End of training
Browse files- README.md +2 -2
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +2788 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# qwen3.5-9b-muspsy-fixed
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [Qwen/Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B) on the
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 1.
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
|
|
| 18 |
|
| 19 |
# qwen3.5-9b-muspsy-fixed
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [Qwen/Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B) on the muspsy_task1, the muspsy_task2 and the muspsy_task3_fixed datasets.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 1.2583
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.2582728862762451,
|
| 4 |
+
"eval_runtime": 307.3942,
|
| 5 |
+
"eval_samples_per_second": 3.5,
|
| 6 |
+
"eval_steps_per_second": 3.5,
|
| 7 |
+
"total_flos": 3.5276122661958113e+18,
|
| 8 |
+
"train_loss": 1.2349582789336941,
|
| 9 |
+
"train_runtime": 87806.6728,
|
| 10 |
+
"train_samples_per_second": 0.698,
|
| 11 |
+
"train_steps_per_second": 0.044
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.2582728862762451,
|
| 4 |
+
"eval_runtime": 307.3942,
|
| 5 |
+
"eval_samples_per_second": 3.5,
|
| 6 |
+
"eval_steps_per_second": 3.5
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 3.5276122661958113e+18,
|
| 4 |
+
"train_loss": 1.2349582789336941,
|
| 5 |
+
"train_runtime": 87806.6728,
|
| 6 |
+
"train_samples_per_second": 0.698,
|
| 7 |
+
"train_steps_per_second": 0.044
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2788 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2500,
|
| 3 |
+
"best_metric": 1.2582728862762451,
|
| 4 |
+
"best_model_checkpoint": "saves/qwen3.5-9b/lora/muspsy-fixed/checkpoint-2500",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3831,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.007832770352964214,
|
| 14 |
+
"grad_norm": 0.8252274990081787,
|
| 15 |
+
"learning_rate": 4.6875000000000004e-06,
|
| 16 |
+
"loss": 2.198759651184082,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.01566554070592843,
|
| 21 |
+
"grad_norm": 0.636273980140686,
|
| 22 |
+
"learning_rate": 9.895833333333333e-06,
|
| 23 |
+
"loss": 2.1897178649902345,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.02349831105889264,
|
| 28 |
+
"grad_norm": 0.5016177892684937,
|
| 29 |
+
"learning_rate": 1.5104166666666667e-05,
|
| 30 |
+
"loss": 2.103467559814453,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03133108141185686,
|
| 35 |
+
"grad_norm": 0.2937227189540863,
|
| 36 |
+
"learning_rate": 2.0312500000000002e-05,
|
| 37 |
+
"loss": 1.9080320358276368,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03916385176482107,
|
| 42 |
+
"grad_norm": 0.26669180393218994,
|
| 43 |
+
"learning_rate": 2.552083333333333e-05,
|
| 44 |
+
"loss": 1.7693399429321288,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04699662211778528,
|
| 49 |
+
"grad_norm": 0.2339705526828766,
|
| 50 |
+
"learning_rate": 3.072916666666667e-05,
|
| 51 |
+
"loss": 1.7049299240112306,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0548293924707495,
|
| 56 |
+
"grad_norm": 0.29707223176956177,
|
| 57 |
+
"learning_rate": 3.59375e-05,
|
| 58 |
+
"loss": 1.5845885276794434,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.06266216282371372,
|
| 63 |
+
"grad_norm": 0.2643619477748871,
|
| 64 |
+
"learning_rate": 4.1145833333333335e-05,
|
| 65 |
+
"loss": 1.5671621322631837,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.07049493317667793,
|
| 70 |
+
"grad_norm": 0.2945081889629364,
|
| 71 |
+
"learning_rate": 4.635416666666667e-05,
|
| 72 |
+
"loss": 1.5295079231262207,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07832770352964213,
|
| 77 |
+
"grad_norm": 0.3696126341819763,
|
| 78 |
+
"learning_rate": 5.15625e-05,
|
| 79 |
+
"loss": 1.4586568832397462,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.08616047388260635,
|
| 84 |
+
"grad_norm": 0.26788729429244995,
|
| 85 |
+
"learning_rate": 5.677083333333334e-05,
|
| 86 |
+
"loss": 1.4245469093322753,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09399324423557057,
|
| 91 |
+
"grad_norm": 0.33440881967544556,
|
| 92 |
+
"learning_rate": 6.197916666666667e-05,
|
| 93 |
+
"loss": 1.473775577545166,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10182601458853478,
|
| 98 |
+
"grad_norm": 0.29041001200675964,
|
| 99 |
+
"learning_rate": 6.71875e-05,
|
| 100 |
+
"loss": 1.4213193893432616,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.109658784941499,
|
| 105 |
+
"grad_norm": 0.28692570328712463,
|
| 106 |
+
"learning_rate": 7.239583333333334e-05,
|
| 107 |
+
"loss": 1.3905920028686523,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.11749155529446322,
|
| 112 |
+
"grad_norm": 0.3533639907836914,
|
| 113 |
+
"learning_rate": 7.760416666666667e-05,
|
| 114 |
+
"loss": 1.409282112121582,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.12532432564742743,
|
| 119 |
+
"grad_norm": 0.3724885582923889,
|
| 120 |
+
"learning_rate": 8.28125e-05,
|
| 121 |
+
"loss": 1.4083166122436523,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13315709600039163,
|
| 126 |
+
"grad_norm": 0.3462872803211212,
|
| 127 |
+
"learning_rate": 8.802083333333334e-05,
|
| 128 |
+
"loss": 1.363219928741455,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14098986635335586,
|
| 133 |
+
"grad_norm": 0.35332179069519043,
|
| 134 |
+
"learning_rate": 9.322916666666667e-05,
|
| 135 |
+
"loss": 1.35781831741333,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.14882263670632007,
|
| 140 |
+
"grad_norm": 0.309860497713089,
|
| 141 |
+
"learning_rate": 9.84375e-05,
|
| 142 |
+
"loss": 1.3855286598205567,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.15665540705928427,
|
| 147 |
+
"grad_norm": 0.35043853521347046,
|
| 148 |
+
"learning_rate": 0.00010364583333333333,
|
| 149 |
+
"loss": 1.3873944282531738,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1644881774122485,
|
| 154 |
+
"grad_norm": 0.3299640119075775,
|
| 155 |
+
"learning_rate": 0.00010885416666666669,
|
| 156 |
+
"loss": 1.4211556434631347,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.1723209477652127,
|
| 161 |
+
"grad_norm": 0.32519766688346863,
|
| 162 |
+
"learning_rate": 0.0001140625,
|
| 163 |
+
"loss": 1.3977481842041015,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.18015371811817693,
|
| 168 |
+
"grad_norm": 0.2888562083244324,
|
| 169 |
+
"learning_rate": 0.00011927083333333333,
|
| 170 |
+
"loss": 1.3984809875488282,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.18798648847114113,
|
| 175 |
+
"grad_norm": 0.30679890513420105,
|
| 176 |
+
"learning_rate": 0.00012447916666666667,
|
| 177 |
+
"loss": 1.3943490028381347,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.19581925882410536,
|
| 182 |
+
"grad_norm": 0.2974647283554077,
|
| 183 |
+
"learning_rate": 0.0001296875,
|
| 184 |
+
"loss": 1.3833021163940429,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.20365202917706957,
|
| 189 |
+
"grad_norm": 0.31096336245536804,
|
| 190 |
+
"learning_rate": 0.00013489583333333333,
|
| 191 |
+
"loss": 1.3720091819763183,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.21148479953003377,
|
| 196 |
+
"grad_norm": 0.2655313014984131,
|
| 197 |
+
"learning_rate": 0.00014010416666666669,
|
| 198 |
+
"loss": 1.358876609802246,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.219317569882998,
|
| 203 |
+
"grad_norm": 0.305414080619812,
|
| 204 |
+
"learning_rate": 0.00014531250000000002,
|
| 205 |
+
"loss": 1.349748420715332,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.2271503402359622,
|
| 210 |
+
"grad_norm": 0.242662712931633,
|
| 211 |
+
"learning_rate": 0.00015052083333333334,
|
| 212 |
+
"loss": 1.392601203918457,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.23498311058892643,
|
| 217 |
+
"grad_norm": 0.28158220648765564,
|
| 218 |
+
"learning_rate": 0.00015572916666666667,
|
| 219 |
+
"loss": 1.3312394142150878,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.24281588094189063,
|
| 224 |
+
"grad_norm": 0.3027257025241852,
|
| 225 |
+
"learning_rate": 0.0001609375,
|
| 226 |
+
"loss": 1.3067551612854005,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.25064865129485486,
|
| 231 |
+
"grad_norm": 0.2552274465560913,
|
| 232 |
+
"learning_rate": 0.00016614583333333333,
|
| 233 |
+
"loss": 1.3163175582885742,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.25848142164781907,
|
| 238 |
+
"grad_norm": 0.28112491965293884,
|
| 239 |
+
"learning_rate": 0.0001713541666666667,
|
| 240 |
+
"loss": 1.3473979949951171,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.26631419200078327,
|
| 245 |
+
"grad_norm": 0.32117217779159546,
|
| 246 |
+
"learning_rate": 0.00017656250000000002,
|
| 247 |
+
"loss": 1.3740118026733399,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.27414696235374747,
|
| 252 |
+
"grad_norm": 0.2886170744895935,
|
| 253 |
+
"learning_rate": 0.00018177083333333334,
|
| 254 |
+
"loss": 1.3543949127197266,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.2819797327067117,
|
| 259 |
+
"grad_norm": 0.30684173107147217,
|
| 260 |
+
"learning_rate": 0.00018697916666666667,
|
| 261 |
+
"loss": 1.3309515953063964,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.28981250305967593,
|
| 266 |
+
"grad_norm": 0.2990993857383728,
|
| 267 |
+
"learning_rate": 0.0001921875,
|
| 268 |
+
"loss": 1.3382258415222168,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.29764527341264013,
|
| 273 |
+
"grad_norm": 0.256516695022583,
|
| 274 |
+
"learning_rate": 0.00019739583333333333,
|
| 275 |
+
"loss": 1.3612469673156737,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.30547804376560433,
|
| 280 |
+
"grad_norm": 0.2611815929412842,
|
| 281 |
+
"learning_rate": 0.0001999989616914266,
|
| 282 |
+
"loss": 1.3669749259948731,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.31331081411856854,
|
| 287 |
+
"grad_norm": 0.2872208058834076,
|
| 288 |
+
"learning_rate": 0.00019999065535220928,
|
| 289 |
+
"loss": 1.3830884933471679,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.3211435844715328,
|
| 294 |
+
"grad_norm": 0.2765055000782013,
|
| 295 |
+
"learning_rate": 0.0001999740433637345,
|
| 296 |
+
"loss": 1.3454028129577638,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.328976354824497,
|
| 301 |
+
"grad_norm": 0.2273096740245819,
|
| 302 |
+
"learning_rate": 0.0001999491271058647,
|
| 303 |
+
"loss": 1.317532730102539,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3368091251774612,
|
| 308 |
+
"grad_norm": 0.29210448265075684,
|
| 309 |
+
"learning_rate": 0.00019991590864825028,
|
| 310 |
+
"loss": 1.3869080543518066,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3446418955304254,
|
| 315 |
+
"grad_norm": 0.2541140615940094,
|
| 316 |
+
"learning_rate": 0.00019987439075015763,
|
| 317 |
+
"loss": 1.380301284790039,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3524746658833896,
|
| 322 |
+
"grad_norm": 0.21736745536327362,
|
| 323 |
+
"learning_rate": 0.00019982457686024005,
|
| 324 |
+
"loss": 1.3478592872619628,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.36030743623635386,
|
| 329 |
+
"grad_norm": 0.2869614064693451,
|
| 330 |
+
"learning_rate": 0.00019976647111625117,
|
| 331 |
+
"loss": 1.3316679000854492,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.36814020658931806,
|
| 336 |
+
"grad_norm": 0.20401088893413544,
|
| 337 |
+
"learning_rate": 0.00019970007834470125,
|
| 338 |
+
"loss": 1.3329079627990723,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.37597297694228227,
|
| 343 |
+
"grad_norm": 0.23920388519763947,
|
| 344 |
+
"learning_rate": 0.00019962540406045647,
|
| 345 |
+
"loss": 1.292691707611084,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.38380574729524647,
|
| 350 |
+
"grad_norm": 0.2719043791294098,
|
| 351 |
+
"learning_rate": 0.0001995424544662805,
|
| 352 |
+
"loss": 1.3550784111022949,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3916385176482107,
|
| 357 |
+
"grad_norm": 0.31126660108566284,
|
| 358 |
+
"learning_rate": 0.0001994512364523196,
|
| 359 |
+
"loss": 1.3583318710327148,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3916385176482107,
|
| 364 |
+
"eval_loss": 1.338908314704895,
|
| 365 |
+
"eval_runtime": 322.798,
|
| 366 |
+
"eval_samples_per_second": 3.333,
|
| 367 |
+
"eval_steps_per_second": 3.333,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.39947128800117493,
|
| 372 |
+
"grad_norm": 0.2459356188774109,
|
| 373 |
+
"learning_rate": 0.0001993517575955301,
|
| 374 |
+
"loss": 1.3324827194213866,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.40730405835413913,
|
| 379 |
+
"grad_norm": 0.23143388330936432,
|
| 380 |
+
"learning_rate": 0.0001992440261590491,
|
| 381 |
+
"loss": 1.3564526557922363,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.41513682870710333,
|
| 386 |
+
"grad_norm": 0.2303522378206253,
|
| 387 |
+
"learning_rate": 0.00019912805109150808,
|
| 388 |
+
"loss": 1.3334980964660645,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.42296959906006754,
|
| 393 |
+
"grad_norm": 0.257758766412735,
|
| 394 |
+
"learning_rate": 0.00019900384202628955,
|
| 395 |
+
"loss": 1.31835880279541,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.4308023694130318,
|
| 400 |
+
"grad_norm": 0.2381799966096878,
|
| 401 |
+
"learning_rate": 0.00019887140928072695,
|
| 402 |
+
"loss": 1.336206817626953,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.438635139765996,
|
| 407 |
+
"grad_norm": 0.19770245254039764,
|
| 408 |
+
"learning_rate": 0.00019873076385524754,
|
| 409 |
+
"loss": 1.3192202568054199,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.4464679101189602,
|
| 414 |
+
"grad_norm": 0.24335643649101257,
|
| 415 |
+
"learning_rate": 0.00019858191743245883,
|
| 416 |
+
"loss": 1.3167536735534668,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.4543006804719244,
|
| 421 |
+
"grad_norm": 0.24658577144145966,
|
| 422 |
+
"learning_rate": 0.00019842488237617793,
|
| 423 |
+
"loss": 1.356813621520996,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.4621334508248886,
|
| 428 |
+
"grad_norm": 0.2352878302335739,
|
| 429 |
+
"learning_rate": 0.00019825967173040478,
|
| 430 |
+
"loss": 1.325494384765625,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.46996622117785286,
|
| 435 |
+
"grad_norm": 0.2151331603527069,
|
| 436 |
+
"learning_rate": 0.0001980862992182385,
|
| 437 |
+
"loss": 1.3333147048950196,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.47779899153081706,
|
| 442 |
+
"grad_norm": 0.21890704333782196,
|
| 443 |
+
"learning_rate": 0.00019790477924073755,
|
| 444 |
+
"loss": 1.3124448776245117,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.48563176188378127,
|
| 449 |
+
"grad_norm": 0.22006437182426453,
|
| 450 |
+
"learning_rate": 0.00019771512687572366,
|
| 451 |
+
"loss": 1.3134421348571776,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.49346453223674547,
|
| 456 |
+
"grad_norm": 0.21201013028621674,
|
| 457 |
+
"learning_rate": 0.00019751735787652908,
|
| 458 |
+
"loss": 1.3247313499450684,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.5012973025897097,
|
| 463 |
+
"grad_norm": 0.2933153212070465,
|
| 464 |
+
"learning_rate": 0.0001973114886706884,
|
| 465 |
+
"loss": 1.3370563507080078,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.5091300729426739,
|
| 470 |
+
"grad_norm": 0.1902574896812439,
|
| 471 |
+
"learning_rate": 0.00019709753635857368,
|
| 472 |
+
"loss": 1.3317548751831054,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.5169628432956381,
|
| 477 |
+
"grad_norm": 0.19506672024726868,
|
| 478 |
+
"learning_rate": 0.00019687551871197428,
|
| 479 |
+
"loss": 1.3153909683227538,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.5247956136486024,
|
| 484 |
+
"grad_norm": 0.23361273109912872,
|
| 485 |
+
"learning_rate": 0.00019664545417262056,
|
| 486 |
+
"loss": 1.3147119522094726,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.5326283840015665,
|
| 491 |
+
"grad_norm": 0.2266688495874405,
|
| 492 |
+
"learning_rate": 0.00019640736185065193,
|
| 493 |
+
"loss": 1.3157793998718261,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.5404611543545308,
|
| 498 |
+
"grad_norm": 0.2243291437625885,
|
| 499 |
+
"learning_rate": 0.00019616126152302975,
|
| 500 |
+
"loss": 1.3172315597534179,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.5482939247074949,
|
| 505 |
+
"grad_norm": 0.24645571410655975,
|
| 506 |
+
"learning_rate": 0.0001959071736318942,
|
| 507 |
+
"loss": 1.3153032302856444,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.5561266950604592,
|
| 512 |
+
"grad_norm": 0.19903017580509186,
|
| 513 |
+
"learning_rate": 0.00019564511928286664,
|
| 514 |
+
"loss": 1.29284610748291,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.5639594654134235,
|
| 519 |
+
"grad_norm": 0.23710037767887115,
|
| 520 |
+
"learning_rate": 0.00019537512024329621,
|
| 521 |
+
"loss": 1.3206818580627442,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.5717922357663876,
|
| 526 |
+
"grad_norm": 0.2107548862695694,
|
| 527 |
+
"learning_rate": 0.00019509719894045193,
|
| 528 |
+
"loss": 1.3023270606994628,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.5796250061193519,
|
| 533 |
+
"grad_norm": 0.20591115951538086,
|
| 534 |
+
"learning_rate": 0.0001948113784596596,
|
| 535 |
+
"loss": 1.3201650619506835,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.587457776472316,
|
| 540 |
+
"grad_norm": 0.23854319751262665,
|
| 541 |
+
"learning_rate": 0.00019451768254238446,
|
| 542 |
+
"loss": 1.290597915649414,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.5952905468252803,
|
| 547 |
+
"grad_norm": 1.604860782623291,
|
| 548 |
+
"learning_rate": 0.00019421613558425892,
|
| 549 |
+
"loss": 1.332386589050293,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.6031233171782445,
|
| 554 |
+
"grad_norm": 0.2258477807044983,
|
| 555 |
+
"learning_rate": 0.00019390676263305635,
|
| 556 |
+
"loss": 1.322896671295166,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.6109560875312087,
|
| 561 |
+
"grad_norm": 0.223722904920578,
|
| 562 |
+
"learning_rate": 0.00019358958938661035,
|
| 563 |
+
"loss": 1.3295950889587402,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.6187888578841729,
|
| 568 |
+
"grad_norm": 0.22754625976085663,
|
| 569 |
+
"learning_rate": 0.0001932646421906802,
|
| 570 |
+
"loss": 1.3002716064453126,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.6266216282371371,
|
| 575 |
+
"grad_norm": 0.22892414033412933,
|
| 576 |
+
"learning_rate": 0.00019293194803676254,
|
| 577 |
+
"loss": 1.3059702873229981,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.6344543985901013,
|
| 582 |
+
"grad_norm": 0.23315167427062988,
|
| 583 |
+
"learning_rate": 0.00019259153455984925,
|
| 584 |
+
"loss": 1.3296985626220703,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.6422871689430656,
|
| 589 |
+
"grad_norm": 0.21056416630744934,
|
| 590 |
+
"learning_rate": 0.00019224343003613212,
|
| 591 |
+
"loss": 1.3305760383605958,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.6501199392960297,
|
| 596 |
+
"grad_norm": 0.2409346103668213,
|
| 597 |
+
"learning_rate": 0.0001918876633806539,
|
| 598 |
+
"loss": 1.325657844543457,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.657952709648994,
|
| 603 |
+
"grad_norm": 0.20294003188610077,
|
| 604 |
+
"learning_rate": 0.00019152426414490665,
|
| 605 |
+
"loss": 1.3022818565368652,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.6657854800019581,
|
| 610 |
+
"grad_norm": 0.24989593029022217,
|
| 611 |
+
"learning_rate": 0.00019115326251437708,
|
| 612 |
+
"loss": 1.3478684425354004,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.6736182503549224,
|
| 617 |
+
"grad_norm": 0.2169850468635559,
|
| 618 |
+
"learning_rate": 0.00019077468930603903,
|
| 619 |
+
"loss": 1.3089423179626465,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.6814510207078867,
|
| 624 |
+
"grad_norm": 0.22551028430461884,
|
| 625 |
+
"learning_rate": 0.0001903885759657939,
|
| 626 |
+
"loss": 1.2900199890136719,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.6892837910608508,
|
| 631 |
+
"grad_norm": 0.21751883625984192,
|
| 632 |
+
"learning_rate": 0.00018999495456585854,
|
| 633 |
+
"loss": 1.3203692436218262,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.6971165614138151,
|
| 638 |
+
"grad_norm": 0.22999724745750427,
|
| 639 |
+
"learning_rate": 0.00018959385780210108,
|
| 640 |
+
"loss": 1.3258060455322265,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.7049493317667792,
|
| 645 |
+
"grad_norm": 0.17664238810539246,
|
| 646 |
+
"learning_rate": 0.00018918531899132527,
|
| 647 |
+
"loss": 1.3001131057739257,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.7127821021197435,
|
| 652 |
+
"grad_norm": 0.24631066620349884,
|
| 653 |
+
"learning_rate": 0.00018876937206850294,
|
| 654 |
+
"loss": 1.318901824951172,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.7206148724727077,
|
| 659 |
+
"grad_norm": 0.2125585526227951,
|
| 660 |
+
"learning_rate": 0.00018834605158395514,
|
| 661 |
+
"loss": 1.2912470817565918,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.7284476428256719,
|
| 666 |
+
"grad_norm": 0.20649491250514984,
|
| 667 |
+
"learning_rate": 0.00018791539270048252,
|
| 668 |
+
"loss": 1.3023160934448241,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.7362804131786361,
|
| 673 |
+
"grad_norm": 0.19381368160247803,
|
| 674 |
+
"learning_rate": 0.00018747743119044405,
|
| 675 |
+
"loss": 1.2719347953796387,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.7441131835316004,
|
| 680 |
+
"grad_norm": 0.21463167667388916,
|
| 681 |
+
"learning_rate": 0.00018703220343278628,
|
| 682 |
+
"loss": 1.3201501846313477,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.7519459538845645,
|
| 687 |
+
"grad_norm": 0.2531944811344147,
|
| 688 |
+
"learning_rate": 0.00018657974641002095,
|
| 689 |
+
"loss": 1.2843191146850585,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.7597787242375288,
|
| 694 |
+
"grad_norm": 0.2112455517053604,
|
| 695 |
+
"learning_rate": 0.0001861200977051535,
|
| 696 |
+
"loss": 1.2866954803466797,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.7676114945904929,
|
| 701 |
+
"grad_norm": 0.23256352543830872,
|
| 702 |
+
"learning_rate": 0.0001856532954985609,
|
| 703 |
+
"loss": 1.3130477905273437,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.7754442649434572,
|
| 708 |
+
"grad_norm": 0.21679829061031342,
|
| 709 |
+
"learning_rate": 0.00018517937856482055,
|
| 710 |
+
"loss": 1.3176883697509765,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.7832770352964215,
|
| 715 |
+
"grad_norm": 0.1952638328075409,
|
| 716 |
+
"learning_rate": 0.00018469838626948928,
|
| 717 |
+
"loss": 1.2918599128723145,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.7832770352964215,
|
| 722 |
+
"eval_loss": 1.3027687072753906,
|
| 723 |
+
"eval_runtime": 331.2141,
|
| 724 |
+
"eval_samples_per_second": 3.249,
|
| 725 |
+
"eval_steps_per_second": 3.249,
|
| 726 |
+
"step": 1000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.7911098056493856,
|
| 730 |
+
"grad_norm": 0.20904706418514252,
|
| 731 |
+
"learning_rate": 0.00018421035856583363,
|
| 732 |
+
"loss": 1.3271790504455567,
|
| 733 |
+
"step": 1010
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.7989425760023499,
|
| 737 |
+
"grad_norm": 0.22483959794044495,
|
| 738 |
+
"learning_rate": 0.00018371533599151105,
|
| 739 |
+
"loss": 1.277155590057373,
|
| 740 |
+
"step": 1020
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.806775346355314,
|
| 744 |
+
"grad_norm": 0.20454975962638855,
|
| 745 |
+
"learning_rate": 0.00018321335966520265,
|
| 746 |
+
"loss": 1.297370147705078,
|
| 747 |
+
"step": 1030
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.8146081167082783,
|
| 751 |
+
"grad_norm": 0.2571241855621338,
|
| 752 |
+
"learning_rate": 0.00018270447128319788,
|
| 753 |
+
"loss": 1.3238384246826171,
|
| 754 |
+
"step": 1040
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.8224408870612425,
|
| 758 |
+
"grad_norm": 0.22101953625679016,
|
| 759 |
+
"learning_rate": 0.00018218871311593085,
|
| 760 |
+
"loss": 1.369068431854248,
|
| 761 |
+
"step": 1050
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.8302736574142067,
|
| 765 |
+
"grad_norm": 0.22725796699523926,
|
| 766 |
+
"learning_rate": 0.0001816661280044693,
|
| 767 |
+
"loss": 1.3287213325500489,
|
| 768 |
+
"step": 1060
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.8381064277671709,
|
| 772 |
+
"grad_norm": 0.2206043154001236,
|
| 773 |
+
"learning_rate": 0.00018113675935695599,
|
| 774 |
+
"loss": 1.3072173118591308,
|
| 775 |
+
"step": 1070
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.8459391981201351,
|
| 779 |
+
"grad_norm": 0.19849057495594025,
|
| 780 |
+
"learning_rate": 0.0001806006511450031,
|
| 781 |
+
"loss": 1.302993392944336,
|
| 782 |
+
"step": 1080
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.8537719684730993,
|
| 786 |
+
"grad_norm": 0.21485477685928345,
|
| 787 |
+
"learning_rate": 0.00018005784790003956,
|
| 788 |
+
"loss": 1.3502599716186523,
|
| 789 |
+
"step": 1090
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.8616047388260636,
|
| 793 |
+
"grad_norm": 0.23242683708667755,
|
| 794 |
+
"learning_rate": 0.00017950839470961245,
|
| 795 |
+
"loss": 1.2745637893676758,
|
| 796 |
+
"step": 1100
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.8694375091790277,
|
| 800 |
+
"grad_norm": 0.20469102263450623,
|
| 801 |
+
"learning_rate": 0.0001789523372136414,
|
| 802 |
+
"loss": 1.3007002830505372,
|
| 803 |
+
"step": 1110
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.877270279531992,
|
| 807 |
+
"grad_norm": 0.18739371001720428,
|
| 808 |
+
"learning_rate": 0.00017838972160062793,
|
| 809 |
+
"loss": 1.314402961730957,
|
| 810 |
+
"step": 1120
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.8851030498849561,
|
| 814 |
+
"grad_norm": 0.22443655133247375,
|
| 815 |
+
"learning_rate": 0.0001778205946038185,
|
| 816 |
+
"loss": 1.2732182502746583,
|
| 817 |
+
"step": 1130
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.8929358202379204,
|
| 821 |
+
"grad_norm": 0.22353816032409668,
|
| 822 |
+
"learning_rate": 0.00017724500349732293,
|
| 823 |
+
"loss": 1.2690745353698731,
|
| 824 |
+
"step": 1140
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.9007685905908847,
|
| 828 |
+
"grad_norm": 0.20211759209632874,
|
| 829 |
+
"learning_rate": 0.00017666299609218745,
|
| 830 |
+
"loss": 1.2555768013000488,
|
| 831 |
+
"step": 1150
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.9086013609438488,
|
| 835 |
+
"grad_norm": 0.20591171085834503,
|
| 836 |
+
"learning_rate": 0.00017607462073242336,
|
| 837 |
+
"loss": 1.2454710960388184,
|
| 838 |
+
"step": 1160
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.9164341312968131,
|
| 842 |
+
"grad_norm": 0.20958608388900757,
|
| 843 |
+
"learning_rate": 0.00017547992629099137,
|
| 844 |
+
"loss": 1.2596508026123048,
|
| 845 |
+
"step": 1170
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.9242669016497772,
|
| 849 |
+
"grad_norm": 0.16288408637046814,
|
| 850 |
+
"learning_rate": 0.000174878962165742,
|
| 851 |
+
"loss": 1.2851850509643554,
|
| 852 |
+
"step": 1180
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.9320996720027415,
|
| 856 |
+
"grad_norm": 0.22015182673931122,
|
| 857 |
+
"learning_rate": 0.00017427177827531235,
|
| 858 |
+
"loss": 1.261205291748047,
|
| 859 |
+
"step": 1190
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.9399324423557057,
|
| 863 |
+
"grad_norm": 0.22604605555534363,
|
| 864 |
+
"learning_rate": 0.00017365842505497975,
|
| 865 |
+
"loss": 1.2355712890625,
|
| 866 |
+
"step": 1200
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.9477652127086699,
|
| 870 |
+
"grad_norm": 0.21393424272537231,
|
| 871 |
+
"learning_rate": 0.00017303895345247225,
|
| 872 |
+
"loss": 1.34449462890625,
|
| 873 |
+
"step": 1210
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.9555979830616341,
|
| 877 |
+
"grad_norm": 0.19904212653636932,
|
| 878 |
+
"learning_rate": 0.00017241341492373678,
|
| 879 |
+
"loss": 1.2960014343261719,
|
| 880 |
+
"step": 1220
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.9634307534145983,
|
| 884 |
+
"grad_norm": 0.21874822676181793,
|
| 885 |
+
"learning_rate": 0.000171781861428665,
|
| 886 |
+
"loss": 1.3022172927856446,
|
| 887 |
+
"step": 1230
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.9712635237675625,
|
| 891 |
+
"grad_norm": 0.21850082278251648,
|
| 892 |
+
"learning_rate": 0.0001711443454267772,
|
| 893 |
+
"loss": 1.2833776473999023,
|
| 894 |
+
"step": 1240
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.9790962941205268,
|
| 898 |
+
"grad_norm": 0.20517125725746155,
|
| 899 |
+
"learning_rate": 0.0001705009198728649,
|
| 900 |
+
"loss": 1.312662696838379,
|
| 901 |
+
"step": 1250
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.9869290644734909,
|
| 905 |
+
"grad_norm": 0.20981578528881073,
|
| 906 |
+
"learning_rate": 0.00016985163821259217,
|
| 907 |
+
"loss": 1.2979387283325194,
|
| 908 |
+
"step": 1260
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.9947618348264552,
|
| 912 |
+
"grad_norm": 0.2150573581457138,
|
| 913 |
+
"learning_rate": 0.00016919655437805617,
|
| 914 |
+
"loss": 1.2593985557556153,
|
| 915 |
+
"step": 1270
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 1.0023498311058892,
|
| 919 |
+
"grad_norm": 0.1821247935295105,
|
| 920 |
+
"learning_rate": 0.00016853572278330737,
|
| 921 |
+
"loss": 1.2760144233703614,
|
| 922 |
+
"step": 1280
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.0101826014588535,
|
| 926 |
+
"grad_norm": 0.20195011794567108,
|
| 927 |
+
"learning_rate": 0.00016786919831982957,
|
| 928 |
+
"loss": 1.2467243194580078,
|
| 929 |
+
"step": 1290
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.0180153718118177,
|
| 933 |
+
"grad_norm": 0.21706004440784454,
|
| 934 |
+
"learning_rate": 0.00016719703635198054,
|
| 935 |
+
"loss": 1.240755558013916,
|
| 936 |
+
"step": 1300
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 1.0258481421647818,
|
| 940 |
+
"grad_norm": 0.22652487456798553,
|
| 941 |
+
"learning_rate": 0.00016651929271239315,
|
| 942 |
+
"loss": 1.2947601318359374,
|
| 943 |
+
"step": 1310
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 1.0336809125177462,
|
| 947 |
+
"grad_norm": 0.21941937506198883,
|
| 948 |
+
"learning_rate": 0.00016583602369733767,
|
| 949 |
+
"loss": 1.2749066352844238,
|
| 950 |
+
"step": 1320
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.0415136828707103,
|
| 954 |
+
"grad_norm": 0.20321761071681976,
|
| 955 |
+
"learning_rate": 0.0001651472860620455,
|
| 956 |
+
"loss": 1.2467523574829102,
|
| 957 |
+
"step": 1330
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.0493464532236745,
|
| 961 |
+
"grad_norm": 0.19074523448944092,
|
| 962 |
+
"learning_rate": 0.00016445313701599502,
|
| 963 |
+
"loss": 1.2307304382324218,
|
| 964 |
+
"step": 1340
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.0571792235766388,
|
| 968 |
+
"grad_norm": 0.2475343942642212,
|
| 969 |
+
"learning_rate": 0.0001637536342181593,
|
| 970 |
+
"loss": 1.189406967163086,
|
| 971 |
+
"step": 1350
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.065011993929603,
|
| 975 |
+
"grad_norm": 0.23844459652900696,
|
| 976 |
+
"learning_rate": 0.00016304883577221684,
|
| 977 |
+
"loss": 1.2495388984680176,
|
| 978 |
+
"step": 1360
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.0728447642825671,
|
| 982 |
+
"grad_norm": 0.23677049577236176,
|
| 983 |
+
"learning_rate": 0.0001623388002217253,
|
| 984 |
+
"loss": 1.2544000625610352,
|
| 985 |
+
"step": 1370
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.0806775346355315,
|
| 989 |
+
"grad_norm": 0.24953368306159973,
|
| 990 |
+
"learning_rate": 0.0001616235865452584,
|
| 991 |
+
"loss": 1.2669272422790527,
|
| 992 |
+
"step": 1380
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.0885103049884957,
|
| 996 |
+
"grad_norm": 0.21519717574119568,
|
| 997 |
+
"learning_rate": 0.00016090325415150708,
|
| 998 |
+
"loss": 1.2439716339111329,
|
| 999 |
+
"step": 1390
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.0963430753414598,
|
| 1003 |
+
"grad_norm": 0.22504104673862457,
|
| 1004 |
+
"learning_rate": 0.00016017786287434468,
|
| 1005 |
+
"loss": 1.2285042762756349,
|
| 1006 |
+
"step": 1400
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.104175845694424,
|
| 1010 |
+
"grad_norm": 0.22723400592803955,
|
| 1011 |
+
"learning_rate": 0.00015944747296785705,
|
| 1012 |
+
"loss": 1.211160373687744,
|
| 1013 |
+
"step": 1410
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.1120086160473883,
|
| 1017 |
+
"grad_norm": 0.2472873479127884,
|
| 1018 |
+
"learning_rate": 0.0001587121451013373,
|
| 1019 |
+
"loss": 1.2765045166015625,
|
| 1020 |
+
"step": 1420
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 1.1198413864003525,
|
| 1024 |
+
"grad_norm": 0.24824367463588715,
|
| 1025 |
+
"learning_rate": 0.00015797194035424661,
|
| 1026 |
+
"loss": 1.222754669189453,
|
| 1027 |
+
"step": 1430
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 1.1276741567533166,
|
| 1031 |
+
"grad_norm": 0.24168582260608673,
|
| 1032 |
+
"learning_rate": 0.00015722692021114052,
|
| 1033 |
+
"loss": 1.1961838722229003,
|
| 1034 |
+
"step": 1440
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.135506927106281,
|
| 1038 |
+
"grad_norm": 0.2568736970424652,
|
| 1039 |
+
"learning_rate": 0.00015647714655656197,
|
| 1040 |
+
"loss": 1.2565688133239745,
|
| 1041 |
+
"step": 1450
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.1433396974592451,
|
| 1045 |
+
"grad_norm": 0.2765266001224518,
|
| 1046 |
+
"learning_rate": 0.00015572268166990066,
|
| 1047 |
+
"loss": 1.2159022331237792,
|
| 1048 |
+
"step": 1460
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.1511724678122093,
|
| 1052 |
+
"grad_norm": 0.22696787118911743,
|
| 1053 |
+
"learning_rate": 0.00015496358822022016,
|
| 1054 |
+
"loss": 1.225893497467041,
|
| 1055 |
+
"step": 1470
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.1590052381651734,
|
| 1059 |
+
"grad_norm": 0.20794737339019775,
|
| 1060 |
+
"learning_rate": 0.00015419992926105192,
|
| 1061 |
+
"loss": 1.2634332656860352,
|
| 1062 |
+
"step": 1480
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 1.1668380085181378,
|
| 1066 |
+
"grad_norm": 0.24258407950401306,
|
| 1067 |
+
"learning_rate": 0.0001534317682251583,
|
| 1068 |
+
"loss": 1.22896146774292,
|
| 1069 |
+
"step": 1490
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 1.174670778871102,
|
| 1073 |
+
"grad_norm": 0.2493806779384613,
|
| 1074 |
+
"learning_rate": 0.00015265916891926313,
|
| 1075 |
+
"loss": 1.209363079071045,
|
| 1076 |
+
"step": 1500
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.174670778871102,
|
| 1080 |
+
"eval_loss": 1.2873718738555908,
|
| 1081 |
+
"eval_runtime": 325.3065,
|
| 1082 |
+
"eval_samples_per_second": 3.308,
|
| 1083 |
+
"eval_steps_per_second": 3.308,
|
| 1084 |
+
"step": 1500
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 1.1825035492240663,
|
| 1088 |
+
"grad_norm": 0.23888343572616577,
|
| 1089 |
+
"learning_rate": 0.0001518821955187519,
|
| 1090 |
+
"loss": 1.226218032836914,
|
| 1091 |
+
"step": 1510
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.1903363195770305,
|
| 1095 |
+
"grad_norm": 0.25171080231666565,
|
| 1096 |
+
"learning_rate": 0.00015110091256234097,
|
| 1097 |
+
"loss": 1.244037914276123,
|
| 1098 |
+
"step": 1520
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.1981690899299946,
|
| 1102 |
+
"grad_norm": 0.23836006224155426,
|
| 1103 |
+
"learning_rate": 0.00015031538494671685,
|
| 1104 |
+
"loss": 1.2562739372253418,
|
| 1105 |
+
"step": 1530
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.2060018602829587,
|
| 1109 |
+
"grad_norm": 0.24001693725585938,
|
| 1110 |
+
"learning_rate": 0.00014952567792114534,
|
| 1111 |
+
"loss": 1.2565048217773438,
|
| 1112 |
+
"step": 1540
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.2138346306359231,
|
| 1116 |
+
"grad_norm": 0.253283828496933,
|
| 1117 |
+
"learning_rate": 0.00014873185708205196,
|
| 1118 |
+
"loss": 1.2531654357910156,
|
| 1119 |
+
"step": 1550
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.2216674009888873,
|
| 1123 |
+
"grad_norm": 0.24608942866325378,
|
| 1124 |
+
"learning_rate": 0.00014793398836757297,
|
| 1125 |
+
"loss": 1.2992531776428222,
|
| 1126 |
+
"step": 1560
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.2295001713418514,
|
| 1130 |
+
"grad_norm": 0.28155362606048584,
|
| 1131 |
+
"learning_rate": 0.0001471321380520785,
|
| 1132 |
+
"loss": 1.2777346611022948,
|
| 1133 |
+
"step": 1570
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.2373329416948158,
|
| 1137 |
+
"grad_norm": 0.20141169428825378,
|
| 1138 |
+
"learning_rate": 0.00014632637274066728,
|
| 1139 |
+
"loss": 1.2318084716796875,
|
| 1140 |
+
"step": 1580
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 1.24516571204778,
|
| 1144 |
+
"grad_norm": 0.25557416677474976,
|
| 1145 |
+
"learning_rate": 0.00014551675936363433,
|
| 1146 |
+
"loss": 1.2323214530944824,
|
| 1147 |
+
"step": 1590
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 1.252998482400744,
|
| 1151 |
+
"grad_norm": 0.21658560633659363,
|
| 1152 |
+
"learning_rate": 0.0001447033651709114,
|
| 1153 |
+
"loss": 1.2233482360839845,
|
| 1154 |
+
"step": 1600
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.2608312527537082,
|
| 1158 |
+
"grad_norm": 0.2324075996875763,
|
| 1159 |
+
"learning_rate": 0.00014388625772648078,
|
| 1160 |
+
"loss": 1.2236601829528808,
|
| 1161 |
+
"step": 1610
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.2686640231066726,
|
| 1165 |
+
"grad_norm": 0.23513862490653992,
|
| 1166 |
+
"learning_rate": 0.00014306550490276334,
|
| 1167 |
+
"loss": 1.2506739616394043,
|
| 1168 |
+
"step": 1620
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 1.2764967934596367,
|
| 1172 |
+
"grad_norm": 0.2221163809299469,
|
| 1173 |
+
"learning_rate": 0.00014224117487498055,
|
| 1174 |
+
"loss": 1.2314607620239257,
|
| 1175 |
+
"step": 1630
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.284329563812601,
|
| 1179 |
+
"grad_norm": 0.21652808785438538,
|
| 1180 |
+
"learning_rate": 0.00014141333611549187,
|
| 1181 |
+
"loss": 1.1632638931274415,
|
| 1182 |
+
"step": 1640
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.2921623341655653,
|
| 1186 |
+
"grad_norm": 0.24344556033611298,
|
| 1187 |
+
"learning_rate": 0.00014058205738810668,
|
| 1188 |
+
"loss": 1.2199566841125489,
|
| 1189 |
+
"step": 1650
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 1.2999951045185294,
|
| 1193 |
+
"grad_norm": 0.2773839831352234,
|
| 1194 |
+
"learning_rate": 0.00013974740774237295,
|
| 1195 |
+
"loss": 1.2052583694458008,
|
| 1196 |
+
"step": 1660
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.3078278748714935,
|
| 1200 |
+
"grad_norm": 0.2699849307537079,
|
| 1201 |
+
"learning_rate": 0.00013890945650784137,
|
| 1202 |
+
"loss": 1.216574001312256,
|
| 1203 |
+
"step": 1670
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.3156606452244577,
|
| 1207 |
+
"grad_norm": 0.2651710510253906,
|
| 1208 |
+
"learning_rate": 0.00013806827328830665,
|
| 1209 |
+
"loss": 1.257701587677002,
|
| 1210 |
+
"step": 1680
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.323493415577422,
|
| 1214 |
+
"grad_norm": 0.21589776873588562,
|
| 1215 |
+
"learning_rate": 0.00013722392795602594,
|
| 1216 |
+
"loss": 1.210690402984619,
|
| 1217 |
+
"step": 1690
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.3313261859303862,
|
| 1221 |
+
"grad_norm": 0.2312076985836029,
|
| 1222 |
+
"learning_rate": 0.00013637649064591488,
|
| 1223 |
+
"loss": 1.2231233596801758,
|
| 1224 |
+
"step": 1700
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 1.3391589562833506,
|
| 1228 |
+
"grad_norm": 0.19364634156227112,
|
| 1229 |
+
"learning_rate": 0.00013552603174972204,
|
| 1230 |
+
"loss": 1.2308096885681152,
|
| 1231 |
+
"step": 1710
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 1.3469917266363147,
|
| 1235 |
+
"grad_norm": 0.24895979464054108,
|
| 1236 |
+
"learning_rate": 0.00013467262191018176,
|
| 1237 |
+
"loss": 1.2260958671569824,
|
| 1238 |
+
"step": 1720
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.3548244969892789,
|
| 1242 |
+
"grad_norm": 0.2678626775741577,
|
| 1243 |
+
"learning_rate": 0.00013381633201514615,
|
| 1244 |
+
"loss": 1.2428908348083496,
|
| 1245 |
+
"step": 1730
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.362657267342243,
|
| 1249 |
+
"grad_norm": 0.30587542057037354,
|
| 1250 |
+
"learning_rate": 0.0001329572331916972,
|
| 1251 |
+
"loss": 1.220828914642334,
|
| 1252 |
+
"step": 1740
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 1.3704900376952074,
|
| 1256 |
+
"grad_norm": 0.2133464366197586,
|
| 1257 |
+
"learning_rate": 0.00013209539680023827,
|
| 1258 |
+
"loss": 1.2177210807800294,
|
| 1259 |
+
"step": 1750
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.3783228080481715,
|
| 1263 |
+
"grad_norm": 0.2186136543750763,
|
| 1264 |
+
"learning_rate": 0.00013123089442856686,
|
| 1265 |
+
"loss": 1.1930458068847656,
|
| 1266 |
+
"step": 1760
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.3861555784011357,
|
| 1270 |
+
"grad_norm": 0.22868897020816803,
|
| 1271 |
+
"learning_rate": 0.00013036379788592813,
|
| 1272 |
+
"loss": 1.2323189735412599,
|
| 1273 |
+
"step": 1770
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 1.3939883487541,
|
| 1277 |
+
"grad_norm": 0.2808184027671814,
|
| 1278 |
+
"learning_rate": 0.00012949417919705007,
|
| 1279 |
+
"loss": 1.238790512084961,
|
| 1280 |
+
"step": 1780
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.4018211191070642,
|
| 1284 |
+
"grad_norm": 0.25862979888916016,
|
| 1285 |
+
"learning_rate": 0.00012862211059616087,
|
| 1286 |
+
"loss": 1.2346102714538574,
|
| 1287 |
+
"step": 1790
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.4096538894600283,
|
| 1291 |
+
"grad_norm": 0.23017963767051697,
|
| 1292 |
+
"learning_rate": 0.0001277476645209889,
|
| 1293 |
+
"loss": 1.2439691543579101,
|
| 1294 |
+
"step": 1800
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 1.4174866598129925,
|
| 1298 |
+
"grad_norm": 0.2433658391237259,
|
| 1299 |
+
"learning_rate": 0.00012687091360674558,
|
| 1300 |
+
"loss": 1.2119647026062013,
|
| 1301 |
+
"step": 1810
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.4253194301659569,
|
| 1305 |
+
"grad_norm": 0.23380210995674133,
|
| 1306 |
+
"learning_rate": 0.0001259919306800921,
|
| 1307 |
+
"loss": 1.1757728576660156,
|
| 1308 |
+
"step": 1820
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 1.433152200518921,
|
| 1312 |
+
"grad_norm": 0.255401074886322,
|
| 1313 |
+
"learning_rate": 0.0001251107887530901,
|
| 1314 |
+
"loss": 1.2686352729797363,
|
| 1315 |
+
"step": 1830
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.4409849708718854,
|
| 1319 |
+
"grad_norm": 0.26713621616363525,
|
| 1320 |
+
"learning_rate": 0.00012422756101713702,
|
| 1321 |
+
"loss": 1.2261395454406738,
|
| 1322 |
+
"step": 1840
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 1.4488177412248495,
|
| 1326 |
+
"grad_norm": 0.22357797622680664,
|
| 1327 |
+
"learning_rate": 0.00012334232083688633,
|
| 1328 |
+
"loss": 1.2072179794311524,
|
| 1329 |
+
"step": 1850
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 1.4566505115778137,
|
| 1333 |
+
"grad_norm": 0.22742509841918945,
|
| 1334 |
+
"learning_rate": 0.0001224551417441539,
|
| 1335 |
+
"loss": 1.215613555908203,
|
| 1336 |
+
"step": 1860
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 1.4644832819307778,
|
| 1340 |
+
"grad_norm": 0.2523942291736603,
|
| 1341 |
+
"learning_rate": 0.00012156609743180969,
|
| 1342 |
+
"loss": 1.203711700439453,
|
| 1343 |
+
"step": 1870
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.4723160522837422,
|
| 1347 |
+
"grad_norm": 0.2069718986749649,
|
| 1348 |
+
"learning_rate": 0.00012067526174765691,
|
| 1349 |
+
"loss": 1.177224349975586,
|
| 1350 |
+
"step": 1880
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.4801488226367063,
|
| 1354 |
+
"grad_norm": 0.29070842266082764,
|
| 1355 |
+
"learning_rate": 0.00011978270868829763,
|
| 1356 |
+
"loss": 1.249673557281494,
|
| 1357 |
+
"step": 1890
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 1.4879815929896705,
|
| 1361 |
+
"grad_norm": 0.2660825848579407,
|
| 1362 |
+
"learning_rate": 0.00011888851239298646,
|
| 1363 |
+
"loss": 1.2102355003356933,
|
| 1364 |
+
"step": 1900
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 1.4958143633426348,
|
| 1368 |
+
"grad_norm": 0.2503930628299713,
|
| 1369 |
+
"learning_rate": 0.00011799274713747219,
|
| 1370 |
+
"loss": 1.2331772804260255,
|
| 1371 |
+
"step": 1910
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 1.503647133695599,
|
| 1375 |
+
"grad_norm": 0.24439479410648346,
|
| 1376 |
+
"learning_rate": 0.00011709548732782806,
|
| 1377 |
+
"loss": 1.225032138824463,
|
| 1378 |
+
"step": 1920
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 1.5114799040485631,
|
| 1382 |
+
"grad_norm": 0.2666963040828705,
|
| 1383 |
+
"learning_rate": 0.00011619680749427142,
|
| 1384 |
+
"loss": 1.197390365600586,
|
| 1385 |
+
"step": 1930
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.5193126744015273,
|
| 1389 |
+
"grad_norm": 0.2462136596441269,
|
| 1390 |
+
"learning_rate": 0.00011529678228497278,
|
| 1391 |
+
"loss": 1.1983027458190918,
|
| 1392 |
+
"step": 1940
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 1.5271454447544917,
|
| 1396 |
+
"grad_norm": 0.2605383098125458,
|
| 1397 |
+
"learning_rate": 0.0001143954864598553,
|
| 1398 |
+
"loss": 1.226210594177246,
|
| 1399 |
+
"step": 1950
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 1.5349782151074558,
|
| 1403 |
+
"grad_norm": 0.2872784435749054,
|
| 1404 |
+
"learning_rate": 0.00011349299488438485,
|
| 1405 |
+
"loss": 1.2140783309936523,
|
| 1406 |
+
"step": 1960
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.5428109854604202,
|
| 1410 |
+
"grad_norm": 0.2374207228422165,
|
| 1411 |
+
"learning_rate": 0.00011258938252335157,
|
| 1412 |
+
"loss": 1.1981720924377441,
|
| 1413 |
+
"step": 1970
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.5506437558133843,
|
| 1417 |
+
"grad_norm": 0.21907652914524078,
|
| 1418 |
+
"learning_rate": 0.00011168472443464261,
|
| 1419 |
+
"loss": 1.2283409118652344,
|
| 1420 |
+
"step": 1980
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.5584765261663485,
|
| 1424 |
+
"grad_norm": 0.26273712515830994,
|
| 1425 |
+
"learning_rate": 0.00011077909576300789,
|
| 1426 |
+
"loss": 1.1771793365478516,
|
| 1427 |
+
"step": 1990
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.5663092965193126,
|
| 1431 |
+
"grad_norm": 0.2728463411331177,
|
| 1432 |
+
"learning_rate": 0.00010987257173381796,
|
| 1433 |
+
"loss": 1.2644289016723633,
|
| 1434 |
+
"step": 2000
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1.5663092965193126,
|
| 1438 |
+
"eval_loss": 1.2753645181655884,
|
| 1439 |
+
"eval_runtime": 332.391,
|
| 1440 |
+
"eval_samples_per_second": 3.237,
|
| 1441 |
+
"eval_steps_per_second": 3.237,
|
| 1442 |
+
"step": 2000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 1.5741420668722768,
|
| 1446 |
+
"grad_norm": 0.2645413279533386,
|
| 1447 |
+
"learning_rate": 0.00010896522764681555,
|
| 1448 |
+
"loss": 1.2775901794433593,
|
| 1449 |
+
"step": 2010
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 1.5819748372252411,
|
| 1453 |
+
"grad_norm": 0.2533264756202698,
|
| 1454 |
+
"learning_rate": 0.00010805713886986106,
|
| 1455 |
+
"loss": 1.1955472946166992,
|
| 1456 |
+
"step": 2020
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 1.5898076075782053,
|
| 1460 |
+
"grad_norm": 0.2823348641395569,
|
| 1461 |
+
"learning_rate": 0.00010714838083267185,
|
| 1462 |
+
"loss": 1.2232410430908203,
|
| 1463 |
+
"step": 2030
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 1.5976403779311696,
|
| 1467 |
+
"grad_norm": 0.2825281023979187,
|
| 1468 |
+
"learning_rate": 0.00010623902902055691,
|
| 1469 |
+
"loss": 1.2160611152648926,
|
| 1470 |
+
"step": 2040
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 1.6054731482841338,
|
| 1474 |
+
"grad_norm": 0.26111525297164917,
|
| 1475 |
+
"learning_rate": 0.00010532915896814672,
|
| 1476 |
+
"loss": 1.2666211128234863,
|
| 1477 |
+
"step": 2050
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 1.613305918637098,
|
| 1481 |
+
"grad_norm": 0.2571956217288971,
|
| 1482 |
+
"learning_rate": 0.000104418846253119,
|
| 1483 |
+
"loss": 1.2443676948547364,
|
| 1484 |
+
"step": 2060
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 1.621138688990062,
|
| 1488 |
+
"grad_norm": 0.24284358322620392,
|
| 1489 |
+
"learning_rate": 0.00010350816648992092,
|
| 1490 |
+
"loss": 1.2544951438903809,
|
| 1491 |
+
"step": 2070
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1.6289714593430262,
|
| 1495 |
+
"grad_norm": 0.2704489529132843,
|
| 1496 |
+
"learning_rate": 0.0001025971953234882,
|
| 1497 |
+
"loss": 1.209574031829834,
|
| 1498 |
+
"step": 2080
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 1.6368042296959906,
|
| 1502 |
+
"grad_norm": 0.26949822902679443,
|
| 1503 |
+
"learning_rate": 0.00010168600842296179,
|
| 1504 |
+
"loss": 1.2597773551940918,
|
| 1505 |
+
"step": 2090
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.644637000048955,
|
| 1509 |
+
"grad_norm": 0.2654898166656494,
|
| 1510 |
+
"learning_rate": 0.00010077468147540239,
|
| 1511 |
+
"loss": 1.2378543853759765,
|
| 1512 |
+
"step": 2100
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 1.6524697704019191,
|
| 1516 |
+
"grad_norm": 0.28754162788391113,
|
| 1517 |
+
"learning_rate": 9.986329017950369e-05,
|
| 1518 |
+
"loss": 1.2315853118896485,
|
| 1519 |
+
"step": 2110
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 1.6603025407548833,
|
| 1523 |
+
"grad_norm": 0.26567715406417847,
|
| 1524 |
+
"learning_rate": 9.895191023930432e-05,
|
| 1525 |
+
"loss": 1.1968579292297363,
|
| 1526 |
+
"step": 2120
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 1.6681353111078474,
|
| 1530 |
+
"grad_norm": 0.23091121017932892,
|
| 1531 |
+
"learning_rate": 9.804061735789972e-05,
|
| 1532 |
+
"loss": 1.1917951583862305,
|
| 1533 |
+
"step": 2130
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1.6759680814608116,
|
| 1537 |
+
"grad_norm": 0.23703314363956451,
|
| 1538 |
+
"learning_rate": 9.712948723115383e-05,
|
| 1539 |
+
"loss": 1.1869622230529786,
|
| 1540 |
+
"step": 2140
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.683800851813776,
|
| 1544 |
+
"grad_norm": 0.2393905222415924,
|
| 1545 |
+
"learning_rate": 9.621859554141152e-05,
|
| 1546 |
+
"loss": 1.2089557647705078,
|
| 1547 |
+
"step": 2150
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1.69163362216674,
|
| 1551 |
+
"grad_norm": 0.2555121183395386,
|
| 1552 |
+
"learning_rate": 9.530801795121201e-05,
|
| 1553 |
+
"loss": 1.2293291091918945,
|
| 1554 |
+
"step": 2160
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.6994663925197044,
|
| 1558 |
+
"grad_norm": 0.23827151954174042,
|
| 1559 |
+
"learning_rate": 9.439783009700418e-05,
|
| 1560 |
+
"loss": 1.2037882804870605,
|
| 1561 |
+
"step": 2170
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 1.7072991628726686,
|
| 1565 |
+
"grad_norm": 0.26843613386154175,
|
| 1566 |
+
"learning_rate": 9.348810758286368e-05,
|
| 1567 |
+
"loss": 1.2214282989501952,
|
| 1568 |
+
"step": 2180
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 1.7151319332256327,
|
| 1572 |
+
"grad_norm": 0.23154932260513306,
|
| 1573 |
+
"learning_rate": 9.25789259742131e-05,
|
| 1574 |
+
"loss": 1.2204715728759765,
|
| 1575 |
+
"step": 2190
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 1.7229647035785969,
|
| 1579 |
+
"grad_norm": 0.2480868548154831,
|
| 1580 |
+
"learning_rate": 9.167036079154504e-05,
|
| 1581 |
+
"loss": 1.2267279624938965,
|
| 1582 |
+
"step": 2200
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 1.730797473931561,
|
| 1586 |
+
"grad_norm": 0.23055171966552734,
|
| 1587 |
+
"learning_rate": 9.076248750414924e-05,
|
| 1588 |
+
"loss": 1.1989973068237305,
|
| 1589 |
+
"step": 2210
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 1.7386302442845254,
|
| 1593 |
+
"grad_norm": 0.27379465103149414,
|
| 1594 |
+
"learning_rate": 8.98553815238436e-05,
|
| 1595 |
+
"loss": 1.2200066566467285,
|
| 1596 |
+
"step": 2220
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 1.7464630146374898,
|
| 1600 |
+
"grad_norm": 0.2760390043258667,
|
| 1601 |
+
"learning_rate": 8.894911819871026e-05,
|
| 1602 |
+
"loss": 1.2310608863830566,
|
| 1603 |
+
"step": 2230
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 1.754295784990454,
|
| 1607 |
+
"grad_norm": 0.22573843598365784,
|
| 1608 |
+
"learning_rate": 8.804377280683683e-05,
|
| 1609 |
+
"loss": 1.2127230644226075,
|
| 1610 |
+
"step": 2240
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 1.762128555343418,
|
| 1614 |
+
"grad_norm": 0.2356061041355133,
|
| 1615 |
+
"learning_rate": 8.713942055006346e-05,
|
| 1616 |
+
"loss": 1.1980902671813964,
|
| 1617 |
+
"step": 2250
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.7699613256963822,
|
| 1621 |
+
"grad_norm": 0.22806137800216675,
|
| 1622 |
+
"learning_rate": 8.62361365477363e-05,
|
| 1623 |
+
"loss": 1.2077306747436523,
|
| 1624 |
+
"step": 2260
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 1.7777940960493464,
|
| 1628 |
+
"grad_norm": 0.2392238974571228,
|
| 1629 |
+
"learning_rate": 8.533399583046769e-05,
|
| 1630 |
+
"loss": 1.2229851722717284,
|
| 1631 |
+
"step": 2270
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 1.7856268664023107,
|
| 1635 |
+
"grad_norm": 0.2972569763660431,
|
| 1636 |
+
"learning_rate": 8.443307333390395e-05,
|
| 1637 |
+
"loss": 1.2483017921447754,
|
| 1638 |
+
"step": 2280
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 1.7934596367552749,
|
| 1642 |
+
"grad_norm": 0.23529693484306335,
|
| 1643 |
+
"learning_rate": 8.35334438925007e-05,
|
| 1644 |
+
"loss": 1.211320972442627,
|
| 1645 |
+
"step": 2290
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.8012924071082392,
|
| 1649 |
+
"grad_norm": 0.24110788106918335,
|
| 1650 |
+
"learning_rate": 8.263518223330697e-05,
|
| 1651 |
+
"loss": 1.2276507377624513,
|
| 1652 |
+
"step": 2300
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 1.8091251774612034,
|
| 1656 |
+
"grad_norm": 0.2633586525917053,
|
| 1657 |
+
"learning_rate": 8.173836296975792e-05,
|
| 1658 |
+
"loss": 1.1740093231201172,
|
| 1659 |
+
"step": 2310
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.8169579478141675,
|
| 1663 |
+
"grad_norm": 0.22816379368305206,
|
| 1664 |
+
"learning_rate": 8.084306059547722e-05,
|
| 1665 |
+
"loss": 1.1880734443664551,
|
| 1666 |
+
"step": 2320
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 1.8247907181671317,
|
| 1670 |
+
"grad_norm": 0.2651647925376892,
|
| 1671 |
+
"learning_rate": 7.994934947808928e-05,
|
| 1672 |
+
"loss": 1.204148769378662,
|
| 1673 |
+
"step": 2330
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 1.8326234885200958,
|
| 1677 |
+
"grad_norm": 0.3007386028766632,
|
| 1678 |
+
"learning_rate": 7.905730385304193e-05,
|
| 1679 |
+
"loss": 1.228445816040039,
|
| 1680 |
+
"step": 2340
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 1.8404562588730602,
|
| 1684 |
+
"grad_norm": 0.28230205178260803,
|
| 1685 |
+
"learning_rate": 7.816699781744011e-05,
|
| 1686 |
+
"loss": 1.2086246490478516,
|
| 1687 |
+
"step": 2350
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 1.8482890292260243,
|
| 1691 |
+
"grad_norm": 0.21306121349334717,
|
| 1692 |
+
"learning_rate": 7.727850532389106e-05,
|
| 1693 |
+
"loss": 1.2209898948669433,
|
| 1694 |
+
"step": 2360
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 1.8561217995789887,
|
| 1698 |
+
"grad_norm": 0.2833939790725708,
|
| 1699 |
+
"learning_rate": 7.639190017436149e-05,
|
| 1700 |
+
"loss": 1.2074974060058594,
|
| 1701 |
+
"step": 2370
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 1.8639545699319529,
|
| 1705 |
+
"grad_norm": 0.2340923398733139,
|
| 1706 |
+
"learning_rate": 7.550725601404733e-05,
|
| 1707 |
+
"loss": 1.1948917388916016,
|
| 1708 |
+
"step": 2380
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 1.871787340284917,
|
| 1712 |
+
"grad_norm": 0.26410800218582153,
|
| 1713 |
+
"learning_rate": 7.462464632525639e-05,
|
| 1714 |
+
"loss": 1.2041044235229492,
|
| 1715 |
+
"step": 2390
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 1.8796201106378811,
|
| 1719 |
+
"grad_norm": 0.3063860535621643,
|
| 1720 |
+
"learning_rate": 7.374414442130458e-05,
|
| 1721 |
+
"loss": 1.1983399391174316,
|
| 1722 |
+
"step": 2400
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 1.8874528809908453,
|
| 1726 |
+
"grad_norm": 0.2856113314628601,
|
| 1727 |
+
"learning_rate": 7.286582344042625e-05,
|
| 1728 |
+
"loss": 1.209029769897461,
|
| 1729 |
+
"step": 2410
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 1.8952856513438097,
|
| 1733 |
+
"grad_norm": 0.26700469851493835,
|
| 1734 |
+
"learning_rate": 7.198975633969902e-05,
|
| 1735 |
+
"loss": 1.1910434722900392,
|
| 1736 |
+
"step": 2420
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 1.903118421696774,
|
| 1740 |
+
"grad_norm": 0.28226250410079956,
|
| 1741 |
+
"learning_rate": 7.111601588898361e-05,
|
| 1742 |
+
"loss": 1.2325517654418945,
|
| 1743 |
+
"step": 2430
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 1.9109511920497382,
|
| 1747 |
+
"grad_norm": 0.25455886125564575,
|
| 1748 |
+
"learning_rate": 7.024467466487934e-05,
|
| 1749 |
+
"loss": 1.2045490264892578,
|
| 1750 |
+
"step": 2440
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 1.9187839624027023,
|
| 1754 |
+
"grad_norm": 0.24883849918842316,
|
| 1755 |
+
"learning_rate": 6.937580504469553e-05,
|
| 1756 |
+
"loss": 1.2177587509155274,
|
| 1757 |
+
"step": 2450
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 1.9266167327556665,
|
| 1761 |
+
"grad_norm": 0.27279818058013916,
|
| 1762 |
+
"learning_rate": 6.850947920043948e-05,
|
| 1763 |
+
"loss": 1.2460152626037597,
|
| 1764 |
+
"step": 2460
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 1.9344495031086306,
|
| 1768 |
+
"grad_norm": 0.2591603398323059,
|
| 1769 |
+
"learning_rate": 6.764576909282177e-05,
|
| 1770 |
+
"loss": 1.1761818885803224,
|
| 1771 |
+
"step": 2470
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 1.942282273461595,
|
| 1775 |
+
"grad_norm": 0.2789348363876343,
|
| 1776 |
+
"learning_rate": 6.678474646527871e-05,
|
| 1777 |
+
"loss": 1.1885599136352538,
|
| 1778 |
+
"step": 2480
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 1.9501150438145591,
|
| 1782 |
+
"grad_norm": 0.2906976342201233,
|
| 1783 |
+
"learning_rate": 6.592648283801308e-05,
|
| 1784 |
+
"loss": 1.2112513542175294,
|
| 1785 |
+
"step": 2490
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 1.9579478141675235,
|
| 1789 |
+
"grad_norm": 0.27513882517814636,
|
| 1790 |
+
"learning_rate": 6.507104950205336e-05,
|
| 1791 |
+
"loss": 1.2198973655700684,
|
| 1792 |
+
"step": 2500
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 1.9579478141675235,
|
| 1796 |
+
"eval_loss": 1.2582728862762451,
|
| 1797 |
+
"eval_runtime": 324.3533,
|
| 1798 |
+
"eval_samples_per_second": 3.317,
|
| 1799 |
+
"eval_steps_per_second": 3.317,
|
| 1800 |
+
"step": 2500
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 1.9657805845204877,
|
| 1804 |
+
"grad_norm": 0.27247440814971924,
|
| 1805 |
+
"learning_rate": 6.421851751333197e-05,
|
| 1806 |
+
"loss": 1.2457717895507812,
|
| 1807 |
+
"step": 2510
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 1.9736133548734518,
|
| 1811 |
+
"grad_norm": 0.3500891923904419,
|
| 1812 |
+
"learning_rate": 6.336895768678314e-05,
|
| 1813 |
+
"loss": 1.2087072372436523,
|
| 1814 |
+
"step": 2520
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 1.981446125226416,
|
| 1818 |
+
"grad_norm": 0.29180240631103516,
|
| 1819 |
+
"learning_rate": 6.252244059046062e-05,
|
| 1820 |
+
"loss": 1.2157578468322754,
|
| 1821 |
+
"step": 2530
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.98927889557938,
|
| 1825 |
+
"grad_norm": 0.2686278820037842,
|
| 1826 |
+
"learning_rate": 6.167903653967606e-05,
|
| 1827 |
+
"loss": 1.2206612586975099,
|
| 1828 |
+
"step": 2540
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 1.9971116659323445,
|
| 1832 |
+
"grad_norm": 0.2538076341152191,
|
| 1833 |
+
"learning_rate": 6.083881559115829e-05,
|
| 1834 |
+
"loss": 1.1985113143920898,
|
| 1835 |
+
"step": 2550
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 2.0046996622117783,
|
| 1839 |
+
"grad_norm": 0.2514108717441559,
|
| 1840 |
+
"learning_rate": 6.00018475372341e-05,
|
| 1841 |
+
"loss": 1.13430757522583,
|
| 1842 |
+
"step": 2560
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 2.012532432564743,
|
| 1846 |
+
"grad_norm": 0.3025277554988861,
|
| 1847 |
+
"learning_rate": 5.91682019000311e-05,
|
| 1848 |
+
"loss": 1.1362914085388183,
|
| 1849 |
+
"step": 2570
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 2.020365202917707,
|
| 1853 |
+
"grad_norm": 0.26406702399253845,
|
| 1854 |
+
"learning_rate": 5.833794792570275e-05,
|
| 1855 |
+
"loss": 1.1333977699279785,
|
| 1856 |
+
"step": 2580
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 2.028197973270671,
|
| 1860 |
+
"grad_norm": 0.2084951251745224,
|
| 1861 |
+
"learning_rate": 5.751115457867653e-05,
|
| 1862 |
+
"loss": 1.111202335357666,
|
| 1863 |
+
"step": 2590
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 2.0360307436236353,
|
| 1867 |
+
"grad_norm": 0.29670166969299316,
|
| 1868 |
+
"learning_rate": 5.668789053592565e-05,
|
| 1869 |
+
"loss": 1.0944523811340332,
|
| 1870 |
+
"step": 2600
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 2.0438635139765995,
|
| 1874 |
+
"grad_norm": 0.3064367175102234,
|
| 1875 |
+
"learning_rate": 5.5868224181264136e-05,
|
| 1876 |
+
"loss": 1.0512296676635742,
|
| 1877 |
+
"step": 2610
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 2.0516962843295636,
|
| 1881 |
+
"grad_norm": 0.2714754641056061,
|
| 1882 |
+
"learning_rate": 5.50522235996668e-05,
|
| 1883 |
+
"loss": 1.1794384956359862,
|
| 1884 |
+
"step": 2620
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 2.0595290546825282,
|
| 1888 |
+
"grad_norm": 0.2643856704235077,
|
| 1889 |
+
"learning_rate": 5.4239956571613715e-05,
|
| 1890 |
+
"loss": 1.125117588043213,
|
| 1891 |
+
"step": 2630
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 2.0673618250354924,
|
| 1895 |
+
"grad_norm": 0.29972824454307556,
|
| 1896 |
+
"learning_rate": 5.343149056746015e-05,
|
| 1897 |
+
"loss": 1.1104560852050782,
|
| 1898 |
+
"step": 2640
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 2.0751945953884565,
|
| 1902 |
+
"grad_norm": 0.3134865462779999,
|
| 1903 |
+
"learning_rate": 5.262689274183215e-05,
|
| 1904 |
+
"loss": 1.1257891654968262,
|
| 1905 |
+
"step": 2650
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 2.0830273657414207,
|
| 1909 |
+
"grad_norm": 0.28377866744995117,
|
| 1910 |
+
"learning_rate": 5.1826229928048363e-05,
|
| 1911 |
+
"loss": 1.1403872489929199,
|
| 1912 |
+
"step": 2660
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 2.090860136094385,
|
| 1916 |
+
"grad_norm": 0.2689083516597748,
|
| 1917 |
+
"learning_rate": 5.1029568632568646e-05,
|
| 1918 |
+
"loss": 1.1211537361145019,
|
| 1919 |
+
"step": 2670
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 2.098692906447349,
|
| 1923 |
+
"grad_norm": 0.27901580929756165,
|
| 1924 |
+
"learning_rate": 5.023697502946969e-05,
|
| 1925 |
+
"loss": 1.1200578689575196,
|
| 1926 |
+
"step": 2680
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 2.106525676800313,
|
| 1930 |
+
"grad_norm": 0.2699909210205078,
|
| 1931 |
+
"learning_rate": 4.9448514954948364e-05,
|
| 1932 |
+
"loss": 1.1365811347961425,
|
| 1933 |
+
"step": 2690
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 2.1143584471532777,
|
| 1937 |
+
"grad_norm": 0.3292345702648163,
|
| 1938 |
+
"learning_rate": 4.8664253901853054e-05,
|
| 1939 |
+
"loss": 1.1002737045288087,
|
| 1940 |
+
"step": 2700
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 2.122191217506242,
|
| 1944 |
+
"grad_norm": 0.3296261727809906,
|
| 1945 |
+
"learning_rate": 4.7884257014243625e-05,
|
| 1946 |
+
"loss": 1.1261274337768554,
|
| 1947 |
+
"step": 2710
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 2.130023987859206,
|
| 1951 |
+
"grad_norm": 0.284611314535141,
|
| 1952 |
+
"learning_rate": 4.710858908198016e-05,
|
| 1953 |
+
"loss": 1.1122213363647462,
|
| 1954 |
+
"step": 2720
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 2.13785675821217,
|
| 1958 |
+
"grad_norm": 0.33854803442955017,
|
| 1959 |
+
"learning_rate": 4.633731453534126e-05,
|
| 1960 |
+
"loss": 1.1449526786804198,
|
| 1961 |
+
"step": 2730
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 2.1456895285651343,
|
| 1965 |
+
"grad_norm": 0.32542046904563904,
|
| 1966 |
+
"learning_rate": 4.557049743967229e-05,
|
| 1967 |
+
"loss": 1.1548108100891112,
|
| 1968 |
+
"step": 2740
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 2.1535222989180984,
|
| 1972 |
+
"grad_norm": 0.28194659948349,
|
| 1973 |
+
"learning_rate": 4.480820149006373e-05,
|
| 1974 |
+
"loss": 1.1169511795043945,
|
| 1975 |
+
"step": 2750
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 2.161355069271063,
|
| 1979 |
+
"grad_norm": 0.2818988859653473,
|
| 1980 |
+
"learning_rate": 4.4050490006060453e-05,
|
| 1981 |
+
"loss": 1.1202062606811523,
|
| 1982 |
+
"step": 2760
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 2.169187839624027,
|
| 1986 |
+
"grad_norm": 0.22729508578777313,
|
| 1987 |
+
"learning_rate": 4.3297425926402115e-05,
|
| 1988 |
+
"loss": 1.126146411895752,
|
| 1989 |
+
"step": 2770
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 2.1770206099769913,
|
| 1993 |
+
"grad_norm": 0.28636181354522705,
|
| 1994 |
+
"learning_rate": 4.254907180379516e-05,
|
| 1995 |
+
"loss": 1.1105897903442383,
|
| 1996 |
+
"step": 2780
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 2.1848533803299555,
|
| 2000 |
+
"grad_norm": 0.34403932094573975,
|
| 2001 |
+
"learning_rate": 4.180548979971701e-05,
|
| 2002 |
+
"loss": 1.1060968399047852,
|
| 2003 |
+
"step": 2790
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 2.1926861506829196,
|
| 2007 |
+
"grad_norm": 0.34418439865112305,
|
| 2008 |
+
"learning_rate": 4.1066741679252597e-05,
|
| 2009 |
+
"loss": 1.131563377380371,
|
| 2010 |
+
"step": 2800
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 2.2005189210358838,
|
| 2014 |
+
"grad_norm": 0.2736617922782898,
|
| 2015 |
+
"learning_rate": 4.033288880596385e-05,
|
| 2016 |
+
"loss": 1.140374755859375,
|
| 2017 |
+
"step": 2810
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 2.208351691388848,
|
| 2021 |
+
"grad_norm": 0.35217058658599854,
|
| 2022 |
+
"learning_rate": 3.960399213679281e-05,
|
| 2023 |
+
"loss": 1.130772876739502,
|
| 2024 |
+
"step": 2820
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 2.2161844617418125,
|
| 2028 |
+
"grad_norm": 0.23825931549072266,
|
| 2029 |
+
"learning_rate": 3.888011221699794e-05,
|
| 2030 |
+
"loss": 1.1130745887756348,
|
| 2031 |
+
"step": 2830
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 2.2240172320947766,
|
| 2035 |
+
"grad_norm": 0.2941889762878418,
|
| 2036 |
+
"learning_rate": 3.816130917512525e-05,
|
| 2037 |
+
"loss": 1.16178617477417,
|
| 2038 |
+
"step": 2840
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 2.231850002447741,
|
| 2042 |
+
"grad_norm": 0.3885084092617035,
|
| 2043 |
+
"learning_rate": 3.744764271801362e-05,
|
| 2044 |
+
"loss": 1.1225702285766601,
|
| 2045 |
+
"step": 2850
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 2.239682772800705,
|
| 2049 |
+
"grad_norm": 0.3275553286075592,
|
| 2050 |
+
"learning_rate": 3.673917212583538e-05,
|
| 2051 |
+
"loss": 1.1144232749938965,
|
| 2052 |
+
"step": 2860
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 2.247515543153669,
|
| 2056 |
+
"grad_norm": 0.3339763283729553,
|
| 2057 |
+
"learning_rate": 3.603595624717221e-05,
|
| 2058 |
+
"loss": 1.1165958404541017,
|
| 2059 |
+
"step": 2870
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 2.2553483135066332,
|
| 2063 |
+
"grad_norm": 0.30861443281173706,
|
| 2064 |
+
"learning_rate": 3.533805349412691e-05,
|
| 2065 |
+
"loss": 1.103284740447998,
|
| 2066 |
+
"step": 2880
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.263181083859598,
|
| 2070 |
+
"grad_norm": 0.2967692017555237,
|
| 2071 |
+
"learning_rate": 3.464552183747147e-05,
|
| 2072 |
+
"loss": 1.1307114601135253,
|
| 2073 |
+
"step": 2890
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 2.271013854212562,
|
| 2077 |
+
"grad_norm": 0.35390040278434753,
|
| 2078 |
+
"learning_rate": 3.395841880183176e-05,
|
| 2079 |
+
"loss": 1.1555585861206055,
|
| 2080 |
+
"step": 2900
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 2.278846624565526,
|
| 2084 |
+
"grad_norm": 0.3211853504180908,
|
| 2085 |
+
"learning_rate": 3.327680146090927e-05,
|
| 2086 |
+
"loss": 1.1091843605041505,
|
| 2087 |
+
"step": 2910
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 2.2866793949184903,
|
| 2091 |
+
"grad_norm": 0.3494303822517395,
|
| 2092 |
+
"learning_rate": 3.260072643274044e-05,
|
| 2093 |
+
"loss": 1.136253547668457,
|
| 2094 |
+
"step": 2920
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 2.2945121652714544,
|
| 2098 |
+
"grad_norm": 0.3768707811832428,
|
| 2099 |
+
"learning_rate": 3.1930249874993514e-05,
|
| 2100 |
+
"loss": 1.1504868507385253,
|
| 2101 |
+
"step": 2930
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 2.3023449356244186,
|
| 2105 |
+
"grad_norm": 0.3536722958087921,
|
| 2106 |
+
"learning_rate": 3.126542748030402e-05,
|
| 2107 |
+
"loss": 1.1369070053100585,
|
| 2108 |
+
"step": 2940
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 2.3101777059773827,
|
| 2112 |
+
"grad_norm": 0.3566012382507324,
|
| 2113 |
+
"learning_rate": 3.0606314471648643e-05,
|
| 2114 |
+
"loss": 1.1335080146789551,
|
| 2115 |
+
"step": 2950
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 2.318010476330347,
|
| 2119 |
+
"grad_norm": 0.3815862834453583,
|
| 2120 |
+
"learning_rate": 2.99529655977582e-05,
|
| 2121 |
+
"loss": 1.079446029663086,
|
| 2122 |
+
"step": 2960
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 2.3258432466833114,
|
| 2126 |
+
"grad_norm": 0.3307647705078125,
|
| 2127 |
+
"learning_rate": 2.9305435128569835e-05,
|
| 2128 |
+
"loss": 1.1216520309448241,
|
| 2129 |
+
"step": 2970
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 2.3336760170362756,
|
| 2133 |
+
"grad_norm": 0.3462022840976715,
|
| 2134 |
+
"learning_rate": 2.8663776850719393e-05,
|
| 2135 |
+
"loss": 1.1272664070129395,
|
| 2136 |
+
"step": 2980
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 2.3415087873892397,
|
| 2140 |
+
"grad_norm": 0.3372374475002289,
|
| 2141 |
+
"learning_rate": 2.8028044063073435e-05,
|
| 2142 |
+
"loss": 1.095858097076416,
|
| 2143 |
+
"step": 2990
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 2.349341557742204,
|
| 2147 |
+
"grad_norm": 0.33324751257896423,
|
| 2148 |
+
"learning_rate": 2.739828957230207e-05,
|
| 2149 |
+
"loss": 1.1110441207885742,
|
| 2150 |
+
"step": 3000
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 2.349341557742204,
|
| 2154 |
+
"eval_loss": 1.2691750526428223,
|
| 2155 |
+
"eval_runtime": 324.8724,
|
| 2156 |
+
"eval_samples_per_second": 3.312,
|
| 2157 |
+
"eval_steps_per_second": 3.312,
|
| 2158 |
+
"step": 3000
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.357174328095168,
|
| 2162 |
+
"grad_norm": 0.2844167649745941,
|
| 2163 |
+
"learning_rate": 2.677456568849277e-05,
|
| 2164 |
+
"loss": 1.0753797531127929,
|
| 2165 |
+
"step": 3010
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.3650070984481326,
|
| 2169 |
+
"grad_norm": 0.3673315644264221,
|
| 2170 |
+
"learning_rate": 2.6156924220804992e-05,
|
| 2171 |
+
"loss": 1.124093723297119,
|
| 2172 |
+
"step": 3020
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.3728398688010968,
|
| 2176 |
+
"grad_norm": 0.30328670144081116,
|
| 2177 |
+
"learning_rate": 2.5545416473166883e-05,
|
| 2178 |
+
"loss": 1.1043742179870606,
|
| 2179 |
+
"step": 3030
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.380672639154061,
|
| 2183 |
+
"grad_norm": 0.29031527042388916,
|
| 2184 |
+
"learning_rate": 2.4940093240013717e-05,
|
| 2185 |
+
"loss": 1.1230106353759766,
|
| 2186 |
+
"step": 3040
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.388505409507025,
|
| 2190 |
+
"grad_norm": 0.3382628262042999,
|
| 2191 |
+
"learning_rate": 2.4341004802068658e-05,
|
| 2192 |
+
"loss": 1.1189377784729004,
|
| 2193 |
+
"step": 3050
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.396338179859989,
|
| 2197 |
+
"grad_norm": 0.3430864214897156,
|
| 2198 |
+
"learning_rate": 2.3748200922166265e-05,
|
| 2199 |
+
"loss": 1.1290170669555664,
|
| 2200 |
+
"step": 3060
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.4041709502129534,
|
| 2204 |
+
"grad_norm": 0.3256888687610626,
|
| 2205 |
+
"learning_rate": 2.3161730841118978e-05,
|
| 2206 |
+
"loss": 1.1303093910217286,
|
| 2207 |
+
"step": 3070
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.4120037205659175,
|
| 2211 |
+
"grad_norm": 0.3369556963443756,
|
| 2212 |
+
"learning_rate": 2.2581643273626928e-05,
|
| 2213 |
+
"loss": 1.1088360786437987,
|
| 2214 |
+
"step": 3080
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.4198364909188816,
|
| 2218 |
+
"grad_norm": 0.32177814841270447,
|
| 2219 |
+
"learning_rate": 2.200798640423156e-05,
|
| 2220 |
+
"loss": 1.0996588706970214,
|
| 2221 |
+
"step": 3090
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.4276692612718462,
|
| 2225 |
+
"grad_norm": 0.3396743834018707,
|
| 2226 |
+
"learning_rate": 2.1440807883313095e-05,
|
| 2227 |
+
"loss": 1.0969182014465333,
|
| 2228 |
+
"step": 3100
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.4355020316248104,
|
| 2232 |
+
"grad_norm": 0.2916245758533478,
|
| 2233 |
+
"learning_rate": 2.088015482313269e-05,
|
| 2234 |
+
"loss": 1.099278736114502,
|
| 2235 |
+
"step": 3110
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.4433348019777745,
|
| 2239 |
+
"grad_norm": 0.3665787875652313,
|
| 2240 |
+
"learning_rate": 2.0326073793918867e-05,
|
| 2241 |
+
"loss": 1.1052253723144532,
|
| 2242 |
+
"step": 3120
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.4511675723307387,
|
| 2246 |
+
"grad_norm": 0.308431476354599,
|
| 2247 |
+
"learning_rate": 1.977861081999931e-05,
|
| 2248 |
+
"loss": 1.1300259590148927,
|
| 2249 |
+
"step": 3130
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.459000342683703,
|
| 2253 |
+
"grad_norm": 0.3617427945137024,
|
| 2254 |
+
"learning_rate": 1.923781137597791e-05,
|
| 2255 |
+
"loss": 1.119156265258789,
|
| 2256 |
+
"step": 3140
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.466833113036667,
|
| 2260 |
+
"grad_norm": 0.4287793040275574,
|
| 2261 |
+
"learning_rate": 1.8703720382957392e-05,
|
| 2262 |
+
"loss": 1.0923386573791505,
|
| 2263 |
+
"step": 3150
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.4746658833896316,
|
| 2267 |
+
"grad_norm": 0.2791413366794586,
|
| 2268 |
+
"learning_rate": 1.8176382204807984e-05,
|
| 2269 |
+
"loss": 1.1367136001586915,
|
| 2270 |
+
"step": 3160
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.4824986537425957,
|
| 2274 |
+
"grad_norm": 0.3458610773086548,
|
| 2275 |
+
"learning_rate": 1.7655840644482403e-05,
|
| 2276 |
+
"loss": 1.13309907913208,
|
| 2277 |
+
"step": 3170
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.49033142409556,
|
| 2281 |
+
"grad_norm": 0.3757402002811432,
|
| 2282 |
+
"learning_rate": 1.714213894037737e-05,
|
| 2283 |
+
"loss": 1.107917308807373,
|
| 2284 |
+
"step": 3180
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.498164194448524,
|
| 2288 |
+
"grad_norm": 0.3789442479610443,
|
| 2289 |
+
"learning_rate": 1.663531976274204e-05,
|
| 2290 |
+
"loss": 1.0986825942993164,
|
| 2291 |
+
"step": 3190
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.505996964801488,
|
| 2295 |
+
"grad_norm": 0.3685928285121918,
|
| 2296 |
+
"learning_rate": 1.6135425210133624e-05,
|
| 2297 |
+
"loss": 1.125806999206543,
|
| 2298 |
+
"step": 3200
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.5138297351544523,
|
| 2302 |
+
"grad_norm": 0.3142971098423004,
|
| 2303 |
+
"learning_rate": 1.5642496805920537e-05,
|
| 2304 |
+
"loss": 1.113065528869629,
|
| 2305 |
+
"step": 3210
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.5216625055074164,
|
| 2309 |
+
"grad_norm": 0.3308829069137573,
|
| 2310 |
+
"learning_rate": 1.515657549483328e-05,
|
| 2311 |
+
"loss": 1.121922206878662,
|
| 2312 |
+
"step": 3220
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.529495275860381,
|
| 2316 |
+
"grad_norm": 0.34523579478263855,
|
| 2317 |
+
"learning_rate": 1.4677701639563368e-05,
|
| 2318 |
+
"loss": 1.1328062057495116,
|
| 2319 |
+
"step": 3230
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.537328046213345,
|
| 2323 |
+
"grad_norm": 0.2806236445903778,
|
| 2324 |
+
"learning_rate": 1.4205915017410654e-05,
|
| 2325 |
+
"loss": 1.1356450080871583,
|
| 2326 |
+
"step": 3240
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.5451608165663093,
|
| 2330 |
+
"grad_norm": 0.2967741787433624,
|
| 2331 |
+
"learning_rate": 1.3741254816979232e-05,
|
| 2332 |
+
"loss": 1.088940715789795,
|
| 2333 |
+
"step": 3250
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.5529935869192735,
|
| 2337 |
+
"grad_norm": 0.35171636939048767,
|
| 2338 |
+
"learning_rate": 1.3283759634922332e-05,
|
| 2339 |
+
"loss": 1.1193583488464356,
|
| 2340 |
+
"step": 3260
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.5608263572722376,
|
| 2344 |
+
"grad_norm": 0.32858696579933167,
|
| 2345 |
+
"learning_rate": 1.2833467472736248e-05,
|
| 2346 |
+
"loss": 1.09879789352417,
|
| 2347 |
+
"step": 3270
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.568659127625202,
|
| 2351 |
+
"grad_norm": 0.34779906272888184,
|
| 2352 |
+
"learning_rate": 1.23904157336038e-05,
|
| 2353 |
+
"loss": 1.126400089263916,
|
| 2354 |
+
"step": 3280
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.5764918979781664,
|
| 2358 |
+
"grad_norm": 0.39374393224716187,
|
| 2359 |
+
"learning_rate": 1.1954641219287466e-05,
|
| 2360 |
+
"loss": 1.16580810546875,
|
| 2361 |
+
"step": 3290
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.5843246683311305,
|
| 2365 |
+
"grad_norm": 0.3365297317504883,
|
| 2366 |
+
"learning_rate": 1.152618012707246e-05,
|
| 2367 |
+
"loss": 1.120203685760498,
|
| 2368 |
+
"step": 3300
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.5921574386840947,
|
| 2372 |
+
"grad_norm": 0.32898035645484924,
|
| 2373 |
+
"learning_rate": 1.1105068046760048e-05,
|
| 2374 |
+
"loss": 1.0829385757446288,
|
| 2375 |
+
"step": 3310
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.599990209037059,
|
| 2379 |
+
"grad_norm": 0.3395327627658844,
|
| 2380 |
+
"learning_rate": 1.0691339957711289e-05,
|
| 2381 |
+
"loss": 1.142617416381836,
|
| 2382 |
+
"step": 3320
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.607822979390023,
|
| 2386 |
+
"grad_norm": 0.32847779989242554,
|
| 2387 |
+
"learning_rate": 1.028503022594155e-05,
|
| 2388 |
+
"loss": 1.110114288330078,
|
| 2389 |
+
"step": 3330
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.615655749742987,
|
| 2393 |
+
"grad_norm": 0.3730526566505432,
|
| 2394 |
+
"learning_rate": 9.886172601265842e-06,
|
| 2395 |
+
"loss": 1.124988079071045,
|
| 2396 |
+
"step": 3340
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.6234885200959512,
|
| 2400 |
+
"grad_norm": 0.3661583662033081,
|
| 2401 |
+
"learning_rate": 9.494800214495459e-06,
|
| 2402 |
+
"loss": 1.1263715744018554,
|
| 2403 |
+
"step": 3350
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.6313212904489154,
|
| 2407 |
+
"grad_norm": 0.36449870467185974,
|
| 2408 |
+
"learning_rate": 9.110945574685992e-06,
|
| 2409 |
+
"loss": 1.0590201377868653,
|
| 2410 |
+
"step": 3360
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.63915406080188,
|
| 2414 |
+
"grad_norm": 0.2658601403236389,
|
| 2415 |
+
"learning_rate": 8.73464056643699e-06,
|
| 2416 |
+
"loss": 1.0944375038146972,
|
| 2417 |
+
"step": 3370
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.646986831154844,
|
| 2421 |
+
"grad_norm": 0.47399842739105225,
|
| 2422 |
+
"learning_rate": 8.365916447243482e-06,
|
| 2423 |
+
"loss": 1.1339784622192384,
|
| 2424 |
+
"step": 3380
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.6548196015078083,
|
| 2428 |
+
"grad_norm": 0.310308575630188,
|
| 2429 |
+
"learning_rate": 8.0048038448996e-06,
|
| 2430 |
+
"loss": 1.134721565246582,
|
| 2431 |
+
"step": 3390
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.6626523718607724,
|
| 2435 |
+
"grad_norm": 0.32272788882255554,
|
| 2436 |
+
"learning_rate": 7.651332754954477e-06,
|
| 2437 |
+
"loss": 1.0940790176391602,
|
| 2438 |
+
"step": 3400
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.6704851422137366,
|
| 2442 |
+
"grad_norm": 0.2760407626628876,
|
| 2443 |
+
"learning_rate": 7.305532538220749e-06,
|
| 2444 |
+
"loss": 1.1149394035339355,
|
| 2445 |
+
"step": 3410
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.678317912566701,
|
| 2449 |
+
"grad_norm": 0.36517298221588135,
|
| 2450 |
+
"learning_rate": 6.967431918335676e-06,
|
| 2451 |
+
"loss": 1.113607692718506,
|
| 2452 |
+
"step": 3420
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.6861506829196653,
|
| 2456 |
+
"grad_norm": 0.3604297935962677,
|
| 2457 |
+
"learning_rate": 6.63705897937521e-06,
|
| 2458 |
+
"loss": 1.1084628105163574,
|
| 2459 |
+
"step": 3430
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.6939834532726294,
|
| 2463 |
+
"grad_norm": 0.3391871452331543,
|
| 2464 |
+
"learning_rate": 6.314441163521334e-06,
|
| 2465 |
+
"loss": 1.109107208251953,
|
| 2466 |
+
"step": 3440
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.7018162236255936,
|
| 2470 |
+
"grad_norm": 0.3342093825340271,
|
| 2471 |
+
"learning_rate": 5.9996052687824045e-06,
|
| 2472 |
+
"loss": 1.147671890258789,
|
| 2473 |
+
"step": 3450
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.7096489939785577,
|
| 2477 |
+
"grad_norm": 0.28410813212394714,
|
| 2478 |
+
"learning_rate": 5.692577446767389e-06,
|
| 2479 |
+
"loss": 1.09722900390625,
|
| 2480 |
+
"step": 3460
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.717481764331522,
|
| 2484 |
+
"grad_norm": 0.3533446490764618,
|
| 2485 |
+
"learning_rate": 5.393383200513447e-06,
|
| 2486 |
+
"loss": 1.1344051361083984,
|
| 2487 |
+
"step": 3470
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.725314534684486,
|
| 2491 |
+
"grad_norm": 0.3195113241672516,
|
| 2492 |
+
"learning_rate": 5.102047382367659e-06,
|
| 2493 |
+
"loss": 1.1147592544555665,
|
| 2494 |
+
"step": 3480
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.73314730503745,
|
| 2498 |
+
"grad_norm": 0.33619624376296997,
|
| 2499 |
+
"learning_rate": 4.818594191922576e-06,
|
| 2500 |
+
"loss": 1.1492552757263184,
|
| 2501 |
+
"step": 3490
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.7409800753904148,
|
| 2505 |
+
"grad_norm": 0.2879767119884491,
|
| 2506 |
+
"learning_rate": 4.543047174006209e-06,
|
| 2507 |
+
"loss": 1.1195759773254395,
|
| 2508 |
+
"step": 3500
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.7409800753904148,
|
| 2512 |
+
"eval_loss": 1.2651914358139038,
|
| 2513 |
+
"eval_runtime": 326.4841,
|
| 2514 |
+
"eval_samples_per_second": 3.296,
|
| 2515 |
+
"eval_steps_per_second": 3.296,
|
| 2516 |
+
"step": 3500
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.748812845743379,
|
| 2520 |
+
"grad_norm": 0.30044785141944885,
|
| 2521 |
+
"learning_rate": 4.275429216726201e-06,
|
| 2522 |
+
"loss": 1.1164446830749513,
|
| 2523 |
+
"step": 3510
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.756645616096343,
|
| 2527 |
+
"grad_norm": 0.3198385536670685,
|
| 2528 |
+
"learning_rate": 4.015762549568724e-06,
|
| 2529 |
+
"loss": 1.0795605659484864,
|
| 2530 |
+
"step": 3520
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 2.764478386449307,
|
| 2534 |
+
"grad_norm": 0.29045701026916504,
|
| 2535 |
+
"learning_rate": 3.7640687415519293e-06,
|
| 2536 |
+
"loss": 1.0812619209289551,
|
| 2537 |
+
"step": 3530
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 2.7723111568022714,
|
| 2541 |
+
"grad_norm": 0.3366210460662842,
|
| 2542 |
+
"learning_rate": 3.520368699434384e-06,
|
| 2543 |
+
"loss": 1.1149346351623535,
|
| 2544 |
+
"step": 3540
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 2.780143927155236,
|
| 2548 |
+
"grad_norm": 0.35936740040779114,
|
| 2549 |
+
"learning_rate": 3.2846826659784334e-06,
|
| 2550 |
+
"loss": 1.099393367767334,
|
| 2551 |
+
"step": 3550
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 2.7879766975082,
|
| 2555 |
+
"grad_norm": 0.32345789670944214,
|
| 2556 |
+
"learning_rate": 3.057030218268764e-06,
|
| 2557 |
+
"loss": 1.1055167198181153,
|
| 2558 |
+
"step": 3560
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 2.7958094678611642,
|
| 2562 |
+
"grad_norm": 0.37679150700569153,
|
| 2563 |
+
"learning_rate": 2.8374302660862826e-06,
|
| 2564 |
+
"loss": 1.1351775169372558,
|
| 2565 |
+
"step": 3570
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 2.8036422382141284,
|
| 2569 |
+
"grad_norm": 0.3136025667190552,
|
| 2570 |
+
"learning_rate": 2.6259010503373206e-06,
|
| 2571 |
+
"loss": 1.0925300598144532,
|
| 2572 |
+
"step": 3580
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 2.8114750085670925,
|
| 2576 |
+
"grad_norm": 0.4157222807407379,
|
| 2577 |
+
"learning_rate": 2.4224601415385294e-06,
|
| 2578 |
+
"loss": 1.0652141571044922,
|
| 2579 |
+
"step": 3590
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 2.8193077789200567,
|
| 2583 |
+
"grad_norm": 0.39427298307418823,
|
| 2584 |
+
"learning_rate": 2.227124438357353e-06,
|
| 2585 |
+
"loss": 1.1031173706054687,
|
| 2586 |
+
"step": 3600
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 2.827140549273021,
|
| 2590 |
+
"grad_norm": 0.38145479559898376,
|
| 2591 |
+
"learning_rate": 2.0399101662083807e-06,
|
| 2592 |
+
"loss": 1.1134588241577148,
|
| 2593 |
+
"step": 3610
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 2.834973319625985,
|
| 2597 |
+
"grad_norm": 0.41683870553970337,
|
| 2598 |
+
"learning_rate": 1.8608328759055716e-06,
|
| 2599 |
+
"loss": 1.094143009185791,
|
| 2600 |
+
"step": 3620
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 2.8428060899789496,
|
| 2604 |
+
"grad_norm": 0.3398721218109131,
|
| 2605 |
+
"learning_rate": 1.6899074423705975e-06,
|
| 2606 |
+
"loss": 1.1329906463623047,
|
| 2607 |
+
"step": 3630
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.8506388603319137,
|
| 2611 |
+
"grad_norm": 0.3933563232421875,
|
| 2612 |
+
"learning_rate": 1.5271480633971768e-06,
|
| 2613 |
+
"loss": 1.0851118087768554,
|
| 2614 |
+
"step": 3640
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 2.858471630684878,
|
| 2618 |
+
"grad_norm": 0.40157073736190796,
|
| 2619 |
+
"learning_rate": 1.372568258471807e-06,
|
| 2620 |
+
"loss": 1.1162568092346192,
|
| 2621 |
+
"step": 3650
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.866304401037842,
|
| 2625 |
+
"grad_norm": 0.336107462644577,
|
| 2626 |
+
"learning_rate": 1.2261808676507724e-06,
|
| 2627 |
+
"loss": 1.0793692588806152,
|
| 2628 |
+
"step": 3660
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 2.874137171390806,
|
| 2632 |
+
"grad_norm": 0.3799566328525543,
|
| 2633 |
+
"learning_rate": 1.0879980504935772e-06,
|
| 2634 |
+
"loss": 1.1028385162353516,
|
| 2635 |
+
"step": 3670
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 2.8819699417437707,
|
| 2639 |
+
"grad_norm": 0.42130786180496216,
|
| 2640 |
+
"learning_rate": 9.580312850529295e-07,
|
| 2641 |
+
"loss": 1.1181939125061036,
|
| 2642 |
+
"step": 3680
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 2.889802712096735,
|
| 2646 |
+
"grad_norm": 0.33488714694976807,
|
| 2647 |
+
"learning_rate": 8.362913669213379e-07,
|
| 2648 |
+
"loss": 1.0907830238342284,
|
| 2649 |
+
"step": 3690
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 2.897635482449699,
|
| 2653 |
+
"grad_norm": 0.3703271448612213,
|
| 2654 |
+
"learning_rate": 7.227884083343406e-07,
|
| 2655 |
+
"loss": 1.080193328857422,
|
| 2656 |
+
"step": 3700
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 2.905468252802663,
|
| 2660 |
+
"grad_norm": 0.35055869817733765,
|
| 2661 |
+
"learning_rate": 6.175318373306205e-07,
|
| 2662 |
+
"loss": 1.0740805625915528,
|
| 2663 |
+
"step": 3710
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 2.9133010231556273,
|
| 2667 |
+
"grad_norm": 0.3170068562030792,
|
| 2668 |
+
"learning_rate": 5.205303969687658e-07,
|
| 2669 |
+
"loss": 1.1151776313781738,
|
| 2670 |
+
"step": 3720
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 2.9211337935085915,
|
| 2674 |
+
"grad_norm": 0.29611772298812866,
|
| 2675 |
+
"learning_rate": 4.317921446011286e-07,
|
| 2676 |
+
"loss": 1.1126981735229493,
|
| 2677 |
+
"step": 3730
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 2.9289665638615556,
|
| 2681 |
+
"grad_norm": 0.37433868646621704,
|
| 2682 |
+
"learning_rate": 3.513244512045155e-07,
|
| 2683 |
+
"loss": 1.1250258445739747,
|
| 2684 |
+
"step": 3740
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 2.9367993342145198,
|
| 2688 |
+
"grad_norm": 0.3662181794643402,
|
| 2689 |
+
"learning_rate": 2.7913400076787776e-07,
|
| 2690 |
+
"loss": 1.0847277641296387,
|
| 2691 |
+
"step": 3750
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 2.9446321045674844,
|
| 2695 |
+
"grad_norm": 0.3808414340019226,
|
| 2696 |
+
"learning_rate": 2.1522678973718847e-07,
|
| 2697 |
+
"loss": 1.1382882118225097,
|
| 2698 |
+
"step": 3760
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 2.9524648749204485,
|
| 2702 |
+
"grad_norm": 0.3531859219074249,
|
| 2703 |
+
"learning_rate": 1.5960812651728552e-07,
|
| 2704 |
+
"loss": 1.1251526832580567,
|
| 2705 |
+
"step": 3770
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 2.9602976452734127,
|
| 2709 |
+
"grad_norm": 0.3787539005279541,
|
| 2710 |
+
"learning_rate": 1.1228263103099101e-07,
|
| 2711 |
+
"loss": 1.097026538848877,
|
| 2712 |
+
"step": 3780
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 2.968130415626377,
|
| 2716 |
+
"grad_norm": 0.3101099133491516,
|
| 2717 |
+
"learning_rate": 7.32542343352849e-08,
|
| 2718 |
+
"loss": 1.0861360549926757,
|
| 2719 |
+
"step": 3790
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 2.975963185979341,
|
| 2723 |
+
"grad_norm": 0.320687860250473,
|
| 2724 |
+
"learning_rate": 4.2526178294877153e-08,
|
| 2725 |
+
"loss": 1.0637449264526366,
|
| 2726 |
+
"step": 3800
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 2.9837959563323055,
|
| 2730 |
+
"grad_norm": 0.2555408775806427,
|
| 2731 |
+
"learning_rate": 2.010101531282338e-08,
|
| 2732 |
+
"loss": 1.1155780792236327,
|
| 2733 |
+
"step": 3810
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 2.9916287266852697,
|
| 2737 |
+
"grad_norm": 0.39955031871795654,
|
| 2738 |
+
"learning_rate": 5.980608118572039e-09,
|
| 2739 |
+
"loss": 1.1184403419494628,
|
| 2740 |
+
"step": 3820
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 2.999461497038234,
|
| 2744 |
+
"grad_norm": 0.35409417748451233,
|
| 2745 |
+
"learning_rate": 1.661296132327017e-10,
|
| 2746 |
+
"loss": 1.114205551147461,
|
| 2747 |
+
"step": 3830
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 3.0,
|
| 2751 |
+
"eval_loss": 1.2653985023498535,
|
| 2752 |
+
"eval_runtime": 324.3577,
|
| 2753 |
+
"eval_samples_per_second": 3.317,
|
| 2754 |
+
"eval_steps_per_second": 3.317,
|
| 2755 |
+
"step": 3831
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 3.0,
|
| 2759 |
+
"step": 3831,
|
| 2760 |
+
"total_flos": 3.5276122661958113e+18,
|
| 2761 |
+
"train_loss": 1.2349582789336941,
|
| 2762 |
+
"train_runtime": 87806.6728,
|
| 2763 |
+
"train_samples_per_second": 0.698,
|
| 2764 |
+
"train_steps_per_second": 0.044
|
| 2765 |
+
}
|
| 2766 |
+
],
|
| 2767 |
+
"logging_steps": 10,
|
| 2768 |
+
"max_steps": 3831,
|
| 2769 |
+
"num_input_tokens_seen": 0,
|
| 2770 |
+
"num_train_epochs": 3,
|
| 2771 |
+
"save_steps": 500,
|
| 2772 |
+
"stateful_callbacks": {
|
| 2773 |
+
"TrainerControl": {
|
| 2774 |
+
"args": {
|
| 2775 |
+
"should_epoch_stop": false,
|
| 2776 |
+
"should_evaluate": false,
|
| 2777 |
+
"should_log": false,
|
| 2778 |
+
"should_save": true,
|
| 2779 |
+
"should_training_stop": true
|
| 2780 |
+
},
|
| 2781 |
+
"attributes": {}
|
| 2782 |
+
}
|
| 2783 |
+
},
|
| 2784 |
+
"total_flos": 3.5276122661958113e+18,
|
| 2785 |
+
"train_batch_size": 1,
|
| 2786 |
+
"trial_name": null,
|
| 2787 |
+
"trial_params": null
|
| 2788 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|