Commit
•
28b6f06
1
Parent(s):
c20551c
End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2520 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ library_name: peft
|
|
| 4 |
license: llama3.1
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: laaj_9k_3epochs
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# laaj_9k_3epochs
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Llama-Guard-3-8B](https://huggingface.co/meta-llama/Llama-Guard-3-8B) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
license: llama3.1
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- lora
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: laaj_9k_3epochs
|
|
|
|
| 16 |
|
| 17 |
# laaj_9k_3epochs
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Llama-Guard-3-8B](https://huggingface.co/meta-llama/Llama-Guard-3-8B) on the laaj_9k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9993650793650795,
|
| 3 |
+
"total_flos": 652111388999680.0,
|
| 4 |
+
"train_loss": 0.34479539529973763,
|
| 5 |
+
"train_runtime": 32429.3956,
|
| 6 |
+
"train_samples_per_second": 0.874,
|
| 7 |
+
"train_steps_per_second": 0.109
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9993650793650795,
|
| 3 |
+
"total_flos": 652111388999680.0,
|
| 4 |
+
"train_loss": 0.34479539529973763,
|
| 5 |
+
"train_runtime": 32429.3956,
|
| 6 |
+
"train_samples_per_second": 0.874,
|
| 7 |
+
"train_steps_per_second": 0.109
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2520 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9993650793650795,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3543,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008465608465608466,
|
| 13 |
+
"grad_norm": 1.2471813174095268,
|
| 14 |
+
"learning_rate": 9.999803440244217e-05,
|
| 15 |
+
"loss": 1.6288,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016931216931216932,
|
| 20 |
+
"grad_norm": 1.0863274390486097,
|
| 21 |
+
"learning_rate": 9.999213776431165e-05,
|
| 22 |
+
"loss": 0.9235,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.025396825396825397,
|
| 27 |
+
"grad_norm": 0.8727777966082182,
|
| 28 |
+
"learning_rate": 9.998231054922511e-05,
|
| 29 |
+
"loss": 0.76,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.033862433862433865,
|
| 34 |
+
"grad_norm": 0.9216466406898077,
|
| 35 |
+
"learning_rate": 9.996855352983658e-05,
|
| 36 |
+
"loss": 0.6727,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.042328042328042326,
|
| 41 |
+
"grad_norm": 0.8685947217114767,
|
| 42 |
+
"learning_rate": 9.995086778777658e-05,
|
| 43 |
+
"loss": 0.6129,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.050793650793650794,
|
| 48 |
+
"grad_norm": 1.2232366822709322,
|
| 49 |
+
"learning_rate": 9.99292547135672e-05,
|
| 50 |
+
"loss": 0.6334,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05925925925925926,
|
| 55 |
+
"grad_norm": 1.0539472852477354,
|
| 56 |
+
"learning_rate": 9.990371600651263e-05,
|
| 57 |
+
"loss": 0.6105,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06772486772486773,
|
| 62 |
+
"grad_norm": 0.9891412672344401,
|
| 63 |
+
"learning_rate": 9.987425367456573e-05,
|
| 64 |
+
"loss": 0.5657,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0761904761904762,
|
| 69 |
+
"grad_norm": 1.2412242369215105,
|
| 70 |
+
"learning_rate": 9.984087003416997e-05,
|
| 71 |
+
"loss": 0.5737,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08465608465608465,
|
| 76 |
+
"grad_norm": 1.040677823054172,
|
| 77 |
+
"learning_rate": 9.980356771007745e-05,
|
| 78 |
+
"loss": 0.59,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09312169312169312,
|
| 83 |
+
"grad_norm": 1.2535421463288563,
|
| 84 |
+
"learning_rate": 9.976234963514244e-05,
|
| 85 |
+
"loss": 0.537,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.10158730158730159,
|
| 90 |
+
"grad_norm": 1.0340076526320245,
|
| 91 |
+
"learning_rate": 9.971721905009086e-05,
|
| 92 |
+
"loss": 0.5238,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.11005291005291006,
|
| 97 |
+
"grad_norm": 0.9971098275169218,
|
| 98 |
+
"learning_rate": 9.966817950326541e-05,
|
| 99 |
+
"loss": 0.5225,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.11851851851851852,
|
| 104 |
+
"grad_norm": 1.227890883155069,
|
| 105 |
+
"learning_rate": 9.961523485034662e-05,
|
| 106 |
+
"loss": 0.5357,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.12698412698412698,
|
| 111 |
+
"grad_norm": 1.0605213954707176,
|
| 112 |
+
"learning_rate": 9.955838925404971e-05,
|
| 113 |
+
"loss": 0.5363,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13544973544973546,
|
| 118 |
+
"grad_norm": 0.9916561072816364,
|
| 119 |
+
"learning_rate": 9.94976471837973e-05,
|
| 120 |
+
"loss": 0.5107,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1439153439153439,
|
| 125 |
+
"grad_norm": 0.9224406793350302,
|
| 126 |
+
"learning_rate": 9.943301341536799e-05,
|
| 127 |
+
"loss": 0.5291,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1523809523809524,
|
| 132 |
+
"grad_norm": 0.9459998168670802,
|
| 133 |
+
"learning_rate": 9.936449303052086e-05,
|
| 134 |
+
"loss": 0.4909,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.16084656084656085,
|
| 139 |
+
"grad_norm": 1.1437589163268378,
|
| 140 |
+
"learning_rate": 9.929209141659598e-05,
|
| 141 |
+
"loss": 0.5365,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1693121693121693,
|
| 146 |
+
"grad_norm": 1.0253073070364553,
|
| 147 |
+
"learning_rate": 9.921581426609074e-05,
|
| 148 |
+
"loss": 0.5505,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.17777777777777778,
|
| 153 |
+
"grad_norm": 1.2156932145478005,
|
| 154 |
+
"learning_rate": 9.91356675762124e-05,
|
| 155 |
+
"loss": 0.5099,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.18624338624338624,
|
| 160 |
+
"grad_norm": 1.0357139504839994,
|
| 161 |
+
"learning_rate": 9.905165764840645e-05,
|
| 162 |
+
"loss": 0.5016,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.19470899470899472,
|
| 167 |
+
"grad_norm": 1.0513703237178957,
|
| 168 |
+
"learning_rate": 9.896379108786126e-05,
|
| 169 |
+
"loss": 0.502,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.20317460317460317,
|
| 174 |
+
"grad_norm": 1.0533509834595176,
|
| 175 |
+
"learning_rate": 9.88720748029887e-05,
|
| 176 |
+
"loss": 0.4945,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.21164021164021163,
|
| 181 |
+
"grad_norm": 1.0842886505846092,
|
| 182 |
+
"learning_rate": 9.877651600488099e-05,
|
| 183 |
+
"loss": 0.4765,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2201058201058201,
|
| 188 |
+
"grad_norm": 1.0079186405663672,
|
| 189 |
+
"learning_rate": 9.867712220674374e-05,
|
| 190 |
+
"loss": 0.4842,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.22857142857142856,
|
| 195 |
+
"grad_norm": 0.9843739521940763,
|
| 196 |
+
"learning_rate": 9.857390122330522e-05,
|
| 197 |
+
"loss": 0.461,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.23703703703703705,
|
| 202 |
+
"grad_norm": 1.0682125354651388,
|
| 203 |
+
"learning_rate": 9.846686117020196e-05,
|
| 204 |
+
"loss": 0.4867,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2455026455026455,
|
| 209 |
+
"grad_norm": 0.9969252730804404,
|
| 210 |
+
"learning_rate": 9.835601046334065e-05,
|
| 211 |
+
"loss": 0.4765,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.25396825396825395,
|
| 216 |
+
"grad_norm": 1.145271346749039,
|
| 217 |
+
"learning_rate": 9.82413578182364e-05,
|
| 218 |
+
"loss": 0.5131,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2624338624338624,
|
| 223 |
+
"grad_norm": 1.0296375563736684,
|
| 224 |
+
"learning_rate": 9.81229122493276e-05,
|
| 225 |
+
"loss": 0.4762,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2708994708994709,
|
| 230 |
+
"grad_norm": 1.0014721034964151,
|
| 231 |
+
"learning_rate": 9.800068306926708e-05,
|
| 232 |
+
"loss": 0.4663,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.27936507936507937,
|
| 237 |
+
"grad_norm": 1.0076225606653653,
|
| 238 |
+
"learning_rate": 9.787467988818997e-05,
|
| 239 |
+
"loss": 0.4918,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2878306878306878,
|
| 244 |
+
"grad_norm": 1.1916493242294524,
|
| 245 |
+
"learning_rate": 9.774491261295807e-05,
|
| 246 |
+
"loss": 0.4712,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2962962962962963,
|
| 251 |
+
"grad_norm": 1.0123926586182874,
|
| 252 |
+
"learning_rate": 9.761139144638094e-05,
|
| 253 |
+
"loss": 0.4648,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3047619047619048,
|
| 258 |
+
"grad_norm": 1.0273059230601556,
|
| 259 |
+
"learning_rate": 9.747412688641372e-05,
|
| 260 |
+
"loss": 0.4956,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.31322751322751324,
|
| 265 |
+
"grad_norm": 1.1785090929345263,
|
| 266 |
+
"learning_rate": 9.733312972533182e-05,
|
| 267 |
+
"loss": 0.4788,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3216931216931217,
|
| 272 |
+
"grad_norm": 1.0144172117749461,
|
| 273 |
+
"learning_rate": 9.71884110488822e-05,
|
| 274 |
+
"loss": 0.4596,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.33015873015873015,
|
| 279 |
+
"grad_norm": 1.043899340203269,
|
| 280 |
+
"learning_rate": 9.703998223541196e-05,
|
| 281 |
+
"loss": 0.4579,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.3386243386243386,
|
| 286 |
+
"grad_norm": 1.1486794203887134,
|
| 287 |
+
"learning_rate": 9.688785495497366e-05,
|
| 288 |
+
"loss": 0.4699,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3470899470899471,
|
| 293 |
+
"grad_norm": 1.0373759945758736,
|
| 294 |
+
"learning_rate": 9.673204116840768e-05,
|
| 295 |
+
"loss": 0.4794,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.35555555555555557,
|
| 300 |
+
"grad_norm": 0.9675582489645671,
|
| 301 |
+
"learning_rate": 9.6572553126402e-05,
|
| 302 |
+
"loss": 0.4506,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.364021164021164,
|
| 307 |
+
"grad_norm": 0.9254293000510532,
|
| 308 |
+
"learning_rate": 9.640940336852884e-05,
|
| 309 |
+
"loss": 0.4633,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3724867724867725,
|
| 314 |
+
"grad_norm": 1.0325559808636224,
|
| 315 |
+
"learning_rate": 9.624260472225882e-05,
|
| 316 |
+
"loss": 0.45,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.38095238095238093,
|
| 321 |
+
"grad_norm": 1.012802385726723,
|
| 322 |
+
"learning_rate": 9.607217030195242e-05,
|
| 323 |
+
"loss": 0.4618,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.38941798941798944,
|
| 328 |
+
"grad_norm": 1.0486189015875456,
|
| 329 |
+
"learning_rate": 9.589811350782884e-05,
|
| 330 |
+
"loss": 0.4694,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.3978835978835979,
|
| 335 |
+
"grad_norm": 1.1012030940322182,
|
| 336 |
+
"learning_rate": 9.572044802491247e-05,
|
| 337 |
+
"loss": 0.4505,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.40634920634920635,
|
| 342 |
+
"grad_norm": 1.0060674054310905,
|
| 343 |
+
"learning_rate": 9.553918782195688e-05,
|
| 344 |
+
"loss": 0.452,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4148148148148148,
|
| 349 |
+
"grad_norm": 1.022772188842258,
|
| 350 |
+
"learning_rate": 9.535434715034656e-05,
|
| 351 |
+
"loss": 0.4356,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.42328042328042326,
|
| 356 |
+
"grad_norm": 0.9036986631596517,
|
| 357 |
+
"learning_rate": 9.516594054297642e-05,
|
| 358 |
+
"loss": 0.454,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.43174603174603177,
|
| 363 |
+
"grad_norm": 0.8998291431871417,
|
| 364 |
+
"learning_rate": 9.497398281310914e-05,
|
| 365 |
+
"loss": 0.4162,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4402116402116402,
|
| 370 |
+
"grad_norm": 0.9267883794464391,
|
| 371 |
+
"learning_rate": 9.477848905321055e-05,
|
| 372 |
+
"loss": 0.453,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4486772486772487,
|
| 377 |
+
"grad_norm": 1.0322711884681597,
|
| 378 |
+
"learning_rate": 9.45794746337629e-05,
|
| 379 |
+
"loss": 0.4505,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.45714285714285713,
|
| 384 |
+
"grad_norm": 1.0203838607248878,
|
| 385 |
+
"learning_rate": 9.43769552020565e-05,
|
| 386 |
+
"loss": 0.4384,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.4656084656084656,
|
| 391 |
+
"grad_norm": 0.9881433507860394,
|
| 392 |
+
"learning_rate": 9.417094668095932e-05,
|
| 393 |
+
"loss": 0.4606,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4740740740740741,
|
| 398 |
+
"grad_norm": 1.0302659809674581,
|
| 399 |
+
"learning_rate": 9.396146526766522e-05,
|
| 400 |
+
"loss": 0.4285,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.48253968253968255,
|
| 405 |
+
"grad_norm": 1.0511136894758872,
|
| 406 |
+
"learning_rate": 9.374852743242037e-05,
|
| 407 |
+
"loss": 0.4643,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.491005291005291,
|
| 412 |
+
"grad_norm": 0.9837494678922593,
|
| 413 |
+
"learning_rate": 9.353214991722834e-05,
|
| 414 |
+
"loss": 0.4461,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.49947089947089945,
|
| 419 |
+
"grad_norm": 0.9920985093408109,
|
| 420 |
+
"learning_rate": 9.331234973453375e-05,
|
| 421 |
+
"loss": 0.4655,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5079365079365079,
|
| 426 |
+
"grad_norm": 0.980959216204584,
|
| 427 |
+
"learning_rate": 9.308914416588468e-05,
|
| 428 |
+
"loss": 0.4155,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5164021164021164,
|
| 433 |
+
"grad_norm": 1.031774153957894,
|
| 434 |
+
"learning_rate": 9.286255076057395e-05,
|
| 435 |
+
"loss": 0.4213,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5248677248677248,
|
| 440 |
+
"grad_norm": 0.8674398091480106,
|
| 441 |
+
"learning_rate": 9.263258733425933e-05,
|
| 442 |
+
"loss": 0.4076,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5333333333333333,
|
| 447 |
+
"grad_norm": 1.0196562220963516,
|
| 448 |
+
"learning_rate": 9.239927196756279e-05,
|
| 449 |
+
"loss": 0.4496,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5417989417989418,
|
| 454 |
+
"grad_norm": 1.015803047506094,
|
| 455 |
+
"learning_rate": 9.216262300464893e-05,
|
| 456 |
+
"loss": 0.4399,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5502645502645502,
|
| 461 |
+
"grad_norm": 0.8444200783279879,
|
| 462 |
+
"learning_rate": 9.192265905178269e-05,
|
| 463 |
+
"loss": 0.4394,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5587301587301587,
|
| 468 |
+
"grad_norm": 0.9377213246241002,
|
| 469 |
+
"learning_rate": 9.167939897586647e-05,
|
| 470 |
+
"loss": 0.4483,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5671957671957671,
|
| 475 |
+
"grad_norm": 1.059524915742627,
|
| 476 |
+
"learning_rate": 9.143286190295668e-05,
|
| 477 |
+
"loss": 0.4307,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5756613756613757,
|
| 482 |
+
"grad_norm": 0.923761212995823,
|
| 483 |
+
"learning_rate": 9.118306721676008e-05,
|
| 484 |
+
"loss": 0.4508,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5841269841269842,
|
| 489 |
+
"grad_norm": 1.0801213897332445,
|
| 490 |
+
"learning_rate": 9.09300345571097e-05,
|
| 491 |
+
"loss": 0.4454,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5925925925925926,
|
| 496 |
+
"grad_norm": 1.0572026584991792,
|
| 497 |
+
"learning_rate": 9.067378381842063e-05,
|
| 498 |
+
"loss": 0.4056,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6010582010582011,
|
| 503 |
+
"grad_norm": 0.887247483674217,
|
| 504 |
+
"learning_rate": 9.041433514812589e-05,
|
| 505 |
+
"loss": 0.4565,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.6095238095238096,
|
| 510 |
+
"grad_norm": 0.9437825913047283,
|
| 511 |
+
"learning_rate": 9.015170894509244e-05,
|
| 512 |
+
"loss": 0.3808,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.617989417989418,
|
| 517 |
+
"grad_norm": 0.9193237876699505,
|
| 518 |
+
"learning_rate": 8.988592585801719e-05,
|
| 519 |
+
"loss": 0.4397,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.6264550264550265,
|
| 524 |
+
"grad_norm": 1.0095223700805926,
|
| 525 |
+
"learning_rate": 8.961700678380359e-05,
|
| 526 |
+
"loss": 0.4667,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6349206349206349,
|
| 531 |
+
"grad_norm": 1.0449169193411385,
|
| 532 |
+
"learning_rate": 8.93449728659187e-05,
|
| 533 |
+
"loss": 0.4269,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.6433862433862434,
|
| 538 |
+
"grad_norm": 0.980016186986963,
|
| 539 |
+
"learning_rate": 8.906984549273069e-05,
|
| 540 |
+
"loss": 0.4324,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6518518518518519,
|
| 545 |
+
"grad_norm": 0.9073046321616872,
|
| 546 |
+
"learning_rate": 8.879164629582726e-05,
|
| 547 |
+
"loss": 0.3978,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.6603174603174603,
|
| 552 |
+
"grad_norm": 1.0375673558788747,
|
| 553 |
+
"learning_rate": 8.851039714831492e-05,
|
| 554 |
+
"loss": 0.4076,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6687830687830688,
|
| 559 |
+
"grad_norm": 0.9857310889370984,
|
| 560 |
+
"learning_rate": 8.822612016309915e-05,
|
| 561 |
+
"loss": 0.4167,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6772486772486772,
|
| 566 |
+
"grad_norm": 0.8471996503361908,
|
| 567 |
+
"learning_rate": 8.793883769114586e-05,
|
| 568 |
+
"loss": 0.4043,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.6857142857142857,
|
| 573 |
+
"grad_norm": 0.8920445010354429,
|
| 574 |
+
"learning_rate": 8.764857231972408e-05,
|
| 575 |
+
"loss": 0.4277,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.6941798941798942,
|
| 580 |
+
"grad_norm": 1.0104461104380098,
|
| 581 |
+
"learning_rate": 8.735534687062999e-05,
|
| 582 |
+
"loss": 0.4141,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.7026455026455026,
|
| 587 |
+
"grad_norm": 0.9774326975200812,
|
| 588 |
+
"learning_rate": 8.70591843983927e-05,
|
| 589 |
+
"loss": 0.4257,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.7111111111111111,
|
| 594 |
+
"grad_norm": 0.8665516757336554,
|
| 595 |
+
"learning_rate": 8.676010818846145e-05,
|
| 596 |
+
"loss": 0.4399,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.7195767195767195,
|
| 601 |
+
"grad_norm": 1.0277004789770665,
|
| 602 |
+
"learning_rate": 8.645814175537497e-05,
|
| 603 |
+
"loss": 0.4514,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.728042328042328,
|
| 608 |
+
"grad_norm": 0.9204427078381576,
|
| 609 |
+
"learning_rate": 8.615330884091261e-05,
|
| 610 |
+
"loss": 0.4144,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7365079365079366,
|
| 615 |
+
"grad_norm": 1.1301407306313882,
|
| 616 |
+
"learning_rate": 8.584563341222765e-05,
|
| 617 |
+
"loss": 0.3885,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.744973544973545,
|
| 622 |
+
"grad_norm": 0.9458002020904854,
|
| 623 |
+
"learning_rate": 8.553513965996293e-05,
|
| 624 |
+
"loss": 0.4365,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7534391534391535,
|
| 629 |
+
"grad_norm": 0.9889479454065517,
|
| 630 |
+
"learning_rate": 8.52218519963489e-05,
|
| 631 |
+
"loss": 0.4229,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7619047619047619,
|
| 636 |
+
"grad_norm": 0.850438656019796,
|
| 637 |
+
"learning_rate": 8.490579505328424e-05,
|
| 638 |
+
"loss": 0.4306,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7703703703703704,
|
| 643 |
+
"grad_norm": 0.9370585479712853,
|
| 644 |
+
"learning_rate": 8.458699368039913e-05,
|
| 645 |
+
"loss": 0.4362,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.7788359788359789,
|
| 650 |
+
"grad_norm": 0.9677664011225651,
|
| 651 |
+
"learning_rate": 8.426547294310159e-05,
|
| 652 |
+
"loss": 0.4025,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.7873015873015873,
|
| 657 |
+
"grad_norm": 1.029715838177075,
|
| 658 |
+
"learning_rate": 8.394125812060666e-05,
|
| 659 |
+
"loss": 0.4093,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7957671957671958,
|
| 664 |
+
"grad_norm": 0.9344248674021697,
|
| 665 |
+
"learning_rate": 8.361437470394889e-05,
|
| 666 |
+
"loss": 0.4039,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.8042328042328042,
|
| 671 |
+
"grad_norm": 0.9458006999291949,
|
| 672 |
+
"learning_rate": 8.328484839397807e-05,
|
| 673 |
+
"loss": 0.3985,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.8126984126984127,
|
| 678 |
+
"grad_norm": 0.9459246300643047,
|
| 679 |
+
"learning_rate": 8.295270509933862e-05,
|
| 680 |
+
"loss": 0.4092,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.8211640211640212,
|
| 685 |
+
"grad_norm": 1.0060543461216613,
|
| 686 |
+
"learning_rate": 8.261797093443249e-05,
|
| 687 |
+
"loss": 0.395,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.8296296296296296,
|
| 692 |
+
"grad_norm": 1.1118301896881613,
|
| 693 |
+
"learning_rate": 8.228067221736595e-05,
|
| 694 |
+
"loss": 0.3968,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.8380952380952381,
|
| 699 |
+
"grad_norm": 0.9258014283812922,
|
| 700 |
+
"learning_rate": 8.19408354678804e-05,
|
| 701 |
+
"loss": 0.4025,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8465608465608465,
|
| 706 |
+
"grad_norm": 0.9611975674417222,
|
| 707 |
+
"learning_rate": 8.159848740526723e-05,
|
| 708 |
+
"loss": 0.4189,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.855026455026455,
|
| 713 |
+
"grad_norm": 1.0106358106141422,
|
| 714 |
+
"learning_rate": 8.125365494626706e-05,
|
| 715 |
+
"loss": 0.4062,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8634920634920635,
|
| 720 |
+
"grad_norm": 1.0337925818103473,
|
| 721 |
+
"learning_rate": 8.090636520295348e-05,
|
| 722 |
+
"loss": 0.4064,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8719576719576719,
|
| 727 |
+
"grad_norm": 0.9347719089761312,
|
| 728 |
+
"learning_rate": 8.055664548060131e-05,
|
| 729 |
+
"loss": 0.4085,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.8804232804232804,
|
| 734 |
+
"grad_norm": 0.9001085371362909,
|
| 735 |
+
"learning_rate": 8.020452327553986e-05,
|
| 736 |
+
"loss": 0.4206,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.8888888888888888,
|
| 741 |
+
"grad_norm": 0.9005574040109541,
|
| 742 |
+
"learning_rate": 7.9850026272991e-05,
|
| 743 |
+
"loss": 0.4139,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.8973544973544973,
|
| 748 |
+
"grad_norm": 0.9668702597620379,
|
| 749 |
+
"learning_rate": 7.94931823448924e-05,
|
| 750 |
+
"loss": 0.4152,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.9058201058201059,
|
| 755 |
+
"grad_norm": 1.055087550179992,
|
| 756 |
+
"learning_rate": 7.913401954770622e-05,
|
| 757 |
+
"loss": 0.3971,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.9142857142857143,
|
| 762 |
+
"grad_norm": 0.8974328370738465,
|
| 763 |
+
"learning_rate": 7.877256612021312e-05,
|
| 764 |
+
"loss": 0.4147,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.9227513227513228,
|
| 769 |
+
"grad_norm": 1.0826599341716459,
|
| 770 |
+
"learning_rate": 7.84088504812921e-05,
|
| 771 |
+
"loss": 0.4012,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.9312169312169312,
|
| 776 |
+
"grad_norm": 0.9939329938043986,
|
| 777 |
+
"learning_rate": 7.804290122768601e-05,
|
| 778 |
+
"loss": 0.3925,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9396825396825397,
|
| 783 |
+
"grad_norm": 0.8740614733530233,
|
| 784 |
+
"learning_rate": 7.767474713175321e-05,
|
| 785 |
+
"loss": 0.4102,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.9481481481481482,
|
| 790 |
+
"grad_norm": 0.9603955711346013,
|
| 791 |
+
"learning_rate": 7.73044171392054e-05,
|
| 792 |
+
"loss": 0.4065,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9566137566137566,
|
| 797 |
+
"grad_norm": 1.127743100139354,
|
| 798 |
+
"learning_rate": 7.69319403668317e-05,
|
| 799 |
+
"loss": 0.3827,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9650793650793651,
|
| 804 |
+
"grad_norm": 0.860392518744168,
|
| 805 |
+
"learning_rate": 7.655734610020952e-05,
|
| 806 |
+
"loss": 0.4088,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.9735449735449735,
|
| 811 |
+
"grad_norm": 0.9182302824948408,
|
| 812 |
+
"learning_rate": 7.618066379140186e-05,
|
| 813 |
+
"loss": 0.4154,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.982010582010582,
|
| 818 |
+
"grad_norm": 0.9472264649293427,
|
| 819 |
+
"learning_rate": 7.580192305664174e-05,
|
| 820 |
+
"loss": 0.3986,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.9904761904761905,
|
| 825 |
+
"grad_norm": 0.8736825227580656,
|
| 826 |
+
"learning_rate": 7.542115367400376e-05,
|
| 827 |
+
"loss": 0.3929,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.9989417989417989,
|
| 832 |
+
"grad_norm": 0.9809536270766679,
|
| 833 |
+
"learning_rate": 7.503838558106263e-05,
|
| 834 |
+
"loss": 0.3868,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.0074074074074073,
|
| 839 |
+
"grad_norm": 1.0636167972989774,
|
| 840 |
+
"learning_rate": 7.465364887253949e-05,
|
| 841 |
+
"loss": 0.3725,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.0158730158730158,
|
| 846 |
+
"grad_norm": 1.107441272645184,
|
| 847 |
+
"learning_rate": 7.426697379793572e-05,
|
| 848 |
+
"loss": 0.3436,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.0243386243386243,
|
| 853 |
+
"grad_norm": 0.9993424372015514,
|
| 854 |
+
"learning_rate": 7.387839075915465e-05,
|
| 855 |
+
"loss": 0.3434,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.0328042328042328,
|
| 860 |
+
"grad_norm": 0.9466656285727851,
|
| 861 |
+
"learning_rate": 7.348793030811111e-05,
|
| 862 |
+
"loss": 0.3333,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.0412698412698413,
|
| 867 |
+
"grad_norm": 0.9712036510252812,
|
| 868 |
+
"learning_rate": 7.309562314432952e-05,
|
| 869 |
+
"loss": 0.3205,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.0497354497354496,
|
| 874 |
+
"grad_norm": 0.9567265246539378,
|
| 875 |
+
"learning_rate": 7.270150011252994e-05,
|
| 876 |
+
"loss": 0.3426,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.0582010582010581,
|
| 881 |
+
"grad_norm": 1.1511281895276615,
|
| 882 |
+
"learning_rate": 7.230559220020319e-05,
|
| 883 |
+
"loss": 0.3197,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.0666666666666667,
|
| 888 |
+
"grad_norm": 1.143315468276855,
|
| 889 |
+
"learning_rate": 7.190793053517424e-05,
|
| 890 |
+
"loss": 0.3317,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.0751322751322752,
|
| 895 |
+
"grad_norm": 0.9958926341045374,
|
| 896 |
+
"learning_rate": 7.1508546383155e-05,
|
| 897 |
+
"loss": 0.3264,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.0835978835978837,
|
| 902 |
+
"grad_norm": 1.028371010948524,
|
| 903 |
+
"learning_rate": 7.110747114528604e-05,
|
| 904 |
+
"loss": 0.3294,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.0920634920634922,
|
| 909 |
+
"grad_norm": 1.0914872277776548,
|
| 910 |
+
"learning_rate": 7.070473635566768e-05,
|
| 911 |
+
"loss": 0.3272,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.1005291005291005,
|
| 916 |
+
"grad_norm": 1.127102783013863,
|
| 917 |
+
"learning_rate": 7.030037367888069e-05,
|
| 918 |
+
"loss": 0.3382,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.108994708994709,
|
| 923 |
+
"grad_norm": 0.9437414070048694,
|
| 924 |
+
"learning_rate": 6.989441490749662e-05,
|
| 925 |
+
"loss": 0.313,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.1174603174603175,
|
| 930 |
+
"grad_norm": 1.018312148913929,
|
| 931 |
+
"learning_rate": 6.948689195957829e-05,
|
| 932 |
+
"loss": 0.3261,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.125925925925926,
|
| 937 |
+
"grad_norm": 1.1453502643848026,
|
| 938 |
+
"learning_rate": 6.907783687617017e-05,
|
| 939 |
+
"loss": 0.3295,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.1343915343915345,
|
| 944 |
+
"grad_norm": 1.0896304992556385,
|
| 945 |
+
"learning_rate": 6.866728181877914e-05,
|
| 946 |
+
"loss": 0.3208,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.1428571428571428,
|
| 951 |
+
"grad_norm": 1.0673296804602648,
|
| 952 |
+
"learning_rate": 6.825525906684593e-05,
|
| 953 |
+
"loss": 0.2993,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.1513227513227513,
|
| 958 |
+
"grad_norm": 1.155110295228161,
|
| 959 |
+
"learning_rate": 6.784180101520717e-05,
|
| 960 |
+
"loss": 0.3519,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.1597883597883598,
|
| 965 |
+
"grad_norm": 1.0507064287253836,
|
| 966 |
+
"learning_rate": 6.74269401715483e-05,
|
| 967 |
+
"loss": 0.3468,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.1682539682539683,
|
| 972 |
+
"grad_norm": 1.1767438102467995,
|
| 973 |
+
"learning_rate": 6.701070915384776e-05,
|
| 974 |
+
"loss": 0.3666,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.1767195767195768,
|
| 979 |
+
"grad_norm": 1.1040760205989266,
|
| 980 |
+
"learning_rate": 6.659314068781241e-05,
|
| 981 |
+
"loss": 0.3516,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.1851851851851851,
|
| 986 |
+
"grad_norm": 1.1334475282531191,
|
| 987 |
+
"learning_rate": 6.617426760430457e-05,
|
| 988 |
+
"loss": 0.3444,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.1936507936507936,
|
| 993 |
+
"grad_norm": 1.020666210262446,
|
| 994 |
+
"learning_rate": 6.575412283676063e-05,
|
| 995 |
+
"loss": 0.3246,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.2021164021164021,
|
| 1000 |
+
"grad_norm": 1.088794196150716,
|
| 1001 |
+
"learning_rate": 6.533273941860174e-05,
|
| 1002 |
+
"loss": 0.3406,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.2105820105820106,
|
| 1007 |
+
"grad_norm": 1.0604023713287243,
|
| 1008 |
+
"learning_rate": 6.491015048063662e-05,
|
| 1009 |
+
"loss": 0.3383,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.2190476190476192,
|
| 1014 |
+
"grad_norm": 1.0596416270343272,
|
| 1015 |
+
"learning_rate": 6.448638924845662e-05,
|
| 1016 |
+
"loss": 0.3217,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.2275132275132274,
|
| 1021 |
+
"grad_norm": 1.1101059810936256,
|
| 1022 |
+
"learning_rate": 6.40614890398235e-05,
|
| 1023 |
+
"loss": 0.3481,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.235978835978836,
|
| 1028 |
+
"grad_norm": 1.0321989665703173,
|
| 1029 |
+
"learning_rate": 6.36354832620497e-05,
|
| 1030 |
+
"loss": 0.3313,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.2444444444444445,
|
| 1035 |
+
"grad_norm": 1.076200733585246,
|
| 1036 |
+
"learning_rate": 6.320840540937196e-05,
|
| 1037 |
+
"loss": 0.3286,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.252910052910053,
|
| 1042 |
+
"grad_norm": 1.2096711574512293,
|
| 1043 |
+
"learning_rate": 6.278028906031757e-05,
|
| 1044 |
+
"loss": 0.337,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.2613756613756615,
|
| 1049 |
+
"grad_norm": 1.10111033029835,
|
| 1050 |
+
"learning_rate": 6.235116787506458e-05,
|
| 1051 |
+
"loss": 0.3373,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.2698412698412698,
|
| 1056 |
+
"grad_norm": 1.0054036677530953,
|
| 1057 |
+
"learning_rate": 6.192107559279513e-05,
|
| 1058 |
+
"loss": 0.344,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.2783068783068783,
|
| 1063 |
+
"grad_norm": 1.1848267908504544,
|
| 1064 |
+
"learning_rate": 6.14900460290428e-05,
|
| 1065 |
+
"loss": 0.3273,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.2867724867724868,
|
| 1070 |
+
"grad_norm": 1.1024985637947646,
|
| 1071 |
+
"learning_rate": 6.105811307303391e-05,
|
| 1072 |
+
"loss": 0.3283,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.2952380952380953,
|
| 1077 |
+
"grad_norm": 0.9670540378850452,
|
| 1078 |
+
"learning_rate": 6.0625310685023006e-05,
|
| 1079 |
+
"loss": 0.3365,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.3037037037037038,
|
| 1084 |
+
"grad_norm": 1.1427343254939735,
|
| 1085 |
+
"learning_rate": 6.0191672893622766e-05,
|
| 1086 |
+
"loss": 0.3634,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.312169312169312,
|
| 1091 |
+
"grad_norm": 1.1927363160896671,
|
| 1092 |
+
"learning_rate": 5.975723379312851e-05,
|
| 1093 |
+
"loss": 0.3269,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.3206349206349206,
|
| 1098 |
+
"grad_norm": 1.0661933157446577,
|
| 1099 |
+
"learning_rate": 5.932202754083765e-05,
|
| 1100 |
+
"loss": 0.3635,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.3291005291005291,
|
| 1105 |
+
"grad_norm": 1.080671677917908,
|
| 1106 |
+
"learning_rate": 5.888608835436408e-05,
|
| 1107 |
+
"loss": 0.3288,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.3375661375661376,
|
| 1112 |
+
"grad_norm": 1.0117605430718117,
|
| 1113 |
+
"learning_rate": 5.844945050894779e-05,
|
| 1114 |
+
"loss": 0.3225,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.3460317460317461,
|
| 1119 |
+
"grad_norm": 0.9733421931133834,
|
| 1120 |
+
"learning_rate": 5.8012148334760077e-05,
|
| 1121 |
+
"loss": 0.3146,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.3544973544973544,
|
| 1126 |
+
"grad_norm": 1.0677601655228577,
|
| 1127 |
+
"learning_rate": 5.757421621420438e-05,
|
| 1128 |
+
"loss": 0.323,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.362962962962963,
|
| 1133 |
+
"grad_norm": 1.027062473997525,
|
| 1134 |
+
"learning_rate": 5.713568857921296e-05,
|
| 1135 |
+
"loss": 0.3071,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.3714285714285714,
|
| 1140 |
+
"grad_norm": 0.979546471375574,
|
| 1141 |
+
"learning_rate": 5.669659990853975e-05,
|
| 1142 |
+
"loss": 0.3317,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.37989417989418,
|
| 1147 |
+
"grad_norm": 1.065648765600161,
|
| 1148 |
+
"learning_rate": 5.62569847250495e-05,
|
| 1149 |
+
"loss": 0.3109,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.3883597883597885,
|
| 1154 |
+
"grad_norm": 0.9904468465024019,
|
| 1155 |
+
"learning_rate": 5.581687759300344e-05,
|
| 1156 |
+
"loss": 0.3295,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.3968253968253967,
|
| 1161 |
+
"grad_norm": 1.032660358149486,
|
| 1162 |
+
"learning_rate": 5.537631311534176e-05,
|
| 1163 |
+
"loss": 0.3112,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.4052910052910053,
|
| 1168 |
+
"grad_norm": 1.2430136826049512,
|
| 1169 |
+
"learning_rate": 5.493532593096289e-05,
|
| 1170 |
+
"loss": 0.3449,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.4137566137566138,
|
| 1175 |
+
"grad_norm": 1.1898277112027362,
|
| 1176 |
+
"learning_rate": 5.449395071200014e-05,
|
| 1177 |
+
"loss": 0.3341,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.4222222222222223,
|
| 1182 |
+
"grad_norm": 1.043179777209729,
|
| 1183 |
+
"learning_rate": 5.40522221610956e-05,
|
| 1184 |
+
"loss": 0.336,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.4306878306878308,
|
| 1189 |
+
"grad_norm": 1.0248865486416825,
|
| 1190 |
+
"learning_rate": 5.3610175008671736e-05,
|
| 1191 |
+
"loss": 0.3238,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.439153439153439,
|
| 1196 |
+
"grad_norm": 1.1689685076211649,
|
| 1197 |
+
"learning_rate": 5.316784401020064e-05,
|
| 1198 |
+
"loss": 0.346,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.4476190476190476,
|
| 1203 |
+
"grad_norm": 0.911540237077825,
|
| 1204 |
+
"learning_rate": 5.272526394347156e-05,
|
| 1205 |
+
"loss": 0.3282,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.456084656084656,
|
| 1210 |
+
"grad_norm": 1.0990655046574285,
|
| 1211 |
+
"learning_rate": 5.228246960585639e-05,
|
| 1212 |
+
"loss": 0.3362,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.4645502645502646,
|
| 1217 |
+
"grad_norm": 1.0636049790842568,
|
| 1218 |
+
"learning_rate": 5.18394958115739e-05,
|
| 1219 |
+
"loss": 0.3326,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.4730158730158731,
|
| 1224 |
+
"grad_norm": 0.9983006726426711,
|
| 1225 |
+
"learning_rate": 5.139637738895243e-05,
|
| 1226 |
+
"loss": 0.3277,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.4814814814814814,
|
| 1231 |
+
"grad_norm": 1.1542167670907033,
|
| 1232 |
+
"learning_rate": 5.0953149177691526e-05,
|
| 1233 |
+
"loss": 0.3094,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.48994708994709,
|
| 1238 |
+
"grad_norm": 1.1340593143811428,
|
| 1239 |
+
"learning_rate": 5.0509846026122796e-05,
|
| 1240 |
+
"loss": 0.3299,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.4984126984126984,
|
| 1245 |
+
"grad_norm": 1.058208878715113,
|
| 1246 |
+
"learning_rate": 5.006650278846991e-05,
|
| 1247 |
+
"loss": 0.3034,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.506878306878307,
|
| 1252 |
+
"grad_norm": 1.0744521042723432,
|
| 1253 |
+
"learning_rate": 4.962315432210827e-05,
|
| 1254 |
+
"loss": 0.3426,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.5153439153439154,
|
| 1259 |
+
"grad_norm": 1.1542452796166005,
|
| 1260 |
+
"learning_rate": 4.917983548482439e-05,
|
| 1261 |
+
"loss": 0.3218,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.5238095238095237,
|
| 1266 |
+
"grad_norm": 1.0168410109864015,
|
| 1267 |
+
"learning_rate": 4.873658113207526e-05,
|
| 1268 |
+
"loss": 0.3349,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.5322751322751322,
|
| 1273 |
+
"grad_norm": 0.9972037732726962,
|
| 1274 |
+
"learning_rate": 4.8293426114247755e-05,
|
| 1275 |
+
"loss": 0.3172,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.5407407407407407,
|
| 1280 |
+
"grad_norm": 1.0573820943940695,
|
| 1281 |
+
"learning_rate": 4.785040527391876e-05,
|
| 1282 |
+
"loss": 0.337,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.5492063492063493,
|
| 1287 |
+
"grad_norm": 0.9881874822329866,
|
| 1288 |
+
"learning_rate": 4.740755344311549e-05,
|
| 1289 |
+
"loss": 0.3067,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.5576719576719578,
|
| 1294 |
+
"grad_norm": 1.1012178266159593,
|
| 1295 |
+
"learning_rate": 4.6964905440577074e-05,
|
| 1296 |
+
"loss": 0.3361,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.566137566137566,
|
| 1301 |
+
"grad_norm": 1.1008138761037878,
|
| 1302 |
+
"learning_rate": 4.6522496069016823e-05,
|
| 1303 |
+
"loss": 0.3143,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.5746031746031746,
|
| 1308 |
+
"grad_norm": 1.0000855307005652,
|
| 1309 |
+
"learning_rate": 4.6080360112385914e-05,
|
| 1310 |
+
"loss": 0.3268,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.583068783068783,
|
| 1315 |
+
"grad_norm": 1.0946371145172427,
|
| 1316 |
+
"learning_rate": 4.5638532333138656e-05,
|
| 1317 |
+
"loss": 0.324,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.5915343915343916,
|
| 1322 |
+
"grad_norm": 1.1665682442710945,
|
| 1323 |
+
"learning_rate": 4.519704746949915e-05,
|
| 1324 |
+
"loss": 0.3056,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.6,
|
| 1329 |
+
"grad_norm": 1.042537667391268,
|
| 1330 |
+
"learning_rate": 4.475594023273024e-05,
|
| 1331 |
+
"loss": 0.3228,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.6084656084656084,
|
| 1336 |
+
"grad_norm": 1.2077526566227794,
|
| 1337 |
+
"learning_rate": 4.4315245304404175e-05,
|
| 1338 |
+
"loss": 0.3219,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.6169312169312169,
|
| 1343 |
+
"grad_norm": 0.9726841884389219,
|
| 1344 |
+
"learning_rate": 4.3874997333676e-05,
|
| 1345 |
+
"loss": 0.3071,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.6253968253968254,
|
| 1350 |
+
"grad_norm": 1.0872803982697865,
|
| 1351 |
+
"learning_rate": 4.343523093455909e-05,
|
| 1352 |
+
"loss": 0.3169,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.633862433862434,
|
| 1357 |
+
"grad_norm": 1.094914483331633,
|
| 1358 |
+
"learning_rate": 4.2995980683203905e-05,
|
| 1359 |
+
"loss": 0.3152,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.6423280423280424,
|
| 1364 |
+
"grad_norm": 1.0181256444534053,
|
| 1365 |
+
"learning_rate": 4.255728111517929e-05,
|
| 1366 |
+
"loss": 0.3403,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.6507936507936507,
|
| 1371 |
+
"grad_norm": 1.126842585732799,
|
| 1372 |
+
"learning_rate": 4.211916672275722e-05,
|
| 1373 |
+
"loss": 0.3132,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.6592592592592592,
|
| 1378 |
+
"grad_norm": 0.9511685681942271,
|
| 1379 |
+
"learning_rate": 4.168167195220088e-05,
|
| 1380 |
+
"loss": 0.299,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.6677248677248677,
|
| 1385 |
+
"grad_norm": 0.9727568749103337,
|
| 1386 |
+
"learning_rate": 4.1244831201056375e-05,
|
| 1387 |
+
"loss": 0.3059,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.6761904761904762,
|
| 1392 |
+
"grad_norm": 1.1382165054479474,
|
| 1393 |
+
"learning_rate": 4.080867881544826e-05,
|
| 1394 |
+
"loss": 0.318,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.6846560846560847,
|
| 1399 |
+
"grad_norm": 1.0381544841427086,
|
| 1400 |
+
"learning_rate": 4.037324908737909e-05,
|
| 1401 |
+
"loss": 0.3076,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.693121693121693,
|
| 1406 |
+
"grad_norm": 1.1334587465684427,
|
| 1407 |
+
"learning_rate": 3.9938576252033256e-05,
|
| 1408 |
+
"loss": 0.3052,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.7015873015873015,
|
| 1413 |
+
"grad_norm": 0.9774012866079483,
|
| 1414 |
+
"learning_rate": 3.950469448508532e-05,
|
| 1415 |
+
"loss": 0.3095,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.71005291005291,
|
| 1420 |
+
"grad_norm": 1.0840882454796814,
|
| 1421 |
+
"learning_rate": 3.907163790001292e-05,
|
| 1422 |
+
"loss": 0.3246,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.7185185185185186,
|
| 1427 |
+
"grad_norm": 1.1578703078806882,
|
| 1428 |
+
"learning_rate": 3.863944054541474e-05,
|
| 1429 |
+
"loss": 0.3148,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.726984126984127,
|
| 1434 |
+
"grad_norm": 1.0210773839195724,
|
| 1435 |
+
"learning_rate": 3.8208136402333314e-05,
|
| 1436 |
+
"loss": 0.3256,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.7354497354497354,
|
| 1441 |
+
"grad_norm": 1.0794889187437258,
|
| 1442 |
+
"learning_rate": 3.7777759381583495e-05,
|
| 1443 |
+
"loss": 0.2897,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.7439153439153439,
|
| 1448 |
+
"grad_norm": 1.0452290154108226,
|
| 1449 |
+
"learning_rate": 3.73483433210861e-05,
|
| 1450 |
+
"loss": 0.3051,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.7523809523809524,
|
| 1455 |
+
"grad_norm": 1.0972215102884366,
|
| 1456 |
+
"learning_rate": 3.691992198320753e-05,
|
| 1457 |
+
"loss": 0.3165,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.7608465608465609,
|
| 1462 |
+
"grad_norm": 1.102108276028283,
|
| 1463 |
+
"learning_rate": 3.649252905210519e-05,
|
| 1464 |
+
"loss": 0.2911,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.7693121693121694,
|
| 1469 |
+
"grad_norm": 1.0290079519183777,
|
| 1470 |
+
"learning_rate": 3.6066198131079146e-05,
|
| 1471 |
+
"loss": 0.3129,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.7777777777777777,
|
| 1476 |
+
"grad_norm": 1.0920319523740303,
|
| 1477 |
+
"learning_rate": 3.564096273993012e-05,
|
| 1478 |
+
"loss": 0.313,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.7862433862433862,
|
| 1483 |
+
"grad_norm": 1.2412824592232703,
|
| 1484 |
+
"learning_rate": 3.521685631232392e-05,
|
| 1485 |
+
"loss": 0.3202,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.7947089947089947,
|
| 1490 |
+
"grad_norm": 0.9652001472992224,
|
| 1491 |
+
"learning_rate": 3.479391219316291e-05,
|
| 1492 |
+
"loss": 0.3058,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.8031746031746032,
|
| 1497 |
+
"grad_norm": 1.2084364290298202,
|
| 1498 |
+
"learning_rate": 3.437216363596418e-05,
|
| 1499 |
+
"loss": 0.3192,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.8116402116402117,
|
| 1504 |
+
"grad_norm": 1.0200010445985095,
|
| 1505 |
+
"learning_rate": 3.395164380024512e-05,
|
| 1506 |
+
"loss": 0.3239,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.82010582010582,
|
| 1511 |
+
"grad_norm": 1.1971574618602432,
|
| 1512 |
+
"learning_rate": 3.3532385748916164e-05,
|
| 1513 |
+
"loss": 0.3229,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.8285714285714287,
|
| 1518 |
+
"grad_norm": 1.1807655262100478,
|
| 1519 |
+
"learning_rate": 3.3114422445681425e-05,
|
| 1520 |
+
"loss": 0.3136,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.837037037037037,
|
| 1525 |
+
"grad_norm": 1.2349063591796927,
|
| 1526 |
+
"learning_rate": 3.269778675244684e-05,
|
| 1527 |
+
"loss": 0.2882,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.8455026455026455,
|
| 1532 |
+
"grad_norm": 1.0765110541279164,
|
| 1533 |
+
"learning_rate": 3.22825114267364e-05,
|
| 1534 |
+
"loss": 0.295,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.853968253968254,
|
| 1539 |
+
"grad_norm": 1.220972473914235,
|
| 1540 |
+
"learning_rate": 3.1868629119116794e-05,
|
| 1541 |
+
"loss": 0.3176,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.8624338624338623,
|
| 1546 |
+
"grad_norm": 1.1122856433998665,
|
| 1547 |
+
"learning_rate": 3.1456172370630124e-05,
|
| 1548 |
+
"loss": 0.3324,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.870899470899471,
|
| 1553 |
+
"grad_norm": 1.1724923812151078,
|
| 1554 |
+
"learning_rate": 3.10451736102355e-05,
|
| 1555 |
+
"loss": 0.3091,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.8793650793650793,
|
| 1560 |
+
"grad_norm": 1.1020637891223475,
|
| 1561 |
+
"learning_rate": 3.0635665152259295e-05,
|
| 1562 |
+
"loss": 0.2955,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.8878306878306879,
|
| 1567 |
+
"grad_norm": 1.0318878630934867,
|
| 1568 |
+
"learning_rate": 3.0227679193854513e-05,
|
| 1569 |
+
"loss": 0.3195,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.8962962962962964,
|
| 1574 |
+
"grad_norm": 1.1032921856519728,
|
| 1575 |
+
"learning_rate": 2.982124781246932e-05,
|
| 1576 |
+
"loss": 0.3039,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.9047619047619047,
|
| 1581 |
+
"grad_norm": 1.131389139262005,
|
| 1582 |
+
"learning_rate": 2.9416402963324896e-05,
|
| 1583 |
+
"loss": 0.2742,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.9132275132275134,
|
| 1588 |
+
"grad_norm": 1.1298513214593129,
|
| 1589 |
+
"learning_rate": 2.9013176476903147e-05,
|
| 1590 |
+
"loss": 0.3048,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.9216931216931217,
|
| 1595 |
+
"grad_norm": 1.081531730138643,
|
| 1596 |
+
"learning_rate": 2.8611600056443944e-05,
|
| 1597 |
+
"loss": 0.3206,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.9301587301587302,
|
| 1602 |
+
"grad_norm": 1.1056344145079804,
|
| 1603 |
+
"learning_rate": 2.821170527545254e-05,
|
| 1604 |
+
"loss": 0.2952,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.9386243386243387,
|
| 1609 |
+
"grad_norm": 1.144349889481454,
|
| 1610 |
+
"learning_rate": 2.7813523575217103e-05,
|
| 1611 |
+
"loss": 0.3037,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.947089947089947,
|
| 1616 |
+
"grad_norm": 1.187655971159083,
|
| 1617 |
+
"learning_rate": 2.7417086262336777e-05,
|
| 1618 |
+
"loss": 0.3054,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.9555555555555557,
|
| 1623 |
+
"grad_norm": 1.1111194236831317,
|
| 1624 |
+
"learning_rate": 2.7022424506260113e-05,
|
| 1625 |
+
"loss": 0.3161,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.964021164021164,
|
| 1630 |
+
"grad_norm": 1.0599427922691493,
|
| 1631 |
+
"learning_rate": 2.662956933683443e-05,
|
| 1632 |
+
"loss": 0.3171,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.9724867724867725,
|
| 1637 |
+
"grad_norm": 1.0305400657350645,
|
| 1638 |
+
"learning_rate": 2.623855164186625e-05,
|
| 1639 |
+
"loss": 0.2773,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.980952380952381,
|
| 1644 |
+
"grad_norm": 1.0485786323318997,
|
| 1645 |
+
"learning_rate": 2.5849402164692593e-05,
|
| 1646 |
+
"loss": 0.3122,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.9894179894179893,
|
| 1651 |
+
"grad_norm": 1.0781221238793617,
|
| 1652 |
+
"learning_rate": 2.546215150176394e-05,
|
| 1653 |
+
"loss": 0.3002,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.997883597883598,
|
| 1658 |
+
"grad_norm": 1.1041144624563863,
|
| 1659 |
+
"learning_rate": 2.507683010023857e-05,
|
| 1660 |
+
"loss": 0.3006,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.0063492063492063,
|
| 1665 |
+
"grad_norm": 1.0896577752739438,
|
| 1666 |
+
"learning_rate": 2.4693468255588752e-05,
|
| 1667 |
+
"loss": 0.2681,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.0148148148148146,
|
| 1672 |
+
"grad_norm": 1.0383913047696638,
|
| 1673 |
+
"learning_rate": 2.4312096109218684e-05,
|
| 1674 |
+
"loss": 0.2509,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.0232804232804233,
|
| 1679 |
+
"grad_norm": 1.1401361260342018,
|
| 1680 |
+
"learning_rate": 2.3932743646094736e-05,
|
| 1681 |
+
"loss": 0.2495,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.0317460317460316,
|
| 1686 |
+
"grad_norm": 1.327049946404094,
|
| 1687 |
+
"learning_rate": 2.3555440692387947e-05,
|
| 1688 |
+
"loss": 0.2394,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.0402116402116404,
|
| 1693 |
+
"grad_norm": 1.1859201205267982,
|
| 1694 |
+
"learning_rate": 2.3180216913128872e-05,
|
| 1695 |
+
"loss": 0.2344,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.0486772486772487,
|
| 1700 |
+
"grad_norm": 1.1188101435575566,
|
| 1701 |
+
"learning_rate": 2.2807101809875275e-05,
|
| 1702 |
+
"loss": 0.243,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.057142857142857,
|
| 1707 |
+
"grad_norm": 1.2283454742009063,
|
| 1708 |
+
"learning_rate": 2.2436124718392565e-05,
|
| 1709 |
+
"loss": 0.248,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.0656084656084657,
|
| 1714 |
+
"grad_norm": 1.2592754589750133,
|
| 1715 |
+
"learning_rate": 2.206731480634739e-05,
|
| 1716 |
+
"loss": 0.2434,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.074074074074074,
|
| 1721 |
+
"grad_norm": 1.237235140900468,
|
| 1722 |
+
"learning_rate": 2.1700701071014222e-05,
|
| 1723 |
+
"loss": 0.2417,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.0825396825396827,
|
| 1728 |
+
"grad_norm": 1.1749453984216076,
|
| 1729 |
+
"learning_rate": 2.1336312336995572e-05,
|
| 1730 |
+
"loss": 0.2402,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.091005291005291,
|
| 1735 |
+
"grad_norm": 1.1524857608371186,
|
| 1736 |
+
"learning_rate": 2.0974177253955657e-05,
|
| 1737 |
+
"loss": 0.2476,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.0994708994708993,
|
| 1742 |
+
"grad_norm": 1.264906725996053,
|
| 1743 |
+
"learning_rate": 2.0614324294367882e-05,
|
| 1744 |
+
"loss": 0.242,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.107936507936508,
|
| 1749 |
+
"grad_norm": 1.1340174663623572,
|
| 1750 |
+
"learning_rate": 2.025678175127623e-05,
|
| 1751 |
+
"loss": 0.247,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.1164021164021163,
|
| 1756 |
+
"grad_norm": 1.2832254402303773,
|
| 1757 |
+
"learning_rate": 1.9901577736070598e-05,
|
| 1758 |
+
"loss": 0.2418,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.124867724867725,
|
| 1763 |
+
"grad_norm": 1.2771941330999887,
|
| 1764 |
+
"learning_rate": 1.954874017627684e-05,
|
| 1765 |
+
"loss": 0.2504,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.1333333333333333,
|
| 1770 |
+
"grad_norm": 1.1685751238482767,
|
| 1771 |
+
"learning_rate": 1.9198296813360777e-05,
|
| 1772 |
+
"loss": 0.2442,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.1417989417989416,
|
| 1777 |
+
"grad_norm": 1.3155782809463616,
|
| 1778 |
+
"learning_rate": 1.8850275200547142e-05,
|
| 1779 |
+
"loss": 0.2453,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.1502645502645503,
|
| 1784 |
+
"grad_norm": 1.2290797778685365,
|
| 1785 |
+
"learning_rate": 1.8504702700653194e-05,
|
| 1786 |
+
"loss": 0.2488,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.1587301587301586,
|
| 1791 |
+
"grad_norm": 1.140239798893309,
|
| 1792 |
+
"learning_rate": 1.8161606483937433e-05,
|
| 1793 |
+
"loss": 0.2377,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.1671957671957673,
|
| 1798 |
+
"grad_norm": 1.2500773194969288,
|
| 1799 |
+
"learning_rate": 1.7821013525963316e-05,
|
| 1800 |
+
"loss": 0.2571,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.1756613756613756,
|
| 1805 |
+
"grad_norm": 1.2713658534451882,
|
| 1806 |
+
"learning_rate": 1.748295060547823e-05,
|
| 1807 |
+
"loss": 0.2449,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.1841269841269844,
|
| 1812 |
+
"grad_norm": 1.227362647774226,
|
| 1813 |
+
"learning_rate": 1.7147444302308273e-05,
|
| 1814 |
+
"loss": 0.2348,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.1925925925925926,
|
| 1819 |
+
"grad_norm": 1.3167422374415048,
|
| 1820 |
+
"learning_rate": 1.6814520995268207e-05,
|
| 1821 |
+
"loss": 0.2578,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.201058201058201,
|
| 1826 |
+
"grad_norm": 1.2496998676048467,
|
| 1827 |
+
"learning_rate": 1.6484206860087653e-05,
|
| 1828 |
+
"loss": 0.2358,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.2095238095238097,
|
| 1833 |
+
"grad_norm": 1.3111014362047553,
|
| 1834 |
+
"learning_rate": 1.6156527867352845e-05,
|
| 1835 |
+
"loss": 0.2288,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.217989417989418,
|
| 1840 |
+
"grad_norm": 1.2159616363632633,
|
| 1841 |
+
"learning_rate": 1.5831509780464947e-05,
|
| 1842 |
+
"loss": 0.2478,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.2264550264550262,
|
| 1847 |
+
"grad_norm": 1.2868454216177594,
|
| 1848 |
+
"learning_rate": 1.5509178153614233e-05,
|
| 1849 |
+
"loss": 0.2462,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.234920634920635,
|
| 1854 |
+
"grad_norm": 1.1236506253498473,
|
| 1855 |
+
"learning_rate": 1.5189558329771092e-05,
|
| 1856 |
+
"loss": 0.2475,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.2433862433862433,
|
| 1861 |
+
"grad_norm": 1.2767351170017573,
|
| 1862 |
+
"learning_rate": 1.48726754386933e-05,
|
| 1863 |
+
"loss": 0.2577,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.251851851851852,
|
| 1868 |
+
"grad_norm": 1.2742999056295827,
|
| 1869 |
+
"learning_rate": 1.4558554394950319e-05,
|
| 1870 |
+
"loss": 0.2257,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.2603174603174603,
|
| 1875 |
+
"grad_norm": 1.1388456844003385,
|
| 1876 |
+
"learning_rate": 1.4247219895964458e-05,
|
| 1877 |
+
"loss": 0.2366,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.268783068783069,
|
| 1882 |
+
"grad_norm": 1.1420407981573506,
|
| 1883 |
+
"learning_rate": 1.3938696420068887e-05,
|
| 1884 |
+
"loss": 0.246,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.2772486772486773,
|
| 1889 |
+
"grad_norm": 1.2447613234115074,
|
| 1890 |
+
"learning_rate": 1.3633008224583276e-05,
|
| 1891 |
+
"loss": 0.2362,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.2857142857142856,
|
| 1896 |
+
"grad_norm": 1.2110210828346344,
|
| 1897 |
+
"learning_rate": 1.3330179343906424e-05,
|
| 1898 |
+
"loss": 0.2271,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.2941798941798943,
|
| 1903 |
+
"grad_norm": 1.1743235820264755,
|
| 1904 |
+
"learning_rate": 1.303023358762669e-05,
|
| 1905 |
+
"loss": 0.2349,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.3026455026455026,
|
| 1910 |
+
"grad_norm": 1.2896552909609165,
|
| 1911 |
+
"learning_rate": 1.27331945386499e-05,
|
| 1912 |
+
"loss": 0.2586,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.311111111111111,
|
| 1917 |
+
"grad_norm": 1.192957421343539,
|
| 1918 |
+
"learning_rate": 1.2439085551345209e-05,
|
| 1919 |
+
"loss": 0.2417,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.3195767195767196,
|
| 1924 |
+
"grad_norm": 1.3450432316093512,
|
| 1925 |
+
"learning_rate": 1.2147929749708931e-05,
|
| 1926 |
+
"loss": 0.2412,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.328042328042328,
|
| 1931 |
+
"grad_norm": 1.3062883724180023,
|
| 1932 |
+
"learning_rate": 1.1859750025546362e-05,
|
| 1933 |
+
"loss": 0.235,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.3365079365079366,
|
| 1938 |
+
"grad_norm": 1.1575476428826368,
|
| 1939 |
+
"learning_rate": 1.1574569036671978e-05,
|
| 1940 |
+
"loss": 0.2429,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.344973544973545,
|
| 1945 |
+
"grad_norm": 1.2906490240628528,
|
| 1946 |
+
"learning_rate": 1.129240920512798e-05,
|
| 1947 |
+
"loss": 0.2359,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.3534391534391537,
|
| 1952 |
+
"grad_norm": 1.3845442578956821,
|
| 1953 |
+
"learning_rate": 1.101329271542143e-05,
|
| 1954 |
+
"loss": 0.2382,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.361904761904762,
|
| 1959 |
+
"grad_norm": 1.2463846728950012,
|
| 1960 |
+
"learning_rate": 1.0737241512779928e-05,
|
| 1961 |
+
"loss": 0.2458,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.3703703703703702,
|
| 1966 |
+
"grad_norm": 1.2537500156225916,
|
| 1967 |
+
"learning_rate": 1.046427730142625e-05,
|
| 1968 |
+
"loss": 0.2487,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.378835978835979,
|
| 1973 |
+
"grad_norm": 1.3091633501767908,
|
| 1974 |
+
"learning_rate": 1.0194421542871918e-05,
|
| 1975 |
+
"loss": 0.2369,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.3873015873015873,
|
| 1980 |
+
"grad_norm": 1.1820540371459332,
|
| 1981 |
+
"learning_rate": 9.927695454229713e-06,
|
| 1982 |
+
"loss": 0.2448,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.3957671957671955,
|
| 1987 |
+
"grad_norm": 1.2731529604104752,
|
| 1988 |
+
"learning_rate": 9.664120006545573e-06,
|
| 1989 |
+
"loss": 0.2419,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.4042328042328043,
|
| 1994 |
+
"grad_norm": 1.3027428739822569,
|
| 1995 |
+
"learning_rate": 9.403715923149742e-06,
|
| 1996 |
+
"loss": 0.2551,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.4126984126984126,
|
| 2001 |
+
"grad_norm": 1.3254079829534025,
|
| 2002 |
+
"learning_rate": 9.146503678027451e-06,
|
| 2003 |
+
"loss": 0.2178,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.4211640211640213,
|
| 2008 |
+
"grad_norm": 1.1613878892535392,
|
| 2009 |
+
"learning_rate": 8.892503494209132e-06,
|
| 2010 |
+
"loss": 0.227,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.4296296296296296,
|
| 2015 |
+
"grad_norm": 1.0705896579237715,
|
| 2016 |
+
"learning_rate": 8.641735342180402e-06,
|
| 2017 |
+
"loss": 0.2449,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.4380952380952383,
|
| 2022 |
+
"grad_norm": 1.2593614315530395,
|
| 2023 |
+
"learning_rate": 8.394218938311982e-06,
|
| 2024 |
+
"loss": 0.2287,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.4465608465608466,
|
| 2029 |
+
"grad_norm": 1.1359699627950683,
|
| 2030 |
+
"learning_rate": 8.149973743309414e-06,
|
| 2031 |
+
"loss": 0.2488,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.455026455026455,
|
| 2036 |
+
"grad_norm": 1.3447659451396483,
|
| 2037 |
+
"learning_rate": 7.909018960683052e-06,
|
| 2038 |
+
"loss": 0.2339,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.4634920634920636,
|
| 2043 |
+
"grad_norm": 1.056442260887814,
|
| 2044 |
+
"learning_rate": 7.671373535238168e-06,
|
| 2045 |
+
"loss": 0.25,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.471957671957672,
|
| 2050 |
+
"grad_norm": 1.240966774372524,
|
| 2051 |
+
"learning_rate": 7.437056151585514e-06,
|
| 2052 |
+
"loss": 0.2348,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.48042328042328,
|
| 2057 |
+
"grad_norm": 1.1279901312845086,
|
| 2058 |
+
"learning_rate": 7.206085232672155e-06,
|
| 2059 |
+
"loss": 0.2319,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.488888888888889,
|
| 2064 |
+
"grad_norm": 1.3442581999609493,
|
| 2065 |
+
"learning_rate": 6.978478938333055e-06,
|
| 2066 |
+
"loss": 0.23,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.497354497354497,
|
| 2071 |
+
"grad_norm": 1.3087593444833816,
|
| 2072 |
+
"learning_rate": 6.75425516386326e-06,
|
| 2073 |
+
"loss": 0.2571,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.505820105820106,
|
| 2078 |
+
"grad_norm": 1.2133121259337931,
|
| 2079 |
+
"learning_rate": 6.533431538610923e-06,
|
| 2080 |
+
"loss": 0.2291,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.5142857142857142,
|
| 2085 |
+
"grad_norm": 1.1919153490283616,
|
| 2086 |
+
"learning_rate": 6.3160254245911825e-06,
|
| 2087 |
+
"loss": 0.2393,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.522751322751323,
|
| 2092 |
+
"grad_norm": 1.4097002557717668,
|
| 2093 |
+
"learning_rate": 6.102053915121092e-06,
|
| 2094 |
+
"loss": 0.2595,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.5312169312169313,
|
| 2099 |
+
"grad_norm": 1.3516367010865522,
|
| 2100 |
+
"learning_rate": 5.891533833475732e-06,
|
| 2101 |
+
"loss": 0.2505,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.5396825396825395,
|
| 2106 |
+
"grad_norm": 1.250969302934421,
|
| 2107 |
+
"learning_rate": 5.684481731565433e-06,
|
| 2108 |
+
"loss": 0.2415,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.5481481481481483,
|
| 2113 |
+
"grad_norm": 1.136756466811525,
|
| 2114 |
+
"learning_rate": 5.48091388863442e-06,
|
| 2115 |
+
"loss": 0.2325,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.5566137566137566,
|
| 2120 |
+
"grad_norm": 1.231370156024587,
|
| 2121 |
+
"learning_rate": 5.280846309980892e-06,
|
| 2122 |
+
"loss": 0.2462,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.565079365079365,
|
| 2127 |
+
"grad_norm": 1.3326225138678371,
|
| 2128 |
+
"learning_rate": 5.0842947256986014e-06,
|
| 2129 |
+
"loss": 0.2412,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.5735449735449736,
|
| 2134 |
+
"grad_norm": 1.1660313881807405,
|
| 2135 |
+
"learning_rate": 4.891274589440148e-06,
|
| 2136 |
+
"loss": 0.2448,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.582010582010582,
|
| 2141 |
+
"grad_norm": 1.3191473663837276,
|
| 2142 |
+
"learning_rate": 4.7018010772018175e-06,
|
| 2143 |
+
"loss": 0.2436,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.5904761904761906,
|
| 2148 |
+
"grad_norm": 1.2540952812205302,
|
| 2149 |
+
"learning_rate": 4.515889086130565e-06,
|
| 2150 |
+
"loss": 0.2479,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.598941798941799,
|
| 2155 |
+
"grad_norm": 1.2017713792968152,
|
| 2156 |
+
"learning_rate": 4.333553233352589e-06,
|
| 2157 |
+
"loss": 0.2566,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.6074074074074076,
|
| 2162 |
+
"grad_norm": 1.29908123750953,
|
| 2163 |
+
"learning_rate": 4.154807854824194e-06,
|
| 2164 |
+
"loss": 0.2602,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.615873015873016,
|
| 2169 |
+
"grad_norm": 1.4243808220545542,
|
| 2170 |
+
"learning_rate": 3.979667004204535e-06,
|
| 2171 |
+
"loss": 0.2398,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.624338624338624,
|
| 2176 |
+
"grad_norm": 1.3457308926738438,
|
| 2177 |
+
"learning_rate": 3.808144451750756e-06,
|
| 2178 |
+
"loss": 0.2308,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.632804232804233,
|
| 2183 |
+
"grad_norm": 1.3407484748421956,
|
| 2184 |
+
"learning_rate": 3.640253683235262e-06,
|
| 2185 |
+
"loss": 0.2511,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.641269841269841,
|
| 2190 |
+
"grad_norm": 1.2093265189235727,
|
| 2191 |
+
"learning_rate": 3.4760078988854305e-06,
|
| 2192 |
+
"loss": 0.2335,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.6497354497354495,
|
| 2197 |
+
"grad_norm": 1.287221034400132,
|
| 2198 |
+
"learning_rate": 3.315420012345777e-06,
|
| 2199 |
+
"loss": 0.2149,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.6582010582010582,
|
| 2204 |
+
"grad_norm": 1.191374675667794,
|
| 2205 |
+
"learning_rate": 3.1585026496625925e-06,
|
| 2206 |
+
"loss": 0.2462,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.6666666666666665,
|
| 2211 |
+
"grad_norm": 1.2788639981032432,
|
| 2212 |
+
"learning_rate": 3.005268148291296e-06,
|
| 2213 |
+
"loss": 0.2404,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.6751322751322753,
|
| 2218 |
+
"grad_norm": 1.4725222188972884,
|
| 2219 |
+
"learning_rate": 2.855728556126319e-06,
|
| 2220 |
+
"loss": 0.2345,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.6835978835978835,
|
| 2225 |
+
"grad_norm": 1.1514884763928235,
|
| 2226 |
+
"learning_rate": 2.709895630553977e-06,
|
| 2227 |
+
"loss": 0.2247,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.6920634920634923,
|
| 2232 |
+
"grad_norm": 1.2744441762954501,
|
| 2233 |
+
"learning_rate": 2.5677808375279443e-06,
|
| 2234 |
+
"loss": 0.2418,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.7005291005291006,
|
| 2239 |
+
"grad_norm": 1.3232083105521923,
|
| 2240 |
+
"learning_rate": 2.42939535066784e-06,
|
| 2241 |
+
"loss": 0.2304,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.708994708994709,
|
| 2246 |
+
"grad_norm": 1.270314937268037,
|
| 2247 |
+
"learning_rate": 2.294750050380662e-06,
|
| 2248 |
+
"loss": 0.2247,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.7174603174603176,
|
| 2253 |
+
"grad_norm": 1.256698656991292,
|
| 2254 |
+
"learning_rate": 2.1638555230053315e-06,
|
| 2255 |
+
"loss": 0.2406,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.725925925925926,
|
| 2260 |
+
"grad_norm": 1.3193829744509844,
|
| 2261 |
+
"learning_rate": 2.036722059980406e-06,
|
| 2262 |
+
"loss": 0.2396,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.734391534391534,
|
| 2267 |
+
"grad_norm": 1.4009607878677235,
|
| 2268 |
+
"learning_rate": 1.9133596570348468e-06,
|
| 2269 |
+
"loss": 0.2329,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.742857142857143,
|
| 2274 |
+
"grad_norm": 1.29045787788227,
|
| 2275 |
+
"learning_rate": 1.7937780134021808e-06,
|
| 2276 |
+
"loss": 0.2301,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.751322751322751,
|
| 2281 |
+
"grad_norm": 1.2476131982200767,
|
| 2282 |
+
"learning_rate": 1.6779865310578623e-06,
|
| 2283 |
+
"loss": 0.2284,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.75978835978836,
|
| 2288 |
+
"grad_norm": 1.0670988501341874,
|
| 2289 |
+
"learning_rate": 1.5659943139801126e-06,
|
| 2290 |
+
"loss": 0.2406,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.768253968253968,
|
| 2295 |
+
"grad_norm": 1.308449437483495,
|
| 2296 |
+
"learning_rate": 1.4578101674340372e-06,
|
| 2297 |
+
"loss": 0.2332,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.776719576719577,
|
| 2302 |
+
"grad_norm": 1.4177869984435298,
|
| 2303 |
+
"learning_rate": 1.3534425972794195e-06,
|
| 2304 |
+
"loss": 0.2233,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.785185185185185,
|
| 2309 |
+
"grad_norm": 1.3267608082684277,
|
| 2310 |
+
"learning_rate": 1.2528998093019051e-06,
|
| 2311 |
+
"loss": 0.2498,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.7936507936507935,
|
| 2316 |
+
"grad_norm": 1.2882825046480746,
|
| 2317 |
+
"learning_rate": 1.1561897085678341e-06,
|
| 2318 |
+
"loss": 0.2286,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.8021164021164022,
|
| 2323 |
+
"grad_norm": 1.2774964089508063,
|
| 2324 |
+
"learning_rate": 1.0633198988027115e-06,
|
| 2325 |
+
"loss": 0.2253,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.8105820105820105,
|
| 2330 |
+
"grad_norm": 1.3527960705502262,
|
| 2331 |
+
"learning_rate": 9.742976817933847e-07,
|
| 2332 |
+
"loss": 0.24,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.819047619047619,
|
| 2337 |
+
"grad_norm": 1.365122676946633,
|
| 2338 |
+
"learning_rate": 8.891300568139638e-07,
|
| 2339 |
+
"loss": 0.2592,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.8275132275132275,
|
| 2344 |
+
"grad_norm": 1.2427303633572429,
|
| 2345 |
+
"learning_rate": 8.078237200754624e-07,
|
| 2346 |
+
"loss": 0.2474,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.835978835978836,
|
| 2351 |
+
"grad_norm": 1.144641112857142,
|
| 2352 |
+
"learning_rate": 7.30385064199357e-07,
|
| 2353 |
+
"loss": 0.2332,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.8444444444444446,
|
| 2358 |
+
"grad_norm": 1.2745498061485745,
|
| 2359 |
+
"learning_rate": 6.568201777149729e-07,
|
| 2360 |
+
"loss": 0.251,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.852910052910053,
|
| 2365 |
+
"grad_norm": 1.422987862547148,
|
| 2366 |
+
"learning_rate": 5.871348445807501e-07,
|
| 2367 |
+
"loss": 0.2449,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.8613756613756616,
|
| 2372 |
+
"grad_norm": 1.0332383237986493,
|
| 2373 |
+
"learning_rate": 5.213345437295181e-07,
|
| 2374 |
+
"loss": 0.2467,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.86984126984127,
|
| 2379 |
+
"grad_norm": 1.4294499006460313,
|
| 2380 |
+
"learning_rate": 4.594244486376964e-07,
|
| 2381 |
+
"loss": 0.2396,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.878306878306878,
|
| 2386 |
+
"grad_norm": 1.3112777173225585,
|
| 2387 |
+
"learning_rate": 4.014094269185642e-07,
|
| 2388 |
+
"loss": 0.2386,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.886772486772487,
|
| 2393 |
+
"grad_norm": 1.2628517758813262,
|
| 2394 |
+
"learning_rate": 3.472940399395108e-07,
|
| 2395 |
+
"loss": 0.2229,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.895238095238095,
|
| 2400 |
+
"grad_norm": 1.235621137865247,
|
| 2401 |
+
"learning_rate": 2.9708254246344494e-07,
|
| 2402 |
+
"loss": 0.2386,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.9037037037037035,
|
| 2407 |
+
"grad_norm": 1.1595358547316623,
|
| 2408 |
+
"learning_rate": 2.5077888231422873e-07,
|
| 2409 |
+
"loss": 0.2407,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.912169312169312,
|
| 2414 |
+
"grad_norm": 1.1332523589105532,
|
| 2415 |
+
"learning_rate": 2.0838670006632643e-07,
|
| 2416 |
+
"loss": 0.224,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.9206349206349205,
|
| 2421 |
+
"grad_norm": 1.431443541686851,
|
| 2422 |
+
"learning_rate": 1.699093287585274e-07,
|
| 2423 |
+
"loss": 0.2334,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.929100529100529,
|
| 2428 |
+
"grad_norm": 1.2372896845957517,
|
| 2429 |
+
"learning_rate": 1.3534979363191701e-07,
|
| 2430 |
+
"loss": 0.2471,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.9375661375661375,
|
| 2435 |
+
"grad_norm": 1.296014823220347,
|
| 2436 |
+
"learning_rate": 1.0471081189200593e-07,
|
| 2437 |
+
"loss": 0.2205,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.9460317460317462,
|
| 2442 |
+
"grad_norm": 1.2004684728241466,
|
| 2443 |
+
"learning_rate": 7.799479249510633e-08,
|
| 2444 |
+
"loss": 0.2376,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.9544973544973545,
|
| 2449 |
+
"grad_norm": 1.2399487682175703,
|
| 2450 |
+
"learning_rate": 5.520383595891132e-08,
|
| 2451 |
+
"loss": 0.2289,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.962962962962963,
|
| 2456 |
+
"grad_norm": 1.439024332184667,
|
| 2457 |
+
"learning_rate": 3.6339734197360365e-08,
|
| 2458 |
+
"loss": 0.2487,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.9714285714285715,
|
| 2463 |
+
"grad_norm": 1.1880176062141694,
|
| 2464 |
+
"learning_rate": 2.14039703797575e-08,
|
| 2465 |
+
"loss": 0.2373,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.97989417989418,
|
| 2470 |
+
"grad_norm": 1.2924381706571177,
|
| 2471 |
+
"learning_rate": 1.0397718814125768e-08,
|
| 2472 |
+
"loss": 0.2468,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.988359788359788,
|
| 2477 |
+
"grad_norm": 1.382013827064607,
|
| 2478 |
+
"learning_rate": 3.321844854925482e-09,
|
| 2479 |
+
"loss": 0.2325,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.996825396825397,
|
| 2484 |
+
"grad_norm": 1.2057884625346706,
|
| 2485 |
+
"learning_rate": 1.7690483496979326e-10,
|
| 2486 |
+
"loss": 0.2409,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.9993650793650795,
|
| 2491 |
+
"step": 3543,
|
| 2492 |
+
"total_flos": 652111388999680.0,
|
| 2493 |
+
"train_loss": 0.34479539529973763,
|
| 2494 |
+
"train_runtime": 32429.3956,
|
| 2495 |
+
"train_samples_per_second": 0.874,
|
| 2496 |
+
"train_steps_per_second": 0.109
|
| 2497 |
+
}
|
| 2498 |
+
],
|
| 2499 |
+
"logging_steps": 10,
|
| 2500 |
+
"max_steps": 3543,
|
| 2501 |
+
"num_input_tokens_seen": 0,
|
| 2502 |
+
"num_train_epochs": 3,
|
| 2503 |
+
"save_steps": 500,
|
| 2504 |
+
"stateful_callbacks": {
|
| 2505 |
+
"TrainerControl": {
|
| 2506 |
+
"args": {
|
| 2507 |
+
"should_epoch_stop": false,
|
| 2508 |
+
"should_evaluate": false,
|
| 2509 |
+
"should_log": false,
|
| 2510 |
+
"should_save": true,
|
| 2511 |
+
"should_training_stop": true
|
| 2512 |
+
},
|
| 2513 |
+
"attributes": {}
|
| 2514 |
+
}
|
| 2515 |
+
},
|
| 2516 |
+
"total_flos": 652111388999680.0,
|
| 2517 |
+
"train_batch_size": 1,
|
| 2518 |
+
"trial_name": null,
|
| 2519 |
+
"trial_params": null
|
| 2520 |
+
}
|
training_loss.png
ADDED
|