.3156028368794326,
|
| 1342 |
+
"grad_norm": 1.0826573371887207,
|
| 1343 |
+
"learning_rate": 1.6843971631205673e-05,
|
| 1344 |
+
"loss": 0.0129,
|
| 1345 |
+
"step": 935
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 3.3333333333333335,
|
| 1349 |
+
"grad_norm": 1.054902195930481,
|
| 1350 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1351 |
+
"loss": 0.0346,
|
| 1352 |
+
"step": 940
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.351063829787234,
|
| 1356 |
+
"grad_norm": 2.4785568714141846,
|
| 1357 |
+
"learning_rate": 1.6489361702127658e-05,
|
| 1358 |
+
"loss": 0.0419,
|
| 1359 |
+
"step": 945
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 3.368794326241135,
|
| 1363 |
+
"grad_norm": 0.26840367913246155,
|
| 1364 |
+
"learning_rate": 1.6312056737588656e-05,
|
| 1365 |
+
"loss": 0.0184,
|
| 1366 |
+
"step": 950
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 3.3865248226950353,
|
| 1370 |
+
"grad_norm": 1.8092659711837769,
|
| 1371 |
+
"learning_rate": 1.6134751773049647e-05,
|
| 1372 |
+
"loss": 0.0374,
|
| 1373 |
+
"step": 955
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 3.404255319148936,
|
| 1377 |
+
"grad_norm": 0.6401882767677307,
|
| 1378 |
+
"learning_rate": 1.595744680851064e-05,
|
| 1379 |
+
"loss": 0.0521,
|
| 1380 |
+
"step": 960
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 3.421985815602837,
|
| 1384 |
+
"grad_norm": 0.3769932985305786,
|
| 1385 |
+
"learning_rate": 1.578014184397163e-05,
|
| 1386 |
+
"loss": 0.0176,
|
| 1387 |
+
"step": 965
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 3.4397163120567376,
|
| 1391 |
+
"grad_norm": 1.343386173248291,
|
| 1392 |
+
"learning_rate": 1.5602836879432626e-05,
|
| 1393 |
+
"loss": 0.0187,
|
| 1394 |
+
"step": 970
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 3.4574468085106385,
|
| 1398 |
+
"grad_norm": 0.7425516247749329,
|
| 1399 |
+
"learning_rate": 1.5425531914893617e-05,
|
| 1400 |
+
"loss": 0.0334,
|
| 1401 |
+
"step": 975
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 3.475177304964539,
|
| 1405 |
+
"grad_norm": 0.5523159503936768,
|
| 1406 |
+
"learning_rate": 1.5248226950354611e-05,
|
| 1407 |
+
"loss": 0.0233,
|
| 1408 |
+
"step": 980
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 3.49290780141844,
|
| 1412 |
+
"grad_norm": 0.3636256754398346,
|
| 1413 |
+
"learning_rate": 1.5070921985815603e-05,
|
| 1414 |
+
"loss": 0.0269,
|
| 1415 |
+
"step": 985
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 3.5106382978723403,
|
| 1419 |
+
"grad_norm": 0.7461684346199036,
|
| 1420 |
+
"learning_rate": 1.4893617021276596e-05,
|
| 1421 |
+
"loss": 0.0313,
|
| 1422 |
+
"step": 990
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 3.528368794326241,
|
| 1426 |
+
"grad_norm": 0.6355846524238586,
|
| 1427 |
+
"learning_rate": 1.4716312056737588e-05,
|
| 1428 |
+
"loss": 0.0216,
|
| 1429 |
+
"step": 995
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 3.546099290780142,
|
| 1433 |
+
"grad_norm": 0.046597789973020554,
|
| 1434 |
+
"learning_rate": 1.4539007092198581e-05,
|
| 1435 |
+
"loss": 0.0247,
|
| 1436 |
+
"step": 1000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 3.5638297872340425,
|
| 1440 |
+
"grad_norm": 0.7104584574699402,
|
| 1441 |
+
"learning_rate": 1.4361702127659577e-05,
|
| 1442 |
+
"loss": 0.0375,
|
| 1443 |
+
"step": 1005
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 3.581560283687943,
|
| 1447 |
+
"grad_norm": 0.7681288719177246,
|
| 1448 |
+
"learning_rate": 1.418439716312057e-05,
|
| 1449 |
+
"loss": 0.013,
|
| 1450 |
+
"step": 1010
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 3.599290780141844,
|
| 1454 |
+
"grad_norm": 2.825113534927368,
|
| 1455 |
+
"learning_rate": 1.4007092198581562e-05,
|
| 1456 |
+
"loss": 0.0138,
|
| 1457 |
+
"step": 1015
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 3.617021276595745,
|
| 1461 |
+
"grad_norm": 0.15184448659420013,
|
| 1462 |
+
"learning_rate": 1.3829787234042554e-05,
|
| 1463 |
+
"loss": 0.0278,
|
| 1464 |
+
"step": 1020
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 3.6347517730496453,
|
| 1468 |
+
"grad_norm": 1.885351300239563,
|
| 1469 |
+
"learning_rate": 1.3652482269503547e-05,
|
| 1470 |
+
"loss": 0.047,
|
| 1471 |
+
"step": 1025
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 3.652482269503546,
|
| 1475 |
+
"grad_norm": 1.271122932434082,
|
| 1476 |
+
"learning_rate": 1.347517730496454e-05,
|
| 1477 |
+
"loss": 0.0309,
|
| 1478 |
+
"step": 1030
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 3.670212765957447,
|
| 1482 |
+
"grad_norm": 0.02030951902270317,
|
| 1483 |
+
"learning_rate": 1.3297872340425532e-05,
|
| 1484 |
+
"loss": 0.0072,
|
| 1485 |
+
"step": 1035
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 3.6879432624113475,
|
| 1489 |
+
"grad_norm": 0.046280715614557266,
|
| 1490 |
+
"learning_rate": 1.3120567375886524e-05,
|
| 1491 |
+
"loss": 0.0219,
|
| 1492 |
+
"step": 1040
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 3.705673758865248,
|
| 1496 |
+
"grad_norm": 2.204589605331421,
|
| 1497 |
+
"learning_rate": 1.2943262411347517e-05,
|
| 1498 |
+
"loss": 0.0402,
|
| 1499 |
+
"step": 1045
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 3.723404255319149,
|
| 1503 |
+
"grad_norm": 0.6903723478317261,
|
| 1504 |
+
"learning_rate": 1.2765957446808511e-05,
|
| 1505 |
+
"loss": 0.0319,
|
| 1506 |
+
"step": 1050
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 3.7411347517730498,
|
| 1510 |
+
"grad_norm": 0.8682162165641785,
|
| 1511 |
+
"learning_rate": 1.2588652482269504e-05,
|
| 1512 |
+
"loss": 0.0296,
|
| 1513 |
+
"step": 1055
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 3.7588652482269502,
|
| 1517 |
+
"grad_norm": 0.9223999381065369,
|
| 1518 |
+
"learning_rate": 1.2411347517730498e-05,
|
| 1519 |
+
"loss": 0.0365,
|
| 1520 |
+
"step": 1060
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 3.776595744680851,
|
| 1524 |
+
"grad_norm": 0.5410795211791992,
|
| 1525 |
+
"learning_rate": 1.223404255319149e-05,
|
| 1526 |
+
"loss": 0.0121,
|
| 1527 |
+
"step": 1065
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 3.794326241134752,
|
| 1531 |
+
"grad_norm": 0.17826128005981445,
|
| 1532 |
+
"learning_rate": 1.2056737588652483e-05,
|
| 1533 |
+
"loss": 0.0231,
|
| 1534 |
+
"step": 1070
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 3.8120567375886525,
|
| 1538 |
+
"grad_norm": 0.47573089599609375,
|
| 1539 |
+
"learning_rate": 1.1879432624113477e-05,
|
| 1540 |
+
"loss": 0.0266,
|
| 1541 |
+
"step": 1075
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 3.829787234042553,
|
| 1545 |
+
"grad_norm": 0.8923744559288025,
|
| 1546 |
+
"learning_rate": 1.170212765957447e-05,
|
| 1547 |
+
"loss": 0.0345,
|
| 1548 |
+
"step": 1080
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 3.847517730496454,
|
| 1552 |
+
"grad_norm": 1.696198582649231,
|
| 1553 |
+
"learning_rate": 1.1524822695035462e-05,
|
| 1554 |
+
"loss": 0.0451,
|
| 1555 |
+
"step": 1085
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 3.8652482269503547,
|
| 1559 |
+
"grad_norm": 0.4135494828224182,
|
| 1560 |
+
"learning_rate": 1.1347517730496454e-05,
|
| 1561 |
+
"loss": 0.0248,
|
| 1562 |
+
"step": 1090
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 3.882978723404255,
|
| 1566 |
+
"grad_norm": 0.4472277760505676,
|
| 1567 |
+
"learning_rate": 1.1170212765957447e-05,
|
| 1568 |
+
"loss": 0.0125,
|
| 1569 |
+
"step": 1095
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 3.900709219858156,
|
| 1573 |
+
"grad_norm": 3.610241174697876,
|
| 1574 |
+
"learning_rate": 1.0992907801418441e-05,
|
| 1575 |
+
"loss": 0.0263,
|
| 1576 |
+
"step": 1100
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 3.918439716312057,
|
| 1580 |
+
"grad_norm": 0.06732177734375,
|
| 1581 |
+
"learning_rate": 1.0815602836879434e-05,
|
| 1582 |
+
"loss": 0.0153,
|
| 1583 |
+
"step": 1105
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 3.9361702127659575,
|
| 1587 |
+
"grad_norm": 0.8051948547363281,
|
| 1588 |
+
"learning_rate": 1.0638297872340426e-05,
|
| 1589 |
+
"loss": 0.0217,
|
| 1590 |
+
"step": 1110
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 3.953900709219858,
|
| 1594 |
+
"grad_norm": 0.24708445370197296,
|
| 1595 |
+
"learning_rate": 1.0460992907801419e-05,
|
| 1596 |
+
"loss": 0.0432,
|
| 1597 |
+
"step": 1115
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 3.971631205673759,
|
| 1601 |
+
"grad_norm": 0.0366179421544075,
|
| 1602 |
+
"learning_rate": 1.0283687943262411e-05,
|
| 1603 |
+
"loss": 0.0065,
|
| 1604 |
+
"step": 1120
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 3.9893617021276597,
|
| 1608 |
+
"grad_norm": 1.3547054529190063,
|
| 1609 |
+
"learning_rate": 1.0106382978723404e-05,
|
| 1610 |
+
"loss": 0.0489,
|
| 1611 |
+
"step": 1125
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 4.0,
|
| 1615 |
+
"eval_accuracy": 0.9988177922269839,
|
| 1616 |
+
"eval_loss": 0.0067949132062494755,
|
| 1617 |
+
"eval_runtime": 99.24,
|
| 1618 |
+
"eval_samples_per_second": 68.188,
|
| 1619 |
+
"eval_steps_per_second": 1.068,
|
| 1620 |
+
"step": 1128
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 4.00709219858156,
|
| 1624 |
+
"grad_norm": 0.2567829191684723,
|
| 1625 |
+
"learning_rate": 9.929078014184398e-06,
|
| 1626 |
+
"loss": 0.0132,
|
| 1627 |
+
"step": 1130
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.024822695035461,
|
| 1631 |
+
"grad_norm": 0.8916745781898499,
|
| 1632 |
+
"learning_rate": 9.75177304964539e-06,
|
| 1633 |
+
"loss": 0.0217,
|
| 1634 |
+
"step": 1135
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.042553191489362,
|
| 1638 |
+
"grad_norm": 0.13239595293998718,
|
| 1639 |
+
"learning_rate": 9.574468085106383e-06,
|
| 1640 |
+
"loss": 0.0171,
|
| 1641 |
+
"step": 1140
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 4.060283687943262,
|
| 1645 |
+
"grad_norm": 0.2502301335334778,
|
| 1646 |
+
"learning_rate": 9.397163120567375e-06,
|
| 1647 |
+
"loss": 0.006,
|
| 1648 |
+
"step": 1145
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 4.078014184397163,
|
| 1652 |
+
"grad_norm": 0.3878186047077179,
|
| 1653 |
+
"learning_rate": 9.219858156028368e-06,
|
| 1654 |
+
"loss": 0.0184,
|
| 1655 |
+
"step": 1150
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 4.095744680851064,
|
| 1659 |
+
"grad_norm": 1.005374789237976,
|
| 1660 |
+
"learning_rate": 9.042553191489362e-06,
|
| 1661 |
+
"loss": 0.0633,
|
| 1662 |
+
"step": 1155
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 4.113475177304965,
|
| 1666 |
+
"grad_norm": 1.9449095726013184,
|
| 1667 |
+
"learning_rate": 8.865248226950355e-06,
|
| 1668 |
+
"loss": 0.027,
|
| 1669 |
+
"step": 1160
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 4.131205673758865,
|
| 1673 |
+
"grad_norm": 0.04838603362441063,
|
| 1674 |
+
"learning_rate": 8.687943262411349e-06,
|
| 1675 |
+
"loss": 0.0077,
|
| 1676 |
+
"step": 1165
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.148936170212766,
|
| 1680 |
+
"grad_norm": 0.07912664860486984,
|
| 1681 |
+
"learning_rate": 8.510638297872341e-06,
|
| 1682 |
+
"loss": 0.0075,
|
| 1683 |
+
"step": 1170
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 4.166666666666667,
|
| 1687 |
+
"grad_norm": 0.24843692779541016,
|
| 1688 |
+
"learning_rate": 8.333333333333334e-06,
|
| 1689 |
+
"loss": 0.0134,
|
| 1690 |
+
"step": 1175
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 4.184397163120567,
|
| 1694 |
+
"grad_norm": 0.7264091968536377,
|
| 1695 |
+
"learning_rate": 8.156028368794328e-06,
|
| 1696 |
+
"loss": 0.0367,
|
| 1697 |
+
"step": 1180
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 4.202127659574468,
|
| 1701 |
+
"grad_norm": 1.442780613899231,
|
| 1702 |
+
"learning_rate": 7.97872340425532e-06,
|
| 1703 |
+
"loss": 0.0143,
|
| 1704 |
+
"step": 1185
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.219858156028369,
|
| 1708 |
+
"grad_norm": 0.4630628228187561,
|
| 1709 |
+
"learning_rate": 7.801418439716313e-06,
|
| 1710 |
+
"loss": 0.0506,
|
| 1711 |
+
"step": 1190
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 4.23758865248227,
|
| 1715 |
+
"grad_norm": 1.1415939331054688,
|
| 1716 |
+
"learning_rate": 7.6241134751773054e-06,
|
| 1717 |
+
"loss": 0.0266,
|
| 1718 |
+
"step": 1195
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.25531914893617,
|
| 1722 |
+
"grad_norm": 0.8882013559341431,
|
| 1723 |
+
"learning_rate": 7.446808510638298e-06,
|
| 1724 |
+
"loss": 0.0159,
|
| 1725 |
+
"step": 1200
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 4.273049645390071,
|
| 1729 |
+
"grad_norm": 0.12462817132472992,
|
| 1730 |
+
"learning_rate": 7.2695035460992904e-06,
|
| 1731 |
+
"loss": 0.0209,
|
| 1732 |
+
"step": 1205
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 4.290780141843972,
|
| 1736 |
+
"grad_norm": 1.5256775617599487,
|
| 1737 |
+
"learning_rate": 7.092198581560285e-06,
|
| 1738 |
+
"loss": 0.0241,
|
| 1739 |
+
"step": 1210
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 4.308510638297872,
|
| 1743 |
+
"grad_norm": 3.1360552310943604,
|
| 1744 |
+
"learning_rate": 6.914893617021277e-06,
|
| 1745 |
+
"loss": 0.0298,
|
| 1746 |
+
"step": 1215
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 4.326241134751773,
|
| 1750 |
+
"grad_norm": 0.7372686266899109,
|
| 1751 |
+
"learning_rate": 6.73758865248227e-06,
|
| 1752 |
+
"loss": 0.0193,
|
| 1753 |
+
"step": 1220
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 4.343971631205674,
|
| 1757 |
+
"grad_norm": 1.058393955230713,
|
| 1758 |
+
"learning_rate": 6.560283687943262e-06,
|
| 1759 |
+
"loss": 0.0293,
|
| 1760 |
+
"step": 1225
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 4.361702127659575,
|
| 1764 |
+
"grad_norm": 0.09787444770336151,
|
| 1765 |
+
"learning_rate": 6.3829787234042555e-06,
|
| 1766 |
+
"loss": 0.0098,
|
| 1767 |
+
"step": 1230
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 4.379432624113475,
|
| 1771 |
+
"grad_norm": 0.4723568856716156,
|
| 1772 |
+
"learning_rate": 6.205673758865249e-06,
|
| 1773 |
+
"loss": 0.0248,
|
| 1774 |
+
"step": 1235
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 4.397163120567376,
|
| 1778 |
+
"grad_norm": 0.3988282382488251,
|
| 1779 |
+
"learning_rate": 6.028368794326241e-06,
|
| 1780 |
+
"loss": 0.0123,
|
| 1781 |
+
"step": 1240
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 4.414893617021277,
|
| 1785 |
+
"grad_norm": 0.080900177359581,
|
| 1786 |
+
"learning_rate": 5.851063829787235e-06,
|
| 1787 |
+
"loss": 0.0361,
|
| 1788 |
+
"step": 1245
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 4.432624113475177,
|
| 1792 |
+
"grad_norm": 0.28941839933395386,
|
| 1793 |
+
"learning_rate": 5.673758865248227e-06,
|
| 1794 |
+
"loss": 0.0173,
|
| 1795 |
+
"step": 1250
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 4.450354609929078,
|
| 1799 |
+
"grad_norm": 0.16654981672763824,
|
| 1800 |
+
"learning_rate": 5.4964539007092206e-06,
|
| 1801 |
+
"loss": 0.0374,
|
| 1802 |
+
"step": 1255
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.468085106382979,
|
| 1806 |
+
"grad_norm": 0.5223755240440369,
|
| 1807 |
+
"learning_rate": 5.319148936170213e-06,
|
| 1808 |
+
"loss": 0.0077,
|
| 1809 |
+
"step": 1260
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 4.48581560283688,
|
| 1813 |
+
"grad_norm": 0.7028219103813171,
|
| 1814 |
+
"learning_rate": 5.1418439716312056e-06,
|
| 1815 |
+
"loss": 0.0299,
|
| 1816 |
+
"step": 1265
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 4.50354609929078,
|
| 1820 |
+
"grad_norm": 0.3213287889957428,
|
| 1821 |
+
"learning_rate": 4.964539007092199e-06,
|
| 1822 |
+
"loss": 0.0569,
|
| 1823 |
+
"step": 1270
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 4.5212765957446805,
|
| 1827 |
+
"grad_norm": 0.26597318053245544,
|
| 1828 |
+
"learning_rate": 4.787234042553191e-06,
|
| 1829 |
+
"loss": 0.0122,
|
| 1830 |
+
"step": 1275
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 4.539007092198582,
|
| 1834 |
+
"grad_norm": 1.5103939771652222,
|
| 1835 |
+
"learning_rate": 4.609929078014184e-06,
|
| 1836 |
+
"loss": 0.021,
|
| 1837 |
+
"step": 1280
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 4.556737588652482,
|
| 1841 |
+
"grad_norm": 0.4471294581890106,
|
| 1842 |
+
"learning_rate": 4.432624113475177e-06,
|
| 1843 |
+
"loss": 0.0361,
|
| 1844 |
+
"step": 1285
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.574468085106383,
|
| 1848 |
+
"grad_norm": 0.21643757820129395,
|
| 1849 |
+
"learning_rate": 4.255319148936171e-06,
|
| 1850 |
+
"loss": 0.0294,
|
| 1851 |
+
"step": 1290
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 4.592198581560284,
|
| 1855 |
+
"grad_norm": 2.008852958679199,
|
| 1856 |
+
"learning_rate": 4.078014184397164e-06,
|
| 1857 |
+
"loss": 0.0301,
|
| 1858 |
+
"step": 1295
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 4.609929078014185,
|
| 1862 |
+
"grad_norm": 1.9454808235168457,
|
| 1863 |
+
"learning_rate": 3.9007092198581565e-06,
|
| 1864 |
+
"loss": 0.0468,
|
| 1865 |
+
"step": 1300
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 4.627659574468085,
|
| 1869 |
+
"grad_norm": 0.3638037443161011,
|
| 1870 |
+
"learning_rate": 3.723404255319149e-06,
|
| 1871 |
+
"loss": 0.009,
|
| 1872 |
+
"step": 1305
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 4.6453900709219855,
|
| 1876 |
+
"grad_norm": 0.5267730951309204,
|
| 1877 |
+
"learning_rate": 3.5460992907801423e-06,
|
| 1878 |
+
"loss": 0.0106,
|
| 1879 |
+
"step": 1310
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 4.663120567375887,
|
| 1883 |
+
"grad_norm": 0.6058443188667297,
|
| 1884 |
+
"learning_rate": 3.368794326241135e-06,
|
| 1885 |
+
"loss": 0.0408,
|
| 1886 |
+
"step": 1315
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.680851063829787,
|
| 1890 |
+
"grad_norm": 0.7102274894714355,
|
| 1891 |
+
"learning_rate": 3.1914893617021277e-06,
|
| 1892 |
+
"loss": 0.0177,
|
| 1893 |
+
"step": 1320
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 4.698581560283688,
|
| 1897 |
+
"grad_norm": 0.03302247077226639,
|
| 1898 |
+
"learning_rate": 3.0141843971631207e-06,
|
| 1899 |
+
"loss": 0.0263,
|
| 1900 |
+
"step": 1325
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 4.716312056737589,
|
| 1904 |
+
"grad_norm": 0.29536840319633484,
|
| 1905 |
+
"learning_rate": 2.8368794326241136e-06,
|
| 1906 |
+
"loss": 0.0176,
|
| 1907 |
+
"step": 1330
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 4.73404255319149,
|
| 1911 |
+
"grad_norm": 0.303792268037796,
|
| 1912 |
+
"learning_rate": 2.6595744680851065e-06,
|
| 1913 |
+
"loss": 0.0149,
|
| 1914 |
+
"step": 1335
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 4.75177304964539,
|
| 1918 |
+
"grad_norm": 1.4943726062774658,
|
| 1919 |
+
"learning_rate": 2.4822695035460995e-06,
|
| 1920 |
+
"loss": 0.0324,
|
| 1921 |
+
"step": 1340
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 4.7695035460992905,
|
| 1925 |
+
"grad_norm": 0.2641482949256897,
|
| 1926 |
+
"learning_rate": 2.304964539007092e-06,
|
| 1927 |
+
"loss": 0.0258,
|
| 1928 |
+
"step": 1345
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.787234042553192,
|
| 1932 |
+
"grad_norm": 1.3771953582763672,
|
| 1933 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 1934 |
+
"loss": 0.0325,
|
| 1935 |
+
"step": 1350
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 4.804964539007092,
|
| 1939 |
+
"grad_norm": 2.062373638153076,
|
| 1940 |
+
"learning_rate": 1.9503546099290782e-06,
|
| 1941 |
+
"loss": 0.0152,
|
| 1942 |
+
"step": 1355
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 4.822695035460993,
|
| 1946 |
+
"grad_norm": 0.5422120094299316,
|
| 1947 |
+
"learning_rate": 1.7730496453900712e-06,
|
| 1948 |
+
"loss": 0.0387,
|
| 1949 |
+
"step": 1360
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 4.840425531914894,
|
| 1953 |
+
"grad_norm": 0.02120036445558071,
|
| 1954 |
+
"learning_rate": 1.5957446808510639e-06,
|
| 1955 |
+
"loss": 0.0122,
|
| 1956 |
+
"step": 1365
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 4.858156028368795,
|
| 1960 |
+
"grad_norm": 0.2176242172718048,
|
| 1961 |
+
"learning_rate": 1.4184397163120568e-06,
|
| 1962 |
+
"loss": 0.0205,
|
| 1963 |
+
"step": 1370
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 4.875886524822695,
|
| 1967 |
+
"grad_norm": 2.1586310863494873,
|
| 1968 |
+
"learning_rate": 1.2411347517730497e-06,
|
| 1969 |
+
"loss": 0.0245,
|
| 1970 |
+
"step": 1375
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.8936170212765955,
|
| 1974 |
+
"grad_norm": 0.3032481074333191,
|
| 1975 |
+
"learning_rate": 1.0638297872340427e-06,
|
| 1976 |
+
"loss": 0.0317,
|
| 1977 |
+
"step": 1380
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 4.911347517730497,
|
| 1981 |
+
"grad_norm": 0.1574692726135254,
|
| 1982 |
+
"learning_rate": 8.865248226950356e-07,
|
| 1983 |
+
"loss": 0.0126,
|
| 1984 |
+
"step": 1385
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 4.929078014184397,
|
| 1988 |
+
"grad_norm": 1.4468116760253906,
|
| 1989 |
+
"learning_rate": 7.092198581560284e-07,
|
| 1990 |
+
"loss": 0.0166,
|
| 1991 |
+
"step": 1390
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 4.946808510638298,
|
| 1995 |
+
"grad_norm": 0.5175489783287048,
|
| 1996 |
+
"learning_rate": 5.319148936170213e-07,
|
| 1997 |
+
"loss": 0.0101,
|
| 1998 |
+
"step": 1395
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 4.964539007092198,
|
| 2002 |
+
"grad_norm": 0.04873587563633919,
|
| 2003 |
+
"learning_rate": 3.546099290780142e-07,
|
| 2004 |
+
"loss": 0.0167,
|
| 2005 |
+
"step": 1400
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 4.9822695035460995,
|
| 2009 |
+
"grad_norm": 3.589585304260254,
|
| 2010 |
+
"learning_rate": 1.773049645390071e-07,
|
| 2011 |
+
"loss": 0.0284,
|
| 2012 |
+
"step": 1405
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 5.0,
|
| 2016 |
+
"grad_norm": 0.1343780905008316,
|
| 2017 |
+
"learning_rate": 0.0,
|
| 2018 |
+
"loss": 0.0284,
|
| 2019 |
+
"step": 1410
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 5.0,
|
| 2023 |
+
"eval_accuracy": 0.9989655681986109,
|
| 2024 |
+
"eval_loss": 0.006617749575525522,
|
| 2025 |
+
"eval_runtime": 99.2975,
|
| 2026 |
+
"eval_samples_per_second": 68.149,
|
| 2027 |
+
"eval_steps_per_second": 1.067,
|
| 2028 |
+
"step": 1410
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 5.0,
|
| 2032 |
+
"step": 1410,
|
| 2033 |
+
"total_flos": 6.98086588443052e+18,
|
| 2034 |
+
"train_loss": 0.042465827911299596,
|
| 2035 |
+
"train_runtime": 3748.8226,
|
| 2036 |
+
"train_samples_per_second": 24.03,
|
| 2037 |
+
"train_steps_per_second": 0.376
|
| 2038 |
+
}
|
| 2039 |
+
],
|
| 2040 |
+
"logging_steps": 5,
|
| 2041 |
+
"max_steps": 1410,
|
| 2042 |
+
"num_input_tokens_seen": 0,
|
| 2043 |
+
"num_train_epochs": 5,
|
| 2044 |
+
"save_steps": 500,
|
| 2045 |
+
"stateful_callbacks": {
|
| 2046 |
+
"TrainerControl": {
|
| 2047 |
+
"args": {
|
| 2048 |
+
"should_epoch_stop": false,
|
| 2049 |
+
"should_evaluate": false,
|
| 2050 |
+
"should_log": false,
|
| 2051 |
+
"should_save": true,
|
| 2052 |
+
"should_training_stop": true
|
| 2053 |
+
},
|
| 2054 |
+
"attributes": {}
|
| 2055 |
+
}
|
| 2056 |
+
},
|
| 2057 |
+
"total_flos": 6.98086588443052e+18,
|
| 2058 |
+
"train_batch_size": 64,
|
| 2059 |
+
"trial_name": null,
|
| 2060 |
+
"trial_params": null
|
| 2061 |
+
}
|