End of training
Browse files- README.md +2 -0
- all_results.json +14 -0
- eval_results.json +8 -0
- runs/Aug26_10-22-55_bubbles.cl.cam.ac.uk/events.out.tfevents.1756202604.bubbles.cl.cam.ac.uk.2004449.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +1142 -0
README.md
CHANGED
|
@@ -14,6 +14,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
# russian-babylm-ft
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [climb-mao/russian-babylm-urop-shivan](https://huggingface.co/climb-mao/russian-babylm-urop-shivan) on an unknown dataset.
|
|
|
|
|
|
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
|
|
| 14 |
# russian-babylm-ft
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [climb-mao/russian-babylm-urop-shivan](https://huggingface.co/climb-mao/russian-babylm-urop-shivan) on an unknown dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 2.6414
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 2.6413557529449463,
|
| 4 |
+
"eval_runtime": 2.9908,
|
| 5 |
+
"eval_samples_per_second": 143.773,
|
| 6 |
+
"eval_steps_per_second": 18.055,
|
| 7 |
+
"perplexity": 14.032214934211854,
|
| 8 |
+
"total_flos": 4871316072038400.0,
|
| 9 |
+
"train_loss": 3.0463915172120695,
|
| 10 |
+
"train_runtime": 2406.5677,
|
| 11 |
+
"train_samples": 41915,
|
| 12 |
+
"train_samples_per_second": 52.251,
|
| 13 |
+
"train_steps_per_second": 3.266
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 2.6413557529449463,
|
| 4 |
+
"eval_runtime": 2.9908,
|
| 5 |
+
"eval_samples_per_second": 143.773,
|
| 6 |
+
"eval_steps_per_second": 18.055,
|
| 7 |
+
"perplexity": 14.032214934211854
|
| 8 |
+
}
|
runs/Aug26_10-22-55_bubbles.cl.cam.ac.uk/events.out.tfevents.1756202604.bubbles.cl.cam.ac.uk.2004449.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cffed173647e71030e3d464bc75b051bff6830291a67443299e32535d5d6c30
|
| 3 |
+
size 359
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 4871316072038400.0,
|
| 4 |
+
"train_loss": 3.0463915172120695,
|
| 5 |
+
"train_runtime": 2406.5677,
|
| 6 |
+
"train_samples": 41915,
|
| 7 |
+
"train_samples_per_second": 52.251,
|
| 8 |
+
"train_steps_per_second": 3.266
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1142 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 2000,
|
| 7 |
+
"global_step": 7860,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.019083969465648856,
|
| 14 |
+
"grad_norm": 1.5055166482925415,
|
| 15 |
+
"learning_rate": 4.152542372881356e-05,
|
| 16 |
+
"loss": 7.5506,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03816793893129771,
|
| 21 |
+
"grad_norm": 2.098764181137085,
|
| 22 |
+
"learning_rate": 8.389830508474577e-05,
|
| 23 |
+
"loss": 6.1122,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05725190839694656,
|
| 28 |
+
"grad_norm": 1.3783698081970215,
|
| 29 |
+
"learning_rate": 0.00012627118644067796,
|
| 30 |
+
"loss": 4.9663,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07633587786259542,
|
| 35 |
+
"grad_norm": 1.1028858423233032,
|
| 36 |
+
"learning_rate": 0.00016864406779661017,
|
| 37 |
+
"loss": 4.4607,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09541984732824428,
|
| 42 |
+
"grad_norm": 1.4427648782730103,
|
| 43 |
+
"learning_rate": 0.0001999985652057071,
|
| 44 |
+
"loss": 4.2266,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11450381679389313,
|
| 49 |
+
"grad_norm": 0.9209873676300049,
|
| 50 |
+
"learning_rate": 0.00019996630537068877,
|
| 51 |
+
"loss": 4.0824,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.13358778625954199,
|
| 56 |
+
"grad_norm": 1.0161939859390259,
|
| 57 |
+
"learning_rate": 0.00019989161170323044,
|
| 58 |
+
"loss": 4.0219,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.15267175572519084,
|
| 63 |
+
"grad_norm": 1.0433142185211182,
|
| 64 |
+
"learning_rate": 0.000199774515909401,
|
| 65 |
+
"loss": 3.9481,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1717557251908397,
|
| 70 |
+
"grad_norm": 0.9553391337394714,
|
| 71 |
+
"learning_rate": 0.00019961506769418144,
|
| 72 |
+
"loss": 3.888,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.19083969465648856,
|
| 77 |
+
"grad_norm": 0.9174131155014038,
|
| 78 |
+
"learning_rate": 0.00019941333474036556,
|
| 79 |
+
"loss": 3.8782,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.2099236641221374,
|
| 84 |
+
"grad_norm": 1.0656416416168213,
|
| 85 |
+
"learning_rate": 0.0001991694026798304,
|
| 86 |
+
"loss": 3.8321,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22900763358778625,
|
| 91 |
+
"grad_norm": 0.8713803887367249,
|
| 92 |
+
"learning_rate": 0.00019888337505718674,
|
| 93 |
+
"loss": 3.7897,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.2480916030534351,
|
| 98 |
+
"grad_norm": 1.0437265634536743,
|
| 99 |
+
"learning_rate": 0.00019855537328582648,
|
| 100 |
+
"loss": 3.7609,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.26717557251908397,
|
| 105 |
+
"grad_norm": 0.9695500731468201,
|
| 106 |
+
"learning_rate": 0.00019818553659638492,
|
| 107 |
+
"loss": 3.7086,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.2862595419847328,
|
| 112 |
+
"grad_norm": 1.0067107677459717,
|
| 113 |
+
"learning_rate": 0.00019777402197763996,
|
| 114 |
+
"loss": 3.6706,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3053435114503817,
|
| 119 |
+
"grad_norm": 1.1255054473876953,
|
| 120 |
+
"learning_rate": 0.00019732100410987328,
|
| 121 |
+
"loss": 3.6353,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.3244274809160305,
|
| 126 |
+
"grad_norm": 1.0459941625595093,
|
| 127 |
+
"learning_rate": 0.0001968266752907218,
|
| 128 |
+
"loss": 3.5962,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.3435114503816794,
|
| 133 |
+
"grad_norm": 1.445251226425171,
|
| 134 |
+
"learning_rate": 0.00019629124535355084,
|
| 135 |
+
"loss": 3.5666,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.36259541984732824,
|
| 140 |
+
"grad_norm": 1.099308729171753,
|
| 141 |
+
"learning_rate": 0.00019571494157838396,
|
| 142 |
+
"loss": 3.5367,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.3816793893129771,
|
| 147 |
+
"grad_norm": 1.1672636270523071,
|
| 148 |
+
"learning_rate": 0.00019509800859542677,
|
| 149 |
+
"loss": 3.5084,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.40076335877862596,
|
| 154 |
+
"grad_norm": 1.1235790252685547,
|
| 155 |
+
"learning_rate": 0.00019444070828122579,
|
| 156 |
+
"loss": 3.462,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.4198473282442748,
|
| 161 |
+
"grad_norm": 1.2316577434539795,
|
| 162 |
+
"learning_rate": 0.00019374331964750694,
|
| 163 |
+
"loss": 3.4475,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4389312977099237,
|
| 168 |
+
"grad_norm": 1.098713755607605,
|
| 169 |
+
"learning_rate": 0.00019300613872274023,
|
| 170 |
+
"loss": 3.4039,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.4580152671755725,
|
| 175 |
+
"grad_norm": 1.1266428232192993,
|
| 176 |
+
"learning_rate": 0.00019222947842648114,
|
| 177 |
+
"loss": 3.4017,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4770992366412214,
|
| 182 |
+
"grad_norm": 1.1996225118637085,
|
| 183 |
+
"learning_rate": 0.0001914136684365421,
|
| 184 |
+
"loss": 3.3432,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.4961832061068702,
|
| 189 |
+
"grad_norm": 1.1482900381088257,
|
| 190 |
+
"learning_rate": 0.0001905590550490505,
|
| 191 |
+
"loss": 3.338,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5152671755725191,
|
| 196 |
+
"grad_norm": 1.2343032360076904,
|
| 197 |
+
"learning_rate": 0.0001896660010314523,
|
| 198 |
+
"loss": 3.3095,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5343511450381679,
|
| 203 |
+
"grad_norm": 1.2895420789718628,
|
| 204 |
+
"learning_rate": 0.0001887348854685242,
|
| 205 |
+
"loss": 3.3011,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5534351145038168,
|
| 210 |
+
"grad_norm": 1.2781686782836914,
|
| 211 |
+
"learning_rate": 0.000187766103601459,
|
| 212 |
+
"loss": 3.2701,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5725190839694656,
|
| 217 |
+
"grad_norm": 1.2175929546356201,
|
| 218 |
+
"learning_rate": 0.00018676006666009328,
|
| 219 |
+
"loss": 3.258,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5916030534351145,
|
| 224 |
+
"grad_norm": 1.163954257965088,
|
| 225 |
+
"learning_rate": 0.00018571720168834786,
|
| 226 |
+
"loss": 3.2395,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6106870229007634,
|
| 231 |
+
"grad_norm": 1.1487972736358643,
|
| 232 |
+
"learning_rate": 0.00018463795136295573,
|
| 233 |
+
"loss": 3.2087,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6297709923664122,
|
| 238 |
+
"grad_norm": 1.2098793983459473,
|
| 239 |
+
"learning_rate": 0.00018352277380555393,
|
| 240 |
+
"loss": 3.2066,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.648854961832061,
|
| 245 |
+
"grad_norm": 1.1558622121810913,
|
| 246 |
+
"learning_rate": 0.0001823721423882196,
|
| 247 |
+
"loss": 3.1798,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.6679389312977099,
|
| 252 |
+
"grad_norm": 1.2244099378585815,
|
| 253 |
+
"learning_rate": 0.0001811865455325322,
|
| 254 |
+
"loss": 3.1717,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6870229007633588,
|
| 259 |
+
"grad_norm": 1.2974528074264526,
|
| 260 |
+
"learning_rate": 0.00017996648650224769,
|
| 261 |
+
"loss": 3.1577,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.7061068702290076,
|
| 266 |
+
"grad_norm": 1.037485122680664,
|
| 267 |
+
"learning_rate": 0.0001787124831896726,
|
| 268 |
+
"loss": 3.1661,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7251908396946565,
|
| 273 |
+
"grad_norm": 1.2607316970825195,
|
| 274 |
+
"learning_rate": 0.0001774250678958282,
|
| 275 |
+
"loss": 3.1344,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7442748091603053,
|
| 280 |
+
"grad_norm": 1.202499508857727,
|
| 281 |
+
"learning_rate": 0.00017610478710449877,
|
| 282 |
+
"loss": 3.1196,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7633587786259542,
|
| 287 |
+
"grad_norm": 1.137555480003357,
|
| 288 |
+
"learning_rate": 0.00017475220125025946,
|
| 289 |
+
"loss": 3.1141,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7824427480916031,
|
| 294 |
+
"grad_norm": 1.1263173818588257,
|
| 295 |
+
"learning_rate": 0.00017336788448058205,
|
| 296 |
+
"loss": 3.0992,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.8015267175572519,
|
| 301 |
+
"grad_norm": 1.1676143407821655,
|
| 302 |
+
"learning_rate": 0.0001719524244121203,
|
| 303 |
+
"loss": 3.0836,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8206106870229007,
|
| 308 |
+
"grad_norm": 1.1205627918243408,
|
| 309 |
+
"learning_rate": 0.00017050642188127756,
|
| 310 |
+
"loss": 3.0902,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8396946564885496,
|
| 315 |
+
"grad_norm": 1.214097023010254,
|
| 316 |
+
"learning_rate": 0.00016903049068916278,
|
| 317 |
+
"loss": 3.0646,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8587786259541985,
|
| 322 |
+
"grad_norm": 1.1289393901824951,
|
| 323 |
+
"learning_rate": 0.0001675252573410436,
|
| 324 |
+
"loss": 3.0548,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8778625954198473,
|
| 329 |
+
"grad_norm": 1.1379965543746948,
|
| 330 |
+
"learning_rate": 0.00016599136078040646,
|
| 331 |
+
"loss": 3.058,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8969465648854962,
|
| 336 |
+
"grad_norm": 1.1136966943740845,
|
| 337 |
+
"learning_rate": 0.00016442945211773697,
|
| 338 |
+
"loss": 3.0348,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.916030534351145,
|
| 343 |
+
"grad_norm": 1.0273264646530151,
|
| 344 |
+
"learning_rate": 0.00016284019435413563,
|
| 345 |
+
"loss": 3.0194,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9351145038167938,
|
| 350 |
+
"grad_norm": 1.0680382251739502,
|
| 351 |
+
"learning_rate": 0.00016122426209988638,
|
| 352 |
+
"loss": 3.0234,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9541984732824428,
|
| 357 |
+
"grad_norm": 1.0025871992111206,
|
| 358 |
+
"learning_rate": 0.00015958234128809687,
|
| 359 |
+
"loss": 3.0161,
|
| 360 |
+
"step": 2500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9732824427480916,
|
| 364 |
+
"grad_norm": 1.1902124881744385,
|
| 365 |
+
"learning_rate": 0.00015791512888353295,
|
| 366 |
+
"loss": 3.0114,
|
| 367 |
+
"step": 2550
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.9923664122137404,
|
| 371 |
+
"grad_norm": 1.0957531929016113,
|
| 372 |
+
"learning_rate": 0.00015622333258676982,
|
| 373 |
+
"loss": 3.0071,
|
| 374 |
+
"step": 2600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.0114503816793894,
|
| 378 |
+
"grad_norm": 1.25103759765625,
|
| 379 |
+
"learning_rate": 0.00015450767053378655,
|
| 380 |
+
"loss": 2.9835,
|
| 381 |
+
"step": 2650
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0305343511450382,
|
| 385 |
+
"grad_norm": 1.067357063293457,
|
| 386 |
+
"learning_rate": 0.00015276887099113069,
|
| 387 |
+
"loss": 2.9665,
|
| 388 |
+
"step": 2700
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.049618320610687,
|
| 392 |
+
"grad_norm": 1.1145495176315308,
|
| 393 |
+
"learning_rate": 0.00015100767204678258,
|
| 394 |
+
"loss": 2.9694,
|
| 395 |
+
"step": 2750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0687022900763359,
|
| 399 |
+
"grad_norm": 1.0184941291809082,
|
| 400 |
+
"learning_rate": 0.00014922482129685097,
|
| 401 |
+
"loss": 2.9663,
|
| 402 |
+
"step": 2800
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0877862595419847,
|
| 406 |
+
"grad_norm": 0.9974035620689392,
|
| 407 |
+
"learning_rate": 0.00014742107552823228,
|
| 408 |
+
"loss": 2.9416,
|
| 409 |
+
"step": 2850
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.1068702290076335,
|
| 413 |
+
"grad_norm": 1.06545889377594,
|
| 414 |
+
"learning_rate": 0.00014559720039736872,
|
| 415 |
+
"loss": 2.9492,
|
| 416 |
+
"step": 2900
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1259541984732824,
|
| 420 |
+
"grad_norm": 0.9931660294532776,
|
| 421 |
+
"learning_rate": 0.00014375397010524135,
|
| 422 |
+
"loss": 2.9452,
|
| 423 |
+
"step": 2950
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1450381679389312,
|
| 427 |
+
"grad_norm": 1.133661150932312,
|
| 428 |
+
"learning_rate": 0.00014189216706873636,
|
| 429 |
+
"loss": 2.9227,
|
| 430 |
+
"step": 3000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.16412213740458,
|
| 434 |
+
"grad_norm": 0.9924533367156982,
|
| 435 |
+
"learning_rate": 0.00014001258158852342,
|
| 436 |
+
"loss": 2.9389,
|
| 437 |
+
"step": 3050
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.183206106870229,
|
| 441 |
+
"grad_norm": 1.1022679805755615,
|
| 442 |
+
"learning_rate": 0.00013811601151358805,
|
| 443 |
+
"loss": 2.9322,
|
| 444 |
+
"step": 3100
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.202290076335878,
|
| 448 |
+
"grad_norm": 1.1133204698562622,
|
| 449 |
+
"learning_rate": 0.0001362032619025595,
|
| 450 |
+
"loss": 2.9197,
|
| 451 |
+
"step": 3150
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2213740458015268,
|
| 455 |
+
"grad_norm": 1.048754334449768,
|
| 456 |
+
"learning_rate": 0.00013427514468197834,
|
| 457 |
+
"loss": 2.9118,
|
| 458 |
+
"step": 3200
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.2404580152671756,
|
| 462 |
+
"grad_norm": 1.0499931573867798,
|
| 463 |
+
"learning_rate": 0.0001323324783016488,
|
| 464 |
+
"loss": 2.9109,
|
| 465 |
+
"step": 3250
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2595419847328244,
|
| 469 |
+
"grad_norm": 1.0267069339752197,
|
| 470 |
+
"learning_rate": 0.00013037608738722216,
|
| 471 |
+
"loss": 2.8971,
|
| 472 |
+
"step": 3300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2786259541984732,
|
| 476 |
+
"grad_norm": 1.036413550376892,
|
| 477 |
+
"learning_rate": 0.00012840680239015846,
|
| 478 |
+
"loss": 2.8914,
|
| 479 |
+
"step": 3350
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.297709923664122,
|
| 483 |
+
"grad_norm": 1.0500997304916382,
|
| 484 |
+
"learning_rate": 0.00012642545923521536,
|
| 485 |
+
"loss": 2.8972,
|
| 486 |
+
"step": 3400
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.316793893129771,
|
| 490 |
+
"grad_norm": 1.1040946245193481,
|
| 491 |
+
"learning_rate": 0.0001244328989656137,
|
| 492 |
+
"loss": 2.8801,
|
| 493 |
+
"step": 3450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.33587786259542,
|
| 497 |
+
"grad_norm": 1.046844720840454,
|
| 498 |
+
"learning_rate": 0.00012242996738603027,
|
| 499 |
+
"loss": 2.8911,
|
| 500 |
+
"step": 3500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3549618320610688,
|
| 504 |
+
"grad_norm": 1.0509463548660278,
|
| 505 |
+
"learning_rate": 0.00012041751470356957,
|
| 506 |
+
"loss": 2.8793,
|
| 507 |
+
"step": 3550
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3740458015267176,
|
| 511 |
+
"grad_norm": 1.0559144020080566,
|
| 512 |
+
"learning_rate": 0.00011839639516686675,
|
| 513 |
+
"loss": 2.8767,
|
| 514 |
+
"step": 3600
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3931297709923665,
|
| 518 |
+
"grad_norm": 1.1194440126419067,
|
| 519 |
+
"learning_rate": 0.00011636746670347493,
|
| 520 |
+
"loss": 2.8698,
|
| 521 |
+
"step": 3650
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.4122137404580153,
|
| 525 |
+
"grad_norm": 1.1573190689086914,
|
| 526 |
+
"learning_rate": 0.000114331590555691,
|
| 527 |
+
"loss": 2.8704,
|
| 528 |
+
"step": 3700
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4312977099236641,
|
| 532 |
+
"grad_norm": 1.0208545923233032,
|
| 533 |
+
"learning_rate": 0.00011228963091497429,
|
| 534 |
+
"loss": 2.8687,
|
| 535 |
+
"step": 3750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.450381679389313,
|
| 539 |
+
"grad_norm": 0.9706463813781738,
|
| 540 |
+
"learning_rate": 0.00011024245455511344,
|
| 541 |
+
"loss": 2.8615,
|
| 542 |
+
"step": 3800
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4694656488549618,
|
| 546 |
+
"grad_norm": 1.095436692237854,
|
| 547 |
+
"learning_rate": 0.00010819093046429709,
|
| 548 |
+
"loss": 2.8573,
|
| 549 |
+
"step": 3850
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4885496183206106,
|
| 553 |
+
"grad_norm": 1.0057562589645386,
|
| 554 |
+
"learning_rate": 0.00010613592947624463,
|
| 555 |
+
"loss": 2.8409,
|
| 556 |
+
"step": 3900
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.5076335877862594,
|
| 560 |
+
"grad_norm": 0.9920310378074646,
|
| 561 |
+
"learning_rate": 0.00010407832390055341,
|
| 562 |
+
"loss": 2.8454,
|
| 563 |
+
"step": 3950
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.5267175572519083,
|
| 567 |
+
"grad_norm": 1.0131624937057495,
|
| 568 |
+
"learning_rate": 0.00010201898715241975,
|
| 569 |
+
"loss": 2.8413,
|
| 570 |
+
"step": 4000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.545801526717557,
|
| 574 |
+
"grad_norm": 0.9802631139755249,
|
| 575 |
+
"learning_rate": 9.995879338189033e-05,
|
| 576 |
+
"loss": 2.8382,
|
| 577 |
+
"step": 4050
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.5648854961832062,
|
| 581 |
+
"grad_norm": 1.2011436223983765,
|
| 582 |
+
"learning_rate": 9.789861710280183e-05,
|
| 583 |
+
"loss": 2.834,
|
| 584 |
+
"step": 4100
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.583969465648855,
|
| 588 |
+
"grad_norm": 1.0407345294952393,
|
| 589 |
+
"learning_rate": 9.583933282156627e-05,
|
| 590 |
+
"loss": 2.8367,
|
| 591 |
+
"step": 4150
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.6030534351145038,
|
| 595 |
+
"grad_norm": 1.061836838722229,
|
| 596 |
+
"learning_rate": 9.378181466595904e-05,
|
| 597 |
+
"loss": 2.8319,
|
| 598 |
+
"step": 4200
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.6221374045801527,
|
| 602 |
+
"grad_norm": 1.1803492307662964,
|
| 603 |
+
"learning_rate": 9.172693601406825e-05,
|
| 604 |
+
"loss": 2.83,
|
| 605 |
+
"step": 4250
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.6412213740458015,
|
| 609 |
+
"grad_norm": 0.9909766316413879,
|
| 610 |
+
"learning_rate": 8.967556912356194e-05,
|
| 611 |
+
"loss": 2.8223,
|
| 612 |
+
"step": 4300
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6603053435114505,
|
| 616 |
+
"grad_norm": 1.0630844831466675,
|
| 617 |
+
"learning_rate": 8.762858476143076e-05,
|
| 618 |
+
"loss": 2.8267,
|
| 619 |
+
"step": 4350
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6793893129770994,
|
| 623 |
+
"grad_norm": 1.0925393104553223,
|
| 624 |
+
"learning_rate": 8.5586851834364e-05,
|
| 625 |
+
"loss": 2.8149,
|
| 626 |
+
"step": 4400
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6984732824427482,
|
| 630 |
+
"grad_norm": 1.1076596975326538,
|
| 631 |
+
"learning_rate": 8.355123701991448e-05,
|
| 632 |
+
"loss": 2.8101,
|
| 633 |
+
"step": 4450
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.717557251908397,
|
| 637 |
+
"grad_norm": 1.0816969871520996,
|
| 638 |
+
"learning_rate": 8.15226043986105e-05,
|
| 639 |
+
"loss": 2.8125,
|
| 640 |
+
"step": 4500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7366412213740459,
|
| 644 |
+
"grad_norm": 1.03250253200531,
|
| 645 |
+
"learning_rate": 7.950181508716951e-05,
|
| 646 |
+
"loss": 2.8163,
|
| 647 |
+
"step": 4550
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.7557251908396947,
|
| 651 |
+
"grad_norm": 1.0649828910827637,
|
| 652 |
+
"learning_rate": 7.748972687297009e-05,
|
| 653 |
+
"loss": 2.8107,
|
| 654 |
+
"step": 4600
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7748091603053435,
|
| 658 |
+
"grad_norm": 1.0636942386627197,
|
| 659 |
+
"learning_rate": 7.548719384993745e-05,
|
| 660 |
+
"loss": 2.7934,
|
| 661 |
+
"step": 4650
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7938931297709924,
|
| 665 |
+
"grad_norm": 1.0865919589996338,
|
| 666 |
+
"learning_rate": 7.349506605599619e-05,
|
| 667 |
+
"loss": 2.8025,
|
| 668 |
+
"step": 4700
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.8129770992366412,
|
| 672 |
+
"grad_norm": 0.9992050528526306,
|
| 673 |
+
"learning_rate": 7.151418911224533e-05,
|
| 674 |
+
"loss": 2.7822,
|
| 675 |
+
"step": 4750
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.83206106870229,
|
| 679 |
+
"grad_norm": 0.9970654249191284,
|
| 680 |
+
"learning_rate": 6.954540386400779e-05,
|
| 681 |
+
"loss": 2.7934,
|
| 682 |
+
"step": 4800
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.8511450381679388,
|
| 686 |
+
"grad_norm": 1.0189416408538818,
|
| 687 |
+
"learning_rate": 6.758954602390706e-05,
|
| 688 |
+
"loss": 2.7864,
|
| 689 |
+
"step": 4850
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8702290076335877,
|
| 693 |
+
"grad_norm": 0.9974378943443298,
|
| 694 |
+
"learning_rate": 6.564744581712318e-05,
|
| 695 |
+
"loss": 2.7992,
|
| 696 |
+
"step": 4900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8893129770992365,
|
| 700 |
+
"grad_norm": 1.0814576148986816,
|
| 701 |
+
"learning_rate": 6.371992762897722e-05,
|
| 702 |
+
"loss": 2.777,
|
| 703 |
+
"step": 4950
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.9083969465648853,
|
| 707 |
+
"grad_norm": 0.9977588057518005,
|
| 708 |
+
"learning_rate": 6.180780965499535e-05,
|
| 709 |
+
"loss": 2.7959,
|
| 710 |
+
"step": 5000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.9274809160305344,
|
| 714 |
+
"grad_norm": 1.000873327255249,
|
| 715 |
+
"learning_rate": 5.9911903553600046e-05,
|
| 716 |
+
"loss": 2.7776,
|
| 717 |
+
"step": 5050
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9465648854961832,
|
| 721 |
+
"grad_norm": 1.0167657136917114,
|
| 722 |
+
"learning_rate": 5.8033014101576086e-05,
|
| 723 |
+
"loss": 2.7896,
|
| 724 |
+
"step": 5100
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.965648854961832,
|
| 728 |
+
"grad_norm": 0.9922732710838318,
|
| 729 |
+
"learning_rate": 5.617193885245806e-05,
|
| 730 |
+
"loss": 2.7811,
|
| 731 |
+
"step": 5150
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.984732824427481,
|
| 735 |
+
"grad_norm": 1.0540660619735718,
|
| 736 |
+
"learning_rate": 5.4329467797983436e-05,
|
| 737 |
+
"loss": 2.769,
|
| 738 |
+
"step": 5200
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.00381679389313,
|
| 742 |
+
"grad_norm": 0.9783818125724792,
|
| 743 |
+
"learning_rate": 5.250638303275619e-05,
|
| 744 |
+
"loss": 2.7773,
|
| 745 |
+
"step": 5250
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 2.0229007633587788,
|
| 749 |
+
"grad_norm": 1.013471245765686,
|
| 750 |
+
"learning_rate": 5.070345842226194e-05,
|
| 751 |
+
"loss": 2.7324,
|
| 752 |
+
"step": 5300
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.0419847328244276,
|
| 756 |
+
"grad_norm": 1.0684107542037964,
|
| 757 |
+
"learning_rate": 4.8921459274376533e-05,
|
| 758 |
+
"loss": 2.744,
|
| 759 |
+
"step": 5350
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 2.0610687022900764,
|
| 763 |
+
"grad_norm": 1.060261845588684,
|
| 764 |
+
"learning_rate": 4.7161142014507235e-05,
|
| 765 |
+
"loss": 2.7408,
|
| 766 |
+
"step": 5400
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.0801526717557253,
|
| 770 |
+
"grad_norm": 1.0751224756240845,
|
| 771 |
+
"learning_rate": 4.5423253864504134e-05,
|
| 772 |
+
"loss": 2.737,
|
| 773 |
+
"step": 5450
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 2.099236641221374,
|
| 777 |
+
"grad_norm": 1.0697996616363525,
|
| 778 |
+
"learning_rate": 4.370853252547844e-05,
|
| 779 |
+
"loss": 2.7446,
|
| 780 |
+
"step": 5500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.118320610687023,
|
| 784 |
+
"grad_norm": 1.0842689275741577,
|
| 785 |
+
"learning_rate": 4.201770586466215e-05,
|
| 786 |
+
"loss": 2.7539,
|
| 787 |
+
"step": 5550
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 2.1374045801526718,
|
| 791 |
+
"grad_norm": 1.0734447240829468,
|
| 792 |
+
"learning_rate": 4.035149160644193e-05,
|
| 793 |
+
"loss": 2.7467,
|
| 794 |
+
"step": 5600
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.1564885496183206,
|
| 798 |
+
"grad_norm": 1.0258666276931763,
|
| 799 |
+
"learning_rate": 3.8710597027698526e-05,
|
| 800 |
+
"loss": 2.7417,
|
| 801 |
+
"step": 5650
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.1755725190839694,
|
| 805 |
+
"grad_norm": 1.1010756492614746,
|
| 806 |
+
"learning_rate": 3.709571865758091e-05,
|
| 807 |
+
"loss": 2.7322,
|
| 808 |
+
"step": 5700
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.1946564885496183,
|
| 812 |
+
"grad_norm": 1.0063403844833374,
|
| 813 |
+
"learning_rate": 3.550754198184266e-05,
|
| 814 |
+
"loss": 2.7414,
|
| 815 |
+
"step": 5750
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 2.213740458015267,
|
| 819 |
+
"grad_norm": 1.0443757772445679,
|
| 820 |
+
"learning_rate": 3.394674115186601e-05,
|
| 821 |
+
"loss": 2.7341,
|
| 822 |
+
"step": 5800
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.232824427480916,
|
| 826 |
+
"grad_norm": 1.0245685577392578,
|
| 827 |
+
"learning_rate": 3.241397869849717e-05,
|
| 828 |
+
"loss": 2.7306,
|
| 829 |
+
"step": 5850
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 2.2519083969465647,
|
| 833 |
+
"grad_norm": 1.020806074142456,
|
| 834 |
+
"learning_rate": 3.090990525081444e-05,
|
| 835 |
+
"loss": 2.7271,
|
| 836 |
+
"step": 5900
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.2709923664122136,
|
| 840 |
+
"grad_norm": 1.0628166198730469,
|
| 841 |
+
"learning_rate": 2.9435159259948018e-05,
|
| 842 |
+
"loss": 2.7213,
|
| 843 |
+
"step": 5950
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 2.2900763358778624,
|
| 847 |
+
"grad_norm": 1.119207501411438,
|
| 848 |
+
"learning_rate": 2.7990366728069795e-05,
|
| 849 |
+
"loss": 2.7189,
|
| 850 |
+
"step": 6000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.3091603053435117,
|
| 854 |
+
"grad_norm": 1.0679359436035156,
|
| 855 |
+
"learning_rate": 2.6576140942666705e-05,
|
| 856 |
+
"loss": 2.7134,
|
| 857 |
+
"step": 6050
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.32824427480916,
|
| 861 |
+
"grad_norm": 1.0782967805862427,
|
| 862 |
+
"learning_rate": 2.5193082216211972e-05,
|
| 863 |
+
"loss": 2.7258,
|
| 864 |
+
"step": 6100
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.3473282442748094,
|
| 868 |
+
"grad_norm": 1.01475989818573,
|
| 869 |
+
"learning_rate": 2.3841777631343444e-05,
|
| 870 |
+
"loss": 2.7249,
|
| 871 |
+
"step": 6150
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.366412213740458,
|
| 875 |
+
"grad_norm": 1.1221365928649902,
|
| 876 |
+
"learning_rate": 2.2522800791657993e-05,
|
| 877 |
+
"loss": 2.739,
|
| 878 |
+
"step": 6200
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.385496183206107,
|
| 882 |
+
"grad_norm": 1.0444400310516357,
|
| 883 |
+
"learning_rate": 2.1236711578227576e-05,
|
| 884 |
+
"loss": 2.7148,
|
| 885 |
+
"step": 6250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.404580152671756,
|
| 889 |
+
"grad_norm": 1.0502898693084717,
|
| 890 |
+
"learning_rate": 1.9984055911940046e-05,
|
| 891 |
+
"loss": 2.7176,
|
| 892 |
+
"step": 6300
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.4236641221374047,
|
| 896 |
+
"grad_norm": 1.0293673276901245,
|
| 897 |
+
"learning_rate": 1.876536552176601e-05,
|
| 898 |
+
"loss": 2.7195,
|
| 899 |
+
"step": 6350
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.4427480916030535,
|
| 903 |
+
"grad_norm": 1.0429584980010986,
|
| 904 |
+
"learning_rate": 1.758115771904971e-05,
|
| 905 |
+
"loss": 2.7158,
|
| 906 |
+
"step": 6400
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.4618320610687023,
|
| 910 |
+
"grad_norm": 1.0905921459197998,
|
| 911 |
+
"learning_rate": 1.643193517792011e-05,
|
| 912 |
+
"loss": 2.7273,
|
| 913 |
+
"step": 6450
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.480916030534351,
|
| 917 |
+
"grad_norm": 1.0426479578018188,
|
| 918 |
+
"learning_rate": 1.531818572191498e-05,
|
| 919 |
+
"loss": 2.7151,
|
| 920 |
+
"step": 6500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.5,
|
| 924 |
+
"grad_norm": 1.0132908821105957,
|
| 925 |
+
"learning_rate": 1.4240382116909001e-05,
|
| 926 |
+
"loss": 2.7145,
|
| 927 |
+
"step": 6550
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.519083969465649,
|
| 931 |
+
"grad_norm": 1.0216840505599976,
|
| 932 |
+
"learning_rate": 1.3198981870433403e-05,
|
| 933 |
+
"loss": 2.7096,
|
| 934 |
+
"step": 6600
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.5381679389312977,
|
| 938 |
+
"grad_norm": 1.0892080068588257,
|
| 939 |
+
"learning_rate": 1.2194427037472655e-05,
|
| 940 |
+
"loss": 2.7115,
|
| 941 |
+
"step": 6650
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.5572519083969465,
|
| 945 |
+
"grad_norm": 1.0239592790603638,
|
| 946 |
+
"learning_rate": 1.1227144032820147e-05,
|
| 947 |
+
"loss": 2.7066,
|
| 948 |
+
"step": 6700
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.5763358778625953,
|
| 952 |
+
"grad_norm": 1.0963391065597534,
|
| 953 |
+
"learning_rate": 1.0297543450073356e-05,
|
| 954 |
+
"loss": 2.7146,
|
| 955 |
+
"step": 6750
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.595419847328244,
|
| 959 |
+
"grad_norm": 1.0348591804504395,
|
| 960 |
+
"learning_rate": 9.406019887344219e-06,
|
| 961 |
+
"loss": 2.7147,
|
| 962 |
+
"step": 6800
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.614503816793893,
|
| 966 |
+
"grad_norm": 1.0796239376068115,
|
| 967 |
+
"learning_rate": 8.552951779759898e-06,
|
| 968 |
+
"loss": 2.7137,
|
| 969 |
+
"step": 6850
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.633587786259542,
|
| 973 |
+
"grad_norm": 1.083822250366211,
|
| 974 |
+
"learning_rate": 7.738701238823842e-06,
|
| 975 |
+
"loss": 2.7053,
|
| 976 |
+
"step": 6900
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.6526717557251906,
|
| 980 |
+
"grad_norm": 1.020037055015564,
|
| 981 |
+
"learning_rate": 6.963613898706412e-06,
|
| 982 |
+
"loss": 2.6944,
|
| 983 |
+
"step": 6950
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.67175572519084,
|
| 987 |
+
"grad_norm": 0.9994911551475525,
|
| 988 |
+
"learning_rate": 6.228018769529509e-06,
|
| 989 |
+
"loss": 2.7013,
|
| 990 |
+
"step": 7000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.6908396946564883,
|
| 994 |
+
"grad_norm": 1.1002082824707031,
|
| 995 |
+
"learning_rate": 5.5322280977079585e-06,
|
| 996 |
+
"loss": 2.7079,
|
| 997 |
+
"step": 7050
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.7099236641221376,
|
| 1001 |
+
"grad_norm": 1.0165749788284302,
|
| 1002 |
+
"learning_rate": 4.8765372334067085e-06,
|
| 1003 |
+
"loss": 2.7197,
|
| 1004 |
+
"step": 7100
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.7290076335877864,
|
| 1008 |
+
"grad_norm": 1.0475736856460571,
|
| 1009 |
+
"learning_rate": 4.261224505170225e-06,
|
| 1010 |
+
"loss": 2.694,
|
| 1011 |
+
"step": 7150
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.7480916030534353,
|
| 1015 |
+
"grad_norm": 1.0303398370742798,
|
| 1016 |
+
"learning_rate": 3.6865511017771802e-06,
|
| 1017 |
+
"loss": 2.7091,
|
| 1018 |
+
"step": 7200
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.767175572519084,
|
| 1022 |
+
"grad_norm": 1.0317213535308838,
|
| 1023 |
+
"learning_rate": 3.1527609613706932e-06,
|
| 1024 |
+
"loss": 2.7119,
|
| 1025 |
+
"step": 7250
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.786259541984733,
|
| 1029 |
+
"grad_norm": 1.033015489578247,
|
| 1030 |
+
"learning_rate": 2.6600806679111466e-06,
|
| 1031 |
+
"loss": 2.7002,
|
| 1032 |
+
"step": 7300
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.8053435114503817,
|
| 1036 |
+
"grad_norm": 1.0410363674163818,
|
| 1037 |
+
"learning_rate": 2.2087193549955253e-06,
|
| 1038 |
+
"loss": 2.7083,
|
| 1039 |
+
"step": 7350
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.8244274809160306,
|
| 1043 |
+
"grad_norm": 1.023459553718567,
|
| 1044 |
+
"learning_rate": 1.798868617084104e-06,
|
| 1045 |
+
"loss": 2.7121,
|
| 1046 |
+
"step": 7400
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.8435114503816794,
|
| 1050 |
+
"grad_norm": 0.9796963930130005,
|
| 1051 |
+
"learning_rate": 1.4307024281721483e-06,
|
| 1052 |
+
"loss": 2.7165,
|
| 1053 |
+
"step": 7450
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.8625954198473282,
|
| 1057 |
+
"grad_norm": 1.0742990970611572,
|
| 1058 |
+
"learning_rate": 1.1043770679412201e-06,
|
| 1059 |
+
"loss": 2.7109,
|
| 1060 |
+
"step": 7500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.881679389312977,
|
| 1064 |
+
"grad_norm": 1.048072338104248,
|
| 1065 |
+
"learning_rate": 8.200310554213552e-07,
|
| 1066 |
+
"loss": 2.7028,
|
| 1067 |
+
"step": 7550
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.900763358778626,
|
| 1071 |
+
"grad_norm": 1.05283784866333,
|
| 1072 |
+
"learning_rate": 5.777850901923509e-07,
|
| 1073 |
+
"loss": 2.7118,
|
| 1074 |
+
"step": 7600
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.9198473282442747,
|
| 1078 |
+
"grad_norm": 1.0603182315826416,
|
| 1079 |
+
"learning_rate": 3.7774200114898404e-07,
|
| 1080 |
+
"loss": 2.6916,
|
| 1081 |
+
"step": 7650
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.9389312977099236,
|
| 1085 |
+
"grad_norm": 1.0822923183441162,
|
| 1086 |
+
"learning_rate": 2.1998670285212632e-07,
|
| 1087 |
+
"loss": 2.7224,
|
| 1088 |
+
"step": 7700
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.9580152671755724,
|
| 1092 |
+
"grad_norm": 1.0677534341812134,
|
| 1093 |
+
"learning_rate": 1.045861594840436e-07,
|
| 1094 |
+
"loss": 2.6941,
|
| 1095 |
+
"step": 7750
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.9770992366412212,
|
| 1099 |
+
"grad_norm": 1.0649638175964355,
|
| 1100 |
+
"learning_rate": 3.158935642338934e-08,
|
| 1101 |
+
"loss": 2.7043,
|
| 1102 |
+
"step": 7800
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.9961832061068705,
|
| 1106 |
+
"grad_norm": 1.0563445091247559,
|
| 1107 |
+
"learning_rate": 1.0272794518040307e-09,
|
| 1108 |
+
"loss": 2.7146,
|
| 1109 |
+
"step": 7850
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 3.0,
|
| 1113 |
+
"step": 7860,
|
| 1114 |
+
"total_flos": 4871316072038400.0,
|
| 1115 |
+
"train_loss": 3.0463915172120695,
|
| 1116 |
+
"train_runtime": 2406.5677,
|
| 1117 |
+
"train_samples_per_second": 52.251,
|
| 1118 |
+
"train_steps_per_second": 3.266
|
| 1119 |
+
}
|
| 1120 |
+
],
|
| 1121 |
+
"logging_steps": 50,
|
| 1122 |
+
"max_steps": 7860,
|
| 1123 |
+
"num_input_tokens_seen": 0,
|
| 1124 |
+
"num_train_epochs": 3,
|
| 1125 |
+
"save_steps": 2000,
|
| 1126 |
+
"stateful_callbacks": {
|
| 1127 |
+
"TrainerControl": {
|
| 1128 |
+
"args": {
|
| 1129 |
+
"should_epoch_stop": false,
|
| 1130 |
+
"should_evaluate": false,
|
| 1131 |
+
"should_log": false,
|
| 1132 |
+
"should_save": true,
|
| 1133 |
+
"should_training_stop": true
|
| 1134 |
+
},
|
| 1135 |
+
"attributes": {}
|
| 1136 |
+
}
|
| 1137 |
+
},
|
| 1138 |
+
"total_flos": 4871316072038400.0,
|
| 1139 |
+
"train_batch_size": 8,
|
| 1140 |
+
"trial_name": null,
|
| 1141 |
+
"trial_params": null
|
| 1142 |
+
}
|