Commit
•
3ddb93f
1
Parent(s):
d9187be
End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +2236 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: mistralai/Mistral-7B-v0.1
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: dcft-orca-agentinstruct-1M-v1-cleaned_mistral
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# dcft-orca-agentinstruct-1M-v1-cleaned_mistral
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.4468
|
| 21 |
|
|
|
|
| 4 |
base_model: mistralai/Mistral-7B-v0.1
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: dcft-orca-agentinstruct-1M-v1-cleaned_mistral
|
|
|
|
| 16 |
|
| 17 |
# dcft-orca-agentinstruct-1M-v1-cleaned_mistral
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mlfoundations-dev/dcft-orca-agentinstruct-1M-v1-cleaned dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.4468
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.998551424432641,
|
| 3 |
+
"eval_loss": 0.4467960000038147,
|
| 4 |
+
"eval_runtime": 180.8976,
|
| 5 |
+
"eval_samples_per_second": 154.248,
|
| 6 |
+
"eval_steps_per_second": 0.603,
|
| 7 |
+
"total_flos": 5200153128468480.0,
|
| 8 |
+
"train_loss": 0.3273588392085691,
|
| 9 |
+
"train_runtime": 26340.0491,
|
| 10 |
+
"train_samples_per_second": 60.381,
|
| 11 |
+
"train_steps_per_second": 0.118
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.998551424432641,
|
| 3 |
+
"eval_loss": 0.4467960000038147,
|
| 4 |
+
"eval_runtime": 180.8976,
|
| 5 |
+
"eval_samples_per_second": 154.248,
|
| 6 |
+
"eval_steps_per_second": 0.603
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.998551424432641,
|
| 3 |
+
"total_flos": 5200153128468480.0,
|
| 4 |
+
"train_loss": 0.3273588392085691,
|
| 5 |
+
"train_runtime": 26340.0491,
|
| 6 |
+
"train_samples_per_second": 60.381,
|
| 7 |
+
"train_steps_per_second": 0.118
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2236 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.998551424432641,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3105,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.009657170449058426,
|
| 13 |
+
"grad_norm": 3.7852027137880246,
|
| 14 |
+
"learning_rate": 5e-06,
|
| 15 |
+
"loss": 0.6254,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01931434089811685,
|
| 20 |
+
"grad_norm": 2.0620749606772866,
|
| 21 |
+
"learning_rate": 5e-06,
|
| 22 |
+
"loss": 0.5439,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.028971511347175277,
|
| 27 |
+
"grad_norm": 2.085747916186358,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.5216,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0386286817962337,
|
| 34 |
+
"grad_norm": 1.8268450890377532,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 0.5097,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04828585224529213,
|
| 41 |
+
"grad_norm": 2.0581630455526123,
|
| 42 |
+
"learning_rate": 5e-06,
|
| 43 |
+
"loss": 0.504,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05794302269435055,
|
| 48 |
+
"grad_norm": 1.602128099496524,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.5013,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06760019314340898,
|
| 55 |
+
"grad_norm": 1.6673623755384994,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.4878,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0772573635924674,
|
| 62 |
+
"grad_norm": 1.8459613232584249,
|
| 63 |
+
"learning_rate": 5e-06,
|
| 64 |
+
"loss": 0.4878,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.08691453404152583,
|
| 69 |
+
"grad_norm": 1.93193366344165,
|
| 70 |
+
"learning_rate": 5e-06,
|
| 71 |
+
"loss": 0.4879,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.09657170449058426,
|
| 76 |
+
"grad_norm": 1.5465868737214155,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.4808,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.10622887493964268,
|
| 83 |
+
"grad_norm": 1.6271945980777192,
|
| 84 |
+
"learning_rate": 5e-06,
|
| 85 |
+
"loss": 0.4782,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1158860453887011,
|
| 90 |
+
"grad_norm": 1.8402435224859226,
|
| 91 |
+
"learning_rate": 5e-06,
|
| 92 |
+
"loss": 0.4767,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.12554321583775954,
|
| 97 |
+
"grad_norm": 1.7618646649255147,
|
| 98 |
+
"learning_rate": 5e-06,
|
| 99 |
+
"loss": 0.4745,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.13520038628681796,
|
| 104 |
+
"grad_norm": 1.5315385028605073,
|
| 105 |
+
"learning_rate": 5e-06,
|
| 106 |
+
"loss": 0.4733,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.14485755673587639,
|
| 111 |
+
"grad_norm": 1.6814511784859512,
|
| 112 |
+
"learning_rate": 5e-06,
|
| 113 |
+
"loss": 0.4733,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1545147271849348,
|
| 118 |
+
"grad_norm": 1.5271479672490604,
|
| 119 |
+
"learning_rate": 5e-06,
|
| 120 |
+
"loss": 0.4716,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.16417189763399323,
|
| 125 |
+
"grad_norm": 1.6777577770303629,
|
| 126 |
+
"learning_rate": 5e-06,
|
| 127 |
+
"loss": 0.471,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.17382906808305165,
|
| 132 |
+
"grad_norm": 1.5955962931770318,
|
| 133 |
+
"learning_rate": 5e-06,
|
| 134 |
+
"loss": 0.4733,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1834862385321101,
|
| 139 |
+
"grad_norm": 1.6328687609928156,
|
| 140 |
+
"learning_rate": 5e-06,
|
| 141 |
+
"loss": 0.4711,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.19314340898116852,
|
| 146 |
+
"grad_norm": 1.6077537949384313,
|
| 147 |
+
"learning_rate": 5e-06,
|
| 148 |
+
"loss": 0.4669,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.20280057943022695,
|
| 153 |
+
"grad_norm": 1.5250866943562713,
|
| 154 |
+
"learning_rate": 5e-06,
|
| 155 |
+
"loss": 0.4647,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21245774987928537,
|
| 160 |
+
"grad_norm": 1.4889880734895962,
|
| 161 |
+
"learning_rate": 5e-06,
|
| 162 |
+
"loss": 0.4585,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2221149203283438,
|
| 167 |
+
"grad_norm": 1.6089621732620885,
|
| 168 |
+
"learning_rate": 5e-06,
|
| 169 |
+
"loss": 0.4578,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2317720907774022,
|
| 174 |
+
"grad_norm": 1.4899292253368934,
|
| 175 |
+
"learning_rate": 5e-06,
|
| 176 |
+
"loss": 0.4555,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.24142926122646063,
|
| 181 |
+
"grad_norm": 1.5257308623369976,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 0.4663,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2510864316755191,
|
| 188 |
+
"grad_norm": 1.8459225661699417,
|
| 189 |
+
"learning_rate": 5e-06,
|
| 190 |
+
"loss": 0.4588,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2607436021245775,
|
| 195 |
+
"grad_norm": 1.723748524461411,
|
| 196 |
+
"learning_rate": 5e-06,
|
| 197 |
+
"loss": 0.4622,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.27040077257363593,
|
| 202 |
+
"grad_norm": 1.6111594714193933,
|
| 203 |
+
"learning_rate": 5e-06,
|
| 204 |
+
"loss": 0.4562,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2800579430226944,
|
| 209 |
+
"grad_norm": 1.489881509289716,
|
| 210 |
+
"learning_rate": 5e-06,
|
| 211 |
+
"loss": 0.4632,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.28971511347175277,
|
| 216 |
+
"grad_norm": 1.7437076586484017,
|
| 217 |
+
"learning_rate": 5e-06,
|
| 218 |
+
"loss": 0.4498,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2993722839208112,
|
| 223 |
+
"grad_norm": 1.5023208696211863,
|
| 224 |
+
"learning_rate": 5e-06,
|
| 225 |
+
"loss": 0.4547,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3090294543698696,
|
| 230 |
+
"grad_norm": 1.47145559032755,
|
| 231 |
+
"learning_rate": 5e-06,
|
| 232 |
+
"loss": 0.448,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.31868662481892807,
|
| 237 |
+
"grad_norm": 1.493111331589544,
|
| 238 |
+
"learning_rate": 5e-06,
|
| 239 |
+
"loss": 0.4493,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.32834379526798646,
|
| 244 |
+
"grad_norm": 1.3731787738087704,
|
| 245 |
+
"learning_rate": 5e-06,
|
| 246 |
+
"loss": 0.4548,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3380009657170449,
|
| 251 |
+
"grad_norm": 1.5097895917724191,
|
| 252 |
+
"learning_rate": 5e-06,
|
| 253 |
+
"loss": 0.4468,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3476581361661033,
|
| 258 |
+
"grad_norm": 1.3977565649576211,
|
| 259 |
+
"learning_rate": 5e-06,
|
| 260 |
+
"loss": 0.4497,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.35731530661516175,
|
| 265 |
+
"grad_norm": 1.597963004052136,
|
| 266 |
+
"learning_rate": 5e-06,
|
| 267 |
+
"loss": 0.446,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3669724770642202,
|
| 272 |
+
"grad_norm": 1.3464438194432793,
|
| 273 |
+
"learning_rate": 5e-06,
|
| 274 |
+
"loss": 0.4462,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3766296475132786,
|
| 279 |
+
"grad_norm": 2.13016639276551,
|
| 280 |
+
"learning_rate": 5e-06,
|
| 281 |
+
"loss": 0.4471,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.38628681796233705,
|
| 286 |
+
"grad_norm": 1.459345081562004,
|
| 287 |
+
"learning_rate": 5e-06,
|
| 288 |
+
"loss": 0.4449,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.39594398841139544,
|
| 293 |
+
"grad_norm": 1.4268521503710732,
|
| 294 |
+
"learning_rate": 5e-06,
|
| 295 |
+
"loss": 0.4484,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4056011588604539,
|
| 300 |
+
"grad_norm": 1.5213089527197126,
|
| 301 |
+
"learning_rate": 5e-06,
|
| 302 |
+
"loss": 0.442,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4152583293095123,
|
| 307 |
+
"grad_norm": 1.4107454080123742,
|
| 308 |
+
"learning_rate": 5e-06,
|
| 309 |
+
"loss": 0.4471,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.42491549975857074,
|
| 314 |
+
"grad_norm": 1.4224146665144493,
|
| 315 |
+
"learning_rate": 5e-06,
|
| 316 |
+
"loss": 0.4421,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.4345726702076292,
|
| 321 |
+
"grad_norm": 1.339648419205393,
|
| 322 |
+
"learning_rate": 5e-06,
|
| 323 |
+
"loss": 0.4459,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4442298406566876,
|
| 328 |
+
"grad_norm": 1.541975682641463,
|
| 329 |
+
"learning_rate": 5e-06,
|
| 330 |
+
"loss": 0.4416,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.45388701110574603,
|
| 335 |
+
"grad_norm": 1.447132284857592,
|
| 336 |
+
"learning_rate": 5e-06,
|
| 337 |
+
"loss": 0.4529,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4635441815548044,
|
| 342 |
+
"grad_norm": 1.385982642001702,
|
| 343 |
+
"learning_rate": 5e-06,
|
| 344 |
+
"loss": 0.4458,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4732013520038629,
|
| 349 |
+
"grad_norm": 2.0640182165227428,
|
| 350 |
+
"learning_rate": 5e-06,
|
| 351 |
+
"loss": 0.4426,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.48285852245292127,
|
| 356 |
+
"grad_norm": 1.348125498390211,
|
| 357 |
+
"learning_rate": 5e-06,
|
| 358 |
+
"loss": 0.4329,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4925156929019797,
|
| 363 |
+
"grad_norm": 1.509261123647944,
|
| 364 |
+
"learning_rate": 5e-06,
|
| 365 |
+
"loss": 0.4437,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5021728633510382,
|
| 370 |
+
"grad_norm": 2.7536862793707493,
|
| 371 |
+
"learning_rate": 5e-06,
|
| 372 |
+
"loss": 0.4359,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5118300338000966,
|
| 377 |
+
"grad_norm": 1.4717359737134839,
|
| 378 |
+
"learning_rate": 5e-06,
|
| 379 |
+
"loss": 0.4382,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.521487204249155,
|
| 384 |
+
"grad_norm": 1.442405927212471,
|
| 385 |
+
"learning_rate": 5e-06,
|
| 386 |
+
"loss": 0.4394,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5311443746982134,
|
| 391 |
+
"grad_norm": 1.3948076832604182,
|
| 392 |
+
"learning_rate": 5e-06,
|
| 393 |
+
"loss": 0.4407,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5408015451472719,
|
| 398 |
+
"grad_norm": 1.3948240431568104,
|
| 399 |
+
"learning_rate": 5e-06,
|
| 400 |
+
"loss": 0.4366,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5504587155963303,
|
| 405 |
+
"grad_norm": 1.417839986279595,
|
| 406 |
+
"learning_rate": 5e-06,
|
| 407 |
+
"loss": 0.4369,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5601158860453888,
|
| 412 |
+
"grad_norm": 1.5109506580715066,
|
| 413 |
+
"learning_rate": 5e-06,
|
| 414 |
+
"loss": 0.4353,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5697730564944471,
|
| 419 |
+
"grad_norm": 1.4806935542579658,
|
| 420 |
+
"learning_rate": 5e-06,
|
| 421 |
+
"loss": 0.4378,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5794302269435055,
|
| 426 |
+
"grad_norm": 1.5398960352117772,
|
| 427 |
+
"learning_rate": 5e-06,
|
| 428 |
+
"loss": 0.4352,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.589087397392564,
|
| 433 |
+
"grad_norm": 1.2925767785772035,
|
| 434 |
+
"learning_rate": 5e-06,
|
| 435 |
+
"loss": 0.4366,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5987445678416224,
|
| 440 |
+
"grad_norm": 1.5201357387636154,
|
| 441 |
+
"learning_rate": 5e-06,
|
| 442 |
+
"loss": 0.4296,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6084017382906808,
|
| 447 |
+
"grad_norm": 1.4717871703145071,
|
| 448 |
+
"learning_rate": 5e-06,
|
| 449 |
+
"loss": 0.431,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6180589087397392,
|
| 454 |
+
"grad_norm": 1.3687678242722554,
|
| 455 |
+
"learning_rate": 5e-06,
|
| 456 |
+
"loss": 0.4345,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6277160791887977,
|
| 461 |
+
"grad_norm": 1.3591593745283608,
|
| 462 |
+
"learning_rate": 5e-06,
|
| 463 |
+
"loss": 0.4313,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6373732496378561,
|
| 468 |
+
"grad_norm": 1.4717658524480044,
|
| 469 |
+
"learning_rate": 5e-06,
|
| 470 |
+
"loss": 0.4291,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6470304200869146,
|
| 475 |
+
"grad_norm": 1.398916059634764,
|
| 476 |
+
"learning_rate": 5e-06,
|
| 477 |
+
"loss": 0.4323,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6566875905359729,
|
| 482 |
+
"grad_norm": 1.4300679905819138,
|
| 483 |
+
"learning_rate": 5e-06,
|
| 484 |
+
"loss": 0.4271,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6663447609850314,
|
| 489 |
+
"grad_norm": 1.4886395606416152,
|
| 490 |
+
"learning_rate": 5e-06,
|
| 491 |
+
"loss": 0.4297,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6760019314340898,
|
| 496 |
+
"grad_norm": 1.5034980432589369,
|
| 497 |
+
"learning_rate": 5e-06,
|
| 498 |
+
"loss": 0.4272,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6856591018831483,
|
| 503 |
+
"grad_norm": 1.5962520697846152,
|
| 504 |
+
"learning_rate": 5e-06,
|
| 505 |
+
"loss": 0.4316,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.6953162723322066,
|
| 510 |
+
"grad_norm": 1.5447138904473863,
|
| 511 |
+
"learning_rate": 5e-06,
|
| 512 |
+
"loss": 0.4324,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7049734427812651,
|
| 517 |
+
"grad_norm": 1.3937453432394886,
|
| 518 |
+
"learning_rate": 5e-06,
|
| 519 |
+
"loss": 0.4311,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7146306132303235,
|
| 524 |
+
"grad_norm": 1.3011464123427467,
|
| 525 |
+
"learning_rate": 5e-06,
|
| 526 |
+
"loss": 0.4292,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.724287783679382,
|
| 531 |
+
"grad_norm": 1.4931973618415784,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.4295,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7339449541284404,
|
| 538 |
+
"grad_norm": 1.3582576932610695,
|
| 539 |
+
"learning_rate": 5e-06,
|
| 540 |
+
"loss": 0.4322,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7436021245774987,
|
| 545 |
+
"grad_norm": 1.5128036261098081,
|
| 546 |
+
"learning_rate": 5e-06,
|
| 547 |
+
"loss": 0.4306,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7532592950265572,
|
| 552 |
+
"grad_norm": 1.3047038324570048,
|
| 553 |
+
"learning_rate": 5e-06,
|
| 554 |
+
"loss": 0.43,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7629164654756156,
|
| 559 |
+
"grad_norm": 1.3637121209115217,
|
| 560 |
+
"learning_rate": 5e-06,
|
| 561 |
+
"loss": 0.4273,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.7725736359246741,
|
| 566 |
+
"grad_norm": 1.3919305669020587,
|
| 567 |
+
"learning_rate": 5e-06,
|
| 568 |
+
"loss": 0.4314,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.7822308063737325,
|
| 573 |
+
"grad_norm": 1.3941190921525122,
|
| 574 |
+
"learning_rate": 5e-06,
|
| 575 |
+
"loss": 0.4255,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7918879768227909,
|
| 580 |
+
"grad_norm": 1.362917573056491,
|
| 581 |
+
"learning_rate": 5e-06,
|
| 582 |
+
"loss": 0.4333,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8015451472718493,
|
| 587 |
+
"grad_norm": 1.415609597869972,
|
| 588 |
+
"learning_rate": 5e-06,
|
| 589 |
+
"loss": 0.4225,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.8112023177209078,
|
| 594 |
+
"grad_norm": 1.489818925031249,
|
| 595 |
+
"learning_rate": 5e-06,
|
| 596 |
+
"loss": 0.425,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.8208594881699662,
|
| 601 |
+
"grad_norm": 1.4658100940320271,
|
| 602 |
+
"learning_rate": 5e-06,
|
| 603 |
+
"loss": 0.4304,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8305166586190246,
|
| 608 |
+
"grad_norm": 1.3385021753495208,
|
| 609 |
+
"learning_rate": 5e-06,
|
| 610 |
+
"loss": 0.4263,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.840173829068083,
|
| 615 |
+
"grad_norm": 1.4519295448528378,
|
| 616 |
+
"learning_rate": 5e-06,
|
| 617 |
+
"loss": 0.4274,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8498309995171415,
|
| 622 |
+
"grad_norm": 1.5271445690200918,
|
| 623 |
+
"learning_rate": 5e-06,
|
| 624 |
+
"loss": 0.4244,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8594881699661999,
|
| 629 |
+
"grad_norm": 9.54446637099851,
|
| 630 |
+
"learning_rate": 5e-06,
|
| 631 |
+
"loss": 0.4294,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8691453404152584,
|
| 636 |
+
"grad_norm": 1.4822497861334036,
|
| 637 |
+
"learning_rate": 5e-06,
|
| 638 |
+
"loss": 0.4206,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.8788025108643167,
|
| 643 |
+
"grad_norm": 1.295865945244785,
|
| 644 |
+
"learning_rate": 5e-06,
|
| 645 |
+
"loss": 0.4223,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.8884596813133752,
|
| 650 |
+
"grad_norm": 1.4498222196547434,
|
| 651 |
+
"learning_rate": 5e-06,
|
| 652 |
+
"loss": 0.4222,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.8981168517624336,
|
| 657 |
+
"grad_norm": 1.4109015353277008,
|
| 658 |
+
"learning_rate": 5e-06,
|
| 659 |
+
"loss": 0.424,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9077740222114921,
|
| 664 |
+
"grad_norm": 1.3782840436535198,
|
| 665 |
+
"learning_rate": 5e-06,
|
| 666 |
+
"loss": 0.4224,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.9174311926605505,
|
| 671 |
+
"grad_norm": 1.3492925117698618,
|
| 672 |
+
"learning_rate": 5e-06,
|
| 673 |
+
"loss": 0.4209,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9270883631096088,
|
| 678 |
+
"grad_norm": 1.3072263370484556,
|
| 679 |
+
"learning_rate": 5e-06,
|
| 680 |
+
"loss": 0.419,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9367455335586673,
|
| 685 |
+
"grad_norm": 1.3359087020152296,
|
| 686 |
+
"learning_rate": 5e-06,
|
| 687 |
+
"loss": 0.4213,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9464027040077257,
|
| 692 |
+
"grad_norm": 1.4027836366902715,
|
| 693 |
+
"learning_rate": 5e-06,
|
| 694 |
+
"loss": 0.4264,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9560598744567842,
|
| 699 |
+
"grad_norm": 1.3083915796128625,
|
| 700 |
+
"learning_rate": 5e-06,
|
| 701 |
+
"loss": 0.4213,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.9657170449058425,
|
| 706 |
+
"grad_norm": 1.283807856099459,
|
| 707 |
+
"learning_rate": 5e-06,
|
| 708 |
+
"loss": 0.4226,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.975374215354901,
|
| 713 |
+
"grad_norm": 1.318305259783526,
|
| 714 |
+
"learning_rate": 5e-06,
|
| 715 |
+
"loss": 0.4177,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.9850313858039594,
|
| 720 |
+
"grad_norm": 1.351081315674525,
|
| 721 |
+
"learning_rate": 5e-06,
|
| 722 |
+
"loss": 0.4184,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.9946885562530179,
|
| 727 |
+
"grad_norm": 1.286367123540384,
|
| 728 |
+
"learning_rate": 5e-06,
|
| 729 |
+
"loss": 0.417,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.9995171414775471,
|
| 734 |
+
"eval_loss": 0.4186817705631256,
|
| 735 |
+
"eval_runtime": 178.2472,
|
| 736 |
+
"eval_samples_per_second": 156.541,
|
| 737 |
+
"eval_steps_per_second": 0.612,
|
| 738 |
+
"step": 1035
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.0043457267020763,
|
| 742 |
+
"grad_norm": 2.083458671261006,
|
| 743 |
+
"learning_rate": 5e-06,
|
| 744 |
+
"loss": 0.3804,
|
| 745 |
+
"step": 1040
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.0140028971511348,
|
| 749 |
+
"grad_norm": 1.6710464918103058,
|
| 750 |
+
"learning_rate": 5e-06,
|
| 751 |
+
"loss": 0.3159,
|
| 752 |
+
"step": 1050
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.0236600676001932,
|
| 756 |
+
"grad_norm": 1.365743230792304,
|
| 757 |
+
"learning_rate": 5e-06,
|
| 758 |
+
"loss": 0.3138,
|
| 759 |
+
"step": 1060
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.0333172380492515,
|
| 763 |
+
"grad_norm": 1.39749368883128,
|
| 764 |
+
"learning_rate": 5e-06,
|
| 765 |
+
"loss": 0.3146,
|
| 766 |
+
"step": 1070
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.04297440849831,
|
| 770 |
+
"grad_norm": 1.5821963324713015,
|
| 771 |
+
"learning_rate": 5e-06,
|
| 772 |
+
"loss": 0.3101,
|
| 773 |
+
"step": 1080
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.0526315789473684,
|
| 777 |
+
"grad_norm": 1.6454840741114403,
|
| 778 |
+
"learning_rate": 5e-06,
|
| 779 |
+
"loss": 0.3152,
|
| 780 |
+
"step": 1090
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0622887493964268,
|
| 784 |
+
"grad_norm": 1.4356941510220327,
|
| 785 |
+
"learning_rate": 5e-06,
|
| 786 |
+
"loss": 0.3211,
|
| 787 |
+
"step": 1100
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.0719459198454853,
|
| 791 |
+
"grad_norm": 1.421015975147645,
|
| 792 |
+
"learning_rate": 5e-06,
|
| 793 |
+
"loss": 0.3185,
|
| 794 |
+
"step": 1110
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.0816030902945437,
|
| 798 |
+
"grad_norm": 1.3637757751998314,
|
| 799 |
+
"learning_rate": 5e-06,
|
| 800 |
+
"loss": 0.3177,
|
| 801 |
+
"step": 1120
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.0912602607436022,
|
| 805 |
+
"grad_norm": 1.4492110772881415,
|
| 806 |
+
"learning_rate": 5e-06,
|
| 807 |
+
"loss": 0.319,
|
| 808 |
+
"step": 1130
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.1009174311926606,
|
| 812 |
+
"grad_norm": 1.4113019794162502,
|
| 813 |
+
"learning_rate": 5e-06,
|
| 814 |
+
"loss": 0.3183,
|
| 815 |
+
"step": 1140
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.110574601641719,
|
| 819 |
+
"grad_norm": 1.4617901618486682,
|
| 820 |
+
"learning_rate": 5e-06,
|
| 821 |
+
"loss": 0.3206,
|
| 822 |
+
"step": 1150
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.1202317720907775,
|
| 826 |
+
"grad_norm": 1.4362393573916201,
|
| 827 |
+
"learning_rate": 5e-06,
|
| 828 |
+
"loss": 0.3161,
|
| 829 |
+
"step": 1160
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.1298889425398357,
|
| 833 |
+
"grad_norm": 1.480531545050256,
|
| 834 |
+
"learning_rate": 5e-06,
|
| 835 |
+
"loss": 0.3176,
|
| 836 |
+
"step": 1170
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.1395461129888942,
|
| 840 |
+
"grad_norm": 1.4429791008774708,
|
| 841 |
+
"learning_rate": 5e-06,
|
| 842 |
+
"loss": 0.3206,
|
| 843 |
+
"step": 1180
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.1492032834379526,
|
| 847 |
+
"grad_norm": 1.580292399548444,
|
| 848 |
+
"learning_rate": 5e-06,
|
| 849 |
+
"loss": 0.3189,
|
| 850 |
+
"step": 1190
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.158860453887011,
|
| 854 |
+
"grad_norm": 1.4710472342154686,
|
| 855 |
+
"learning_rate": 5e-06,
|
| 856 |
+
"loss": 0.3176,
|
| 857 |
+
"step": 1200
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.1685176243360695,
|
| 861 |
+
"grad_norm": 1.5510111053127804,
|
| 862 |
+
"learning_rate": 5e-06,
|
| 863 |
+
"loss": 0.3256,
|
| 864 |
+
"step": 1210
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.178174794785128,
|
| 868 |
+
"grad_norm": 1.479606545432114,
|
| 869 |
+
"learning_rate": 5e-06,
|
| 870 |
+
"loss": 0.3194,
|
| 871 |
+
"step": 1220
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.1878319652341864,
|
| 875 |
+
"grad_norm": 1.4470509715427249,
|
| 876 |
+
"learning_rate": 5e-06,
|
| 877 |
+
"loss": 0.315,
|
| 878 |
+
"step": 1230
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.1974891356832449,
|
| 882 |
+
"grad_norm": 1.688180637289201,
|
| 883 |
+
"learning_rate": 5e-06,
|
| 884 |
+
"loss": 0.3203,
|
| 885 |
+
"step": 1240
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.2071463061323033,
|
| 889 |
+
"grad_norm": 1.4261984744846448,
|
| 890 |
+
"learning_rate": 5e-06,
|
| 891 |
+
"loss": 0.3182,
|
| 892 |
+
"step": 1250
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.2168034765813616,
|
| 896 |
+
"grad_norm": 1.5604626368447907,
|
| 897 |
+
"learning_rate": 5e-06,
|
| 898 |
+
"loss": 0.3204,
|
| 899 |
+
"step": 1260
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.22646064703042,
|
| 903 |
+
"grad_norm": 1.5055974462561796,
|
| 904 |
+
"learning_rate": 5e-06,
|
| 905 |
+
"loss": 0.3194,
|
| 906 |
+
"step": 1270
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.2361178174794785,
|
| 910 |
+
"grad_norm": 1.472413847791833,
|
| 911 |
+
"learning_rate": 5e-06,
|
| 912 |
+
"loss": 0.3164,
|
| 913 |
+
"step": 1280
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.245774987928537,
|
| 917 |
+
"grad_norm": 1.4227053664977902,
|
| 918 |
+
"learning_rate": 5e-06,
|
| 919 |
+
"loss": 0.3195,
|
| 920 |
+
"step": 1290
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.2554321583775954,
|
| 924 |
+
"grad_norm": 1.4555119482883845,
|
| 925 |
+
"learning_rate": 5e-06,
|
| 926 |
+
"loss": 0.3206,
|
| 927 |
+
"step": 1300
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.2650893288266538,
|
| 931 |
+
"grad_norm": 1.3939863833537502,
|
| 932 |
+
"learning_rate": 5e-06,
|
| 933 |
+
"loss": 0.3174,
|
| 934 |
+
"step": 1310
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.2747464992757123,
|
| 938 |
+
"grad_norm": 1.6503658733951097,
|
| 939 |
+
"learning_rate": 5e-06,
|
| 940 |
+
"loss": 0.3212,
|
| 941 |
+
"step": 1320
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.2844036697247707,
|
| 945 |
+
"grad_norm": 1.5193288982909279,
|
| 946 |
+
"learning_rate": 5e-06,
|
| 947 |
+
"loss": 0.3185,
|
| 948 |
+
"step": 1330
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.2940608401738292,
|
| 952 |
+
"grad_norm": 1.8992907121697116,
|
| 953 |
+
"learning_rate": 5e-06,
|
| 954 |
+
"loss": 0.3195,
|
| 955 |
+
"step": 1340
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.3037180106228874,
|
| 959 |
+
"grad_norm": 1.462324131369204,
|
| 960 |
+
"learning_rate": 5e-06,
|
| 961 |
+
"loss": 0.3181,
|
| 962 |
+
"step": 1350
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.3133751810719458,
|
| 966 |
+
"grad_norm": 1.5578779143588841,
|
| 967 |
+
"learning_rate": 5e-06,
|
| 968 |
+
"loss": 0.3203,
|
| 969 |
+
"step": 1360
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.3230323515210043,
|
| 973 |
+
"grad_norm": 1.5253185582586557,
|
| 974 |
+
"learning_rate": 5e-06,
|
| 975 |
+
"loss": 0.3226,
|
| 976 |
+
"step": 1370
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.3326895219700627,
|
| 980 |
+
"grad_norm": 1.598448718585017,
|
| 981 |
+
"learning_rate": 5e-06,
|
| 982 |
+
"loss": 0.3237,
|
| 983 |
+
"step": 1380
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.3423466924191212,
|
| 987 |
+
"grad_norm": 1.4405857624240854,
|
| 988 |
+
"learning_rate": 5e-06,
|
| 989 |
+
"loss": 0.3223,
|
| 990 |
+
"step": 1390
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.3520038628681796,
|
| 994 |
+
"grad_norm": 1.517684638613212,
|
| 995 |
+
"learning_rate": 5e-06,
|
| 996 |
+
"loss": 0.3231,
|
| 997 |
+
"step": 1400
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.361661033317238,
|
| 1001 |
+
"grad_norm": 1.483051649435277,
|
| 1002 |
+
"learning_rate": 5e-06,
|
| 1003 |
+
"loss": 0.3239,
|
| 1004 |
+
"step": 1410
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.3713182037662965,
|
| 1008 |
+
"grad_norm": 1.5390732079855134,
|
| 1009 |
+
"learning_rate": 5e-06,
|
| 1010 |
+
"loss": 0.3218,
|
| 1011 |
+
"step": 1420
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.380975374215355,
|
| 1015 |
+
"grad_norm": 1.5122394104164938,
|
| 1016 |
+
"learning_rate": 5e-06,
|
| 1017 |
+
"loss": 0.322,
|
| 1018 |
+
"step": 1430
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.3906325446644132,
|
| 1022 |
+
"grad_norm": 1.5432820441426014,
|
| 1023 |
+
"learning_rate": 5e-06,
|
| 1024 |
+
"loss": 0.3145,
|
| 1025 |
+
"step": 1440
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.4002897151134719,
|
| 1029 |
+
"grad_norm": 1.4643744673012662,
|
| 1030 |
+
"learning_rate": 5e-06,
|
| 1031 |
+
"loss": 0.3178,
|
| 1032 |
+
"step": 1450
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.4099468855625301,
|
| 1036 |
+
"grad_norm": 1.4925264181574245,
|
| 1037 |
+
"learning_rate": 5e-06,
|
| 1038 |
+
"loss": 0.3243,
|
| 1039 |
+
"step": 1460
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.4196040560115886,
|
| 1043 |
+
"grad_norm": 1.4092376086387746,
|
| 1044 |
+
"learning_rate": 5e-06,
|
| 1045 |
+
"loss": 0.3211,
|
| 1046 |
+
"step": 1470
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.429261226460647,
|
| 1050 |
+
"grad_norm": 1.4232165437715756,
|
| 1051 |
+
"learning_rate": 5e-06,
|
| 1052 |
+
"loss": 0.3221,
|
| 1053 |
+
"step": 1480
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.4389183969097055,
|
| 1057 |
+
"grad_norm": 1.4049090252857714,
|
| 1058 |
+
"learning_rate": 5e-06,
|
| 1059 |
+
"loss": 0.3259,
|
| 1060 |
+
"step": 1490
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.448575567358764,
|
| 1064 |
+
"grad_norm": 1.5709064686147567,
|
| 1065 |
+
"learning_rate": 5e-06,
|
| 1066 |
+
"loss": 0.3192,
|
| 1067 |
+
"step": 1500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.4582327378078224,
|
| 1071 |
+
"grad_norm": 1.530143106655807,
|
| 1072 |
+
"learning_rate": 5e-06,
|
| 1073 |
+
"loss": 0.3198,
|
| 1074 |
+
"step": 1510
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.4678899082568808,
|
| 1078 |
+
"grad_norm": 1.4642072322356405,
|
| 1079 |
+
"learning_rate": 5e-06,
|
| 1080 |
+
"loss": 0.3242,
|
| 1081 |
+
"step": 1520
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.477547078705939,
|
| 1085 |
+
"grad_norm": 1.5078287858320247,
|
| 1086 |
+
"learning_rate": 5e-06,
|
| 1087 |
+
"loss": 0.3229,
|
| 1088 |
+
"step": 1530
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.4872042491549977,
|
| 1092 |
+
"grad_norm": 1.6543135999245175,
|
| 1093 |
+
"learning_rate": 5e-06,
|
| 1094 |
+
"loss": 0.3258,
|
| 1095 |
+
"step": 1540
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.496861419604056,
|
| 1099 |
+
"grad_norm": 1.4353311149318353,
|
| 1100 |
+
"learning_rate": 5e-06,
|
| 1101 |
+
"loss": 0.3218,
|
| 1102 |
+
"step": 1550
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.5065185900531144,
|
| 1106 |
+
"grad_norm": 1.427362675163451,
|
| 1107 |
+
"learning_rate": 5e-06,
|
| 1108 |
+
"loss": 0.3203,
|
| 1109 |
+
"step": 1560
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.5161757605021728,
|
| 1113 |
+
"grad_norm": 1.469685263972175,
|
| 1114 |
+
"learning_rate": 5e-06,
|
| 1115 |
+
"loss": 0.3233,
|
| 1116 |
+
"step": 1570
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.5258329309512313,
|
| 1120 |
+
"grad_norm": 1.3556540711826985,
|
| 1121 |
+
"learning_rate": 5e-06,
|
| 1122 |
+
"loss": 0.32,
|
| 1123 |
+
"step": 1580
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.5354901014002897,
|
| 1127 |
+
"grad_norm": 1.7785146026384562,
|
| 1128 |
+
"learning_rate": 5e-06,
|
| 1129 |
+
"loss": 0.3255,
|
| 1130 |
+
"step": 1590
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.5451472718493482,
|
| 1134 |
+
"grad_norm": 1.6406483748665126,
|
| 1135 |
+
"learning_rate": 5e-06,
|
| 1136 |
+
"loss": 0.325,
|
| 1137 |
+
"step": 1600
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.5548044422984066,
|
| 1141 |
+
"grad_norm": 1.487355254108548,
|
| 1142 |
+
"learning_rate": 5e-06,
|
| 1143 |
+
"loss": 0.3195,
|
| 1144 |
+
"step": 1610
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.5644616127474649,
|
| 1148 |
+
"grad_norm": 1.5580699551469968,
|
| 1149 |
+
"learning_rate": 5e-06,
|
| 1150 |
+
"loss": 0.324,
|
| 1151 |
+
"step": 1620
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.5741187831965235,
|
| 1155 |
+
"grad_norm": 1.6649624650081492,
|
| 1156 |
+
"learning_rate": 5e-06,
|
| 1157 |
+
"loss": 0.319,
|
| 1158 |
+
"step": 1630
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.5837759536455818,
|
| 1162 |
+
"grad_norm": 1.5071762619332458,
|
| 1163 |
+
"learning_rate": 5e-06,
|
| 1164 |
+
"loss": 0.319,
|
| 1165 |
+
"step": 1640
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.5934331240946402,
|
| 1169 |
+
"grad_norm": 1.4857917379863423,
|
| 1170 |
+
"learning_rate": 5e-06,
|
| 1171 |
+
"loss": 0.3246,
|
| 1172 |
+
"step": 1650
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.6030902945436987,
|
| 1176 |
+
"grad_norm": 1.539444045216992,
|
| 1177 |
+
"learning_rate": 5e-06,
|
| 1178 |
+
"loss": 0.325,
|
| 1179 |
+
"step": 1660
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.6127474649927571,
|
| 1183 |
+
"grad_norm": 1.4202629921913885,
|
| 1184 |
+
"learning_rate": 5e-06,
|
| 1185 |
+
"loss": 0.3249,
|
| 1186 |
+
"step": 1670
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.6224046354418156,
|
| 1190 |
+
"grad_norm": 1.3824989312845557,
|
| 1191 |
+
"learning_rate": 5e-06,
|
| 1192 |
+
"loss": 0.3174,
|
| 1193 |
+
"step": 1680
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.632061805890874,
|
| 1197 |
+
"grad_norm": 1.4180567800715427,
|
| 1198 |
+
"learning_rate": 5e-06,
|
| 1199 |
+
"loss": 0.32,
|
| 1200 |
+
"step": 1690
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.6417189763399325,
|
| 1204 |
+
"grad_norm": 1.5877926258862851,
|
| 1205 |
+
"learning_rate": 5e-06,
|
| 1206 |
+
"loss": 0.3251,
|
| 1207 |
+
"step": 1700
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.6513761467889907,
|
| 1211 |
+
"grad_norm": 1.4625682063733663,
|
| 1212 |
+
"learning_rate": 5e-06,
|
| 1213 |
+
"loss": 0.3251,
|
| 1214 |
+
"step": 1710
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.6610333172380494,
|
| 1218 |
+
"grad_norm": 1.4726154892277907,
|
| 1219 |
+
"learning_rate": 5e-06,
|
| 1220 |
+
"loss": 0.3217,
|
| 1221 |
+
"step": 1720
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.6706904876871076,
|
| 1225 |
+
"grad_norm": 1.4737546342638634,
|
| 1226 |
+
"learning_rate": 5e-06,
|
| 1227 |
+
"loss": 0.3241,
|
| 1228 |
+
"step": 1730
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.6803476581361663,
|
| 1232 |
+
"grad_norm": 1.5913560262828996,
|
| 1233 |
+
"learning_rate": 5e-06,
|
| 1234 |
+
"loss": 0.3216,
|
| 1235 |
+
"step": 1740
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.6900048285852245,
|
| 1239 |
+
"grad_norm": 1.6014273903495235,
|
| 1240 |
+
"learning_rate": 5e-06,
|
| 1241 |
+
"loss": 0.3228,
|
| 1242 |
+
"step": 1750
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.699661999034283,
|
| 1246 |
+
"grad_norm": 1.4444614753017166,
|
| 1247 |
+
"learning_rate": 5e-06,
|
| 1248 |
+
"loss": 0.3219,
|
| 1249 |
+
"step": 1760
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.7093191694833414,
|
| 1253 |
+
"grad_norm": 1.4711587303385112,
|
| 1254 |
+
"learning_rate": 5e-06,
|
| 1255 |
+
"loss": 0.3204,
|
| 1256 |
+
"step": 1770
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.7189763399323998,
|
| 1260 |
+
"grad_norm": 1.4911136718247058,
|
| 1261 |
+
"learning_rate": 5e-06,
|
| 1262 |
+
"loss": 0.3225,
|
| 1263 |
+
"step": 1780
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.7286335103814583,
|
| 1267 |
+
"grad_norm": 1.4836367449856465,
|
| 1268 |
+
"learning_rate": 5e-06,
|
| 1269 |
+
"loss": 0.3275,
|
| 1270 |
+
"step": 1790
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.7382906808305165,
|
| 1274 |
+
"grad_norm": 1.5104260942145018,
|
| 1275 |
+
"learning_rate": 5e-06,
|
| 1276 |
+
"loss": 0.3261,
|
| 1277 |
+
"step": 1800
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.7479478512795752,
|
| 1281 |
+
"grad_norm": 1.4587337741103181,
|
| 1282 |
+
"learning_rate": 5e-06,
|
| 1283 |
+
"loss": 0.3247,
|
| 1284 |
+
"step": 1810
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.7576050217286334,
|
| 1288 |
+
"grad_norm": 1.5339818170022481,
|
| 1289 |
+
"learning_rate": 5e-06,
|
| 1290 |
+
"loss": 0.3237,
|
| 1291 |
+
"step": 1820
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.767262192177692,
|
| 1295 |
+
"grad_norm": 1.4650422841731612,
|
| 1296 |
+
"learning_rate": 5e-06,
|
| 1297 |
+
"loss": 0.3218,
|
| 1298 |
+
"step": 1830
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.7769193626267503,
|
| 1302 |
+
"grad_norm": 1.6273768535871191,
|
| 1303 |
+
"learning_rate": 5e-06,
|
| 1304 |
+
"loss": 0.3239,
|
| 1305 |
+
"step": 1840
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.7865765330758088,
|
| 1309 |
+
"grad_norm": 1.6280662513787525,
|
| 1310 |
+
"learning_rate": 5e-06,
|
| 1311 |
+
"loss": 0.3267,
|
| 1312 |
+
"step": 1850
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.7962337035248672,
|
| 1316 |
+
"grad_norm": 1.469702386246834,
|
| 1317 |
+
"learning_rate": 5e-06,
|
| 1318 |
+
"loss": 0.322,
|
| 1319 |
+
"step": 1860
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.8058908739739257,
|
| 1323 |
+
"grad_norm": 1.6861190386571836,
|
| 1324 |
+
"learning_rate": 5e-06,
|
| 1325 |
+
"loss": 0.3246,
|
| 1326 |
+
"step": 1870
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.8155480444229841,
|
| 1330 |
+
"grad_norm": 1.481497242440989,
|
| 1331 |
+
"learning_rate": 5e-06,
|
| 1332 |
+
"loss": 0.3226,
|
| 1333 |
+
"step": 1880
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.8252052148720423,
|
| 1337 |
+
"grad_norm": 1.6155053021371237,
|
| 1338 |
+
"learning_rate": 5e-06,
|
| 1339 |
+
"loss": 0.3236,
|
| 1340 |
+
"step": 1890
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.834862385321101,
|
| 1344 |
+
"grad_norm": 1.416112013187697,
|
| 1345 |
+
"learning_rate": 5e-06,
|
| 1346 |
+
"loss": 0.3228,
|
| 1347 |
+
"step": 1900
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.8445195557701592,
|
| 1351 |
+
"grad_norm": 1.581676674672017,
|
| 1352 |
+
"learning_rate": 5e-06,
|
| 1353 |
+
"loss": 0.3241,
|
| 1354 |
+
"step": 1910
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.854176726219218,
|
| 1358 |
+
"grad_norm": 1.5569951920933318,
|
| 1359 |
+
"learning_rate": 5e-06,
|
| 1360 |
+
"loss": 0.3259,
|
| 1361 |
+
"step": 1920
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.8638338966682761,
|
| 1365 |
+
"grad_norm": 1.6829953573077656,
|
| 1366 |
+
"learning_rate": 5e-06,
|
| 1367 |
+
"loss": 0.3239,
|
| 1368 |
+
"step": 1930
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.8734910671173346,
|
| 1372 |
+
"grad_norm": 1.4499124168214599,
|
| 1373 |
+
"learning_rate": 5e-06,
|
| 1374 |
+
"loss": 0.3237,
|
| 1375 |
+
"step": 1940
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.883148237566393,
|
| 1379 |
+
"grad_norm": 1.3921897240084418,
|
| 1380 |
+
"learning_rate": 5e-06,
|
| 1381 |
+
"loss": 0.3202,
|
| 1382 |
+
"step": 1950
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.8928054080154515,
|
| 1386 |
+
"grad_norm": 1.7399387456220219,
|
| 1387 |
+
"learning_rate": 5e-06,
|
| 1388 |
+
"loss": 0.3185,
|
| 1389 |
+
"step": 1960
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.90246257846451,
|
| 1393 |
+
"grad_norm": 1.5558707224477395,
|
| 1394 |
+
"learning_rate": 5e-06,
|
| 1395 |
+
"loss": 0.323,
|
| 1396 |
+
"step": 1970
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.9121197489135682,
|
| 1400 |
+
"grad_norm": 1.49802902853701,
|
| 1401 |
+
"learning_rate": 5e-06,
|
| 1402 |
+
"loss": 0.3234,
|
| 1403 |
+
"step": 1980
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.9217769193626268,
|
| 1407 |
+
"grad_norm": 1.4807948737028658,
|
| 1408 |
+
"learning_rate": 5e-06,
|
| 1409 |
+
"loss": 0.322,
|
| 1410 |
+
"step": 1990
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.931434089811685,
|
| 1414 |
+
"grad_norm": 1.3868348286955445,
|
| 1415 |
+
"learning_rate": 5e-06,
|
| 1416 |
+
"loss": 0.3247,
|
| 1417 |
+
"step": 2000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.9410912602607437,
|
| 1421 |
+
"grad_norm": 1.6163693821676344,
|
| 1422 |
+
"learning_rate": 5e-06,
|
| 1423 |
+
"loss": 0.3248,
|
| 1424 |
+
"step": 2010
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.950748430709802,
|
| 1428 |
+
"grad_norm": 1.552439354775519,
|
| 1429 |
+
"learning_rate": 5e-06,
|
| 1430 |
+
"loss": 0.3233,
|
| 1431 |
+
"step": 2020
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.9604056011588604,
|
| 1435 |
+
"grad_norm": 1.5224507289579325,
|
| 1436 |
+
"learning_rate": 5e-06,
|
| 1437 |
+
"loss": 0.3311,
|
| 1438 |
+
"step": 2030
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.9700627716079189,
|
| 1442 |
+
"grad_norm": 1.3715156278923533,
|
| 1443 |
+
"learning_rate": 5e-06,
|
| 1444 |
+
"loss": 0.3224,
|
| 1445 |
+
"step": 2040
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.9797199420569773,
|
| 1449 |
+
"grad_norm": 1.4690837804253396,
|
| 1450 |
+
"learning_rate": 5e-06,
|
| 1451 |
+
"loss": 0.3245,
|
| 1452 |
+
"step": 2050
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.9893771125060358,
|
| 1456 |
+
"grad_norm": 1.5304270102797093,
|
| 1457 |
+
"learning_rate": 5e-06,
|
| 1458 |
+
"loss": 0.3283,
|
| 1459 |
+
"step": 2060
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.999034282955094,
|
| 1463 |
+
"grad_norm": 1.5064039907716613,
|
| 1464 |
+
"learning_rate": 5e-06,
|
| 1465 |
+
"loss": 0.326,
|
| 1466 |
+
"step": 2070
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.0,
|
| 1470 |
+
"eval_loss": 0.41363242268562317,
|
| 1471 |
+
"eval_runtime": 183.9864,
|
| 1472 |
+
"eval_samples_per_second": 151.658,
|
| 1473 |
+
"eval_steps_per_second": 0.592,
|
| 1474 |
+
"step": 2071
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 2.0086914534041527,
|
| 1478 |
+
"grad_norm": 1.9529357489293813,
|
| 1479 |
+
"learning_rate": 5e-06,
|
| 1480 |
+
"loss": 0.2222,
|
| 1481 |
+
"step": 2080
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 2.018348623853211,
|
| 1485 |
+
"grad_norm": 1.755518669301023,
|
| 1486 |
+
"learning_rate": 5e-06,
|
| 1487 |
+
"loss": 0.2019,
|
| 1488 |
+
"step": 2090
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 2.0280057943022696,
|
| 1492 |
+
"grad_norm": 1.740635003243075,
|
| 1493 |
+
"learning_rate": 5e-06,
|
| 1494 |
+
"loss": 0.203,
|
| 1495 |
+
"step": 2100
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 2.037662964751328,
|
| 1499 |
+
"grad_norm": 1.8237819578036996,
|
| 1500 |
+
"learning_rate": 5e-06,
|
| 1501 |
+
"loss": 0.1982,
|
| 1502 |
+
"step": 2110
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 2.0473201352003865,
|
| 1506 |
+
"grad_norm": 1.5602633442114324,
|
| 1507 |
+
"learning_rate": 5e-06,
|
| 1508 |
+
"loss": 0.1992,
|
| 1509 |
+
"step": 2120
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 2.0569773056494447,
|
| 1513 |
+
"grad_norm": 1.816471476491465,
|
| 1514 |
+
"learning_rate": 5e-06,
|
| 1515 |
+
"loss": 0.2049,
|
| 1516 |
+
"step": 2130
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 2.066634476098503,
|
| 1520 |
+
"grad_norm": 1.7673442658203458,
|
| 1521 |
+
"learning_rate": 5e-06,
|
| 1522 |
+
"loss": 0.1993,
|
| 1523 |
+
"step": 2140
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 2.0762916465475616,
|
| 1527 |
+
"grad_norm": 1.6973737687900528,
|
| 1528 |
+
"learning_rate": 5e-06,
|
| 1529 |
+
"loss": 0.1998,
|
| 1530 |
+
"step": 2150
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 2.08594881699662,
|
| 1534 |
+
"grad_norm": 1.5869449008271677,
|
| 1535 |
+
"learning_rate": 5e-06,
|
| 1536 |
+
"loss": 0.202,
|
| 1537 |
+
"step": 2160
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 2.0956059874456785,
|
| 1541 |
+
"grad_norm": 1.64683922032891,
|
| 1542 |
+
"learning_rate": 5e-06,
|
| 1543 |
+
"loss": 0.2002,
|
| 1544 |
+
"step": 2170
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 2.1052631578947367,
|
| 1548 |
+
"grad_norm": 1.9386924431067427,
|
| 1549 |
+
"learning_rate": 5e-06,
|
| 1550 |
+
"loss": 0.2076,
|
| 1551 |
+
"step": 2180
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 2.1149203283437954,
|
| 1555 |
+
"grad_norm": 1.6210647566768155,
|
| 1556 |
+
"learning_rate": 5e-06,
|
| 1557 |
+
"loss": 0.2028,
|
| 1558 |
+
"step": 2190
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 2.1245774987928536,
|
| 1562 |
+
"grad_norm": 1.6287574889203644,
|
| 1563 |
+
"learning_rate": 5e-06,
|
| 1564 |
+
"loss": 0.2042,
|
| 1565 |
+
"step": 2200
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 2.1342346692419123,
|
| 1569 |
+
"grad_norm": 1.655477861818872,
|
| 1570 |
+
"learning_rate": 5e-06,
|
| 1571 |
+
"loss": 0.2013,
|
| 1572 |
+
"step": 2210
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 2.1438918396909705,
|
| 1576 |
+
"grad_norm": 1.5729371185522592,
|
| 1577 |
+
"learning_rate": 5e-06,
|
| 1578 |
+
"loss": 0.2015,
|
| 1579 |
+
"step": 2220
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 2.153549010140029,
|
| 1583 |
+
"grad_norm": 1.6476425123948135,
|
| 1584 |
+
"learning_rate": 5e-06,
|
| 1585 |
+
"loss": 0.2012,
|
| 1586 |
+
"step": 2230
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 2.1632061805890874,
|
| 1590 |
+
"grad_norm": 1.8001638906124682,
|
| 1591 |
+
"learning_rate": 5e-06,
|
| 1592 |
+
"loss": 0.2008,
|
| 1593 |
+
"step": 2240
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 2.1728633510381457,
|
| 1597 |
+
"grad_norm": 1.7044240854710664,
|
| 1598 |
+
"learning_rate": 5e-06,
|
| 1599 |
+
"loss": 0.2054,
|
| 1600 |
+
"step": 2250
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 2.1825205214872043,
|
| 1604 |
+
"grad_norm": 1.7542655066679125,
|
| 1605 |
+
"learning_rate": 5e-06,
|
| 1606 |
+
"loss": 0.2034,
|
| 1607 |
+
"step": 2260
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 2.1921776919362626,
|
| 1611 |
+
"grad_norm": 1.6322702947383616,
|
| 1612 |
+
"learning_rate": 5e-06,
|
| 1613 |
+
"loss": 0.2057,
|
| 1614 |
+
"step": 2270
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 2.2018348623853212,
|
| 1618 |
+
"grad_norm": 1.6677542060673687,
|
| 1619 |
+
"learning_rate": 5e-06,
|
| 1620 |
+
"loss": 0.207,
|
| 1621 |
+
"step": 2280
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 2.2114920328343795,
|
| 1625 |
+
"grad_norm": 1.7199005988447569,
|
| 1626 |
+
"learning_rate": 5e-06,
|
| 1627 |
+
"loss": 0.2078,
|
| 1628 |
+
"step": 2290
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 2.221149203283438,
|
| 1632 |
+
"grad_norm": 1.7417097237727022,
|
| 1633 |
+
"learning_rate": 5e-06,
|
| 1634 |
+
"loss": 0.2056,
|
| 1635 |
+
"step": 2300
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 2.2308063737324964,
|
| 1639 |
+
"grad_norm": 1.7881569830854456,
|
| 1640 |
+
"learning_rate": 5e-06,
|
| 1641 |
+
"loss": 0.2018,
|
| 1642 |
+
"step": 2310
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 2.240463544181555,
|
| 1646 |
+
"grad_norm": 1.778258239931456,
|
| 1647 |
+
"learning_rate": 5e-06,
|
| 1648 |
+
"loss": 0.2085,
|
| 1649 |
+
"step": 2320
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 2.2501207146306133,
|
| 1653 |
+
"grad_norm": 1.7066617550978782,
|
| 1654 |
+
"learning_rate": 5e-06,
|
| 1655 |
+
"loss": 0.2081,
|
| 1656 |
+
"step": 2330
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 2.2597778850796715,
|
| 1660 |
+
"grad_norm": 1.6359235244009738,
|
| 1661 |
+
"learning_rate": 5e-06,
|
| 1662 |
+
"loss": 0.2068,
|
| 1663 |
+
"step": 2340
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 2.26943505552873,
|
| 1667 |
+
"grad_norm": 1.9690603812490861,
|
| 1668 |
+
"learning_rate": 5e-06,
|
| 1669 |
+
"loss": 0.2093,
|
| 1670 |
+
"step": 2350
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 2.2790922259777884,
|
| 1674 |
+
"grad_norm": 1.8155935869005713,
|
| 1675 |
+
"learning_rate": 5e-06,
|
| 1676 |
+
"loss": 0.2086,
|
| 1677 |
+
"step": 2360
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 2.288749396426847,
|
| 1681 |
+
"grad_norm": 1.7505898538326212,
|
| 1682 |
+
"learning_rate": 5e-06,
|
| 1683 |
+
"loss": 0.2082,
|
| 1684 |
+
"step": 2370
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 2.2984065668759053,
|
| 1688 |
+
"grad_norm": 1.687851682529015,
|
| 1689 |
+
"learning_rate": 5e-06,
|
| 1690 |
+
"loss": 0.2094,
|
| 1691 |
+
"step": 2380
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 2.308063737324964,
|
| 1695 |
+
"grad_norm": 2.0069895205569535,
|
| 1696 |
+
"learning_rate": 5e-06,
|
| 1697 |
+
"loss": 0.2114,
|
| 1698 |
+
"step": 2390
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 2.317720907774022,
|
| 1702 |
+
"grad_norm": 1.8064841179245696,
|
| 1703 |
+
"learning_rate": 5e-06,
|
| 1704 |
+
"loss": 0.2059,
|
| 1705 |
+
"step": 2400
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 2.327378078223081,
|
| 1709 |
+
"grad_norm": 1.7513896709251378,
|
| 1710 |
+
"learning_rate": 5e-06,
|
| 1711 |
+
"loss": 0.2073,
|
| 1712 |
+
"step": 2410
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 2.337035248672139,
|
| 1716 |
+
"grad_norm": 1.7915692908106093,
|
| 1717 |
+
"learning_rate": 5e-06,
|
| 1718 |
+
"loss": 0.2097,
|
| 1719 |
+
"step": 2420
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 2.3466924191211973,
|
| 1723 |
+
"grad_norm": 1.6328915879521833,
|
| 1724 |
+
"learning_rate": 5e-06,
|
| 1725 |
+
"loss": 0.2126,
|
| 1726 |
+
"step": 2430
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 2.356349589570256,
|
| 1730 |
+
"grad_norm": 1.6676621870989963,
|
| 1731 |
+
"learning_rate": 5e-06,
|
| 1732 |
+
"loss": 0.2082,
|
| 1733 |
+
"step": 2440
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 2.366006760019314,
|
| 1737 |
+
"grad_norm": 1.71621065929775,
|
| 1738 |
+
"learning_rate": 5e-06,
|
| 1739 |
+
"loss": 0.2079,
|
| 1740 |
+
"step": 2450
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 2.375663930468373,
|
| 1744 |
+
"grad_norm": 1.6107515870333837,
|
| 1745 |
+
"learning_rate": 5e-06,
|
| 1746 |
+
"loss": 0.213,
|
| 1747 |
+
"step": 2460
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 2.385321100917431,
|
| 1751 |
+
"grad_norm": 1.723592646527568,
|
| 1752 |
+
"learning_rate": 5e-06,
|
| 1753 |
+
"loss": 0.2112,
|
| 1754 |
+
"step": 2470
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 2.3949782713664898,
|
| 1758 |
+
"grad_norm": 1.810751615032428,
|
| 1759 |
+
"learning_rate": 5e-06,
|
| 1760 |
+
"loss": 0.214,
|
| 1761 |
+
"step": 2480
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 2.404635441815548,
|
| 1765 |
+
"grad_norm": 1.8583457997365094,
|
| 1766 |
+
"learning_rate": 5e-06,
|
| 1767 |
+
"loss": 0.2133,
|
| 1768 |
+
"step": 2490
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 2.4142926122646067,
|
| 1772 |
+
"grad_norm": 1.8098882466647033,
|
| 1773 |
+
"learning_rate": 5e-06,
|
| 1774 |
+
"loss": 0.2118,
|
| 1775 |
+
"step": 2500
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 2.423949782713665,
|
| 1779 |
+
"grad_norm": 1.7310394353267804,
|
| 1780 |
+
"learning_rate": 5e-06,
|
| 1781 |
+
"loss": 0.2091,
|
| 1782 |
+
"step": 2510
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 2.433606953162723,
|
| 1786 |
+
"grad_norm": 1.9482929822769175,
|
| 1787 |
+
"learning_rate": 5e-06,
|
| 1788 |
+
"loss": 0.2103,
|
| 1789 |
+
"step": 2520
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 2.443264123611782,
|
| 1793 |
+
"grad_norm": 1.8075435579847086,
|
| 1794 |
+
"learning_rate": 5e-06,
|
| 1795 |
+
"loss": 0.2111,
|
| 1796 |
+
"step": 2530
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 2.45292129406084,
|
| 1800 |
+
"grad_norm": 1.6649997057492705,
|
| 1801 |
+
"learning_rate": 5e-06,
|
| 1802 |
+
"loss": 0.2119,
|
| 1803 |
+
"step": 2540
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 2.4625784645098987,
|
| 1807 |
+
"grad_norm": 1.595570416233996,
|
| 1808 |
+
"learning_rate": 5e-06,
|
| 1809 |
+
"loss": 0.209,
|
| 1810 |
+
"step": 2550
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 2.472235634958957,
|
| 1814 |
+
"grad_norm": 1.707621313978413,
|
| 1815 |
+
"learning_rate": 5e-06,
|
| 1816 |
+
"loss": 0.2124,
|
| 1817 |
+
"step": 2560
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 2.4818928054080156,
|
| 1821 |
+
"grad_norm": 1.7540598817249424,
|
| 1822 |
+
"learning_rate": 5e-06,
|
| 1823 |
+
"loss": 0.2134,
|
| 1824 |
+
"step": 2570
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 2.491549975857074,
|
| 1828 |
+
"grad_norm": 1.7644906677930094,
|
| 1829 |
+
"learning_rate": 5e-06,
|
| 1830 |
+
"loss": 0.212,
|
| 1831 |
+
"step": 2580
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 2.5012071463061325,
|
| 1835 |
+
"grad_norm": 1.7556689404760746,
|
| 1836 |
+
"learning_rate": 5e-06,
|
| 1837 |
+
"loss": 0.2141,
|
| 1838 |
+
"step": 2590
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 2.5108643167551907,
|
| 1842 |
+
"grad_norm": 1.7473267171283928,
|
| 1843 |
+
"learning_rate": 5e-06,
|
| 1844 |
+
"loss": 0.2124,
|
| 1845 |
+
"step": 2600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 2.520521487204249,
|
| 1849 |
+
"grad_norm": 1.9136137702521907,
|
| 1850 |
+
"learning_rate": 5e-06,
|
| 1851 |
+
"loss": 0.2127,
|
| 1852 |
+
"step": 2610
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 2.5301786576533076,
|
| 1856 |
+
"grad_norm": 1.7481532118907182,
|
| 1857 |
+
"learning_rate": 5e-06,
|
| 1858 |
+
"loss": 0.2131,
|
| 1859 |
+
"step": 2620
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 2.539835828102366,
|
| 1863 |
+
"grad_norm": 1.832460065168467,
|
| 1864 |
+
"learning_rate": 5e-06,
|
| 1865 |
+
"loss": 0.2153,
|
| 1866 |
+
"step": 2630
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 2.5494929985514245,
|
| 1870 |
+
"grad_norm": 1.6966236245813655,
|
| 1871 |
+
"learning_rate": 5e-06,
|
| 1872 |
+
"loss": 0.2131,
|
| 1873 |
+
"step": 2640
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 2.5591501690004828,
|
| 1877 |
+
"grad_norm": 1.7181221486120444,
|
| 1878 |
+
"learning_rate": 5e-06,
|
| 1879 |
+
"loss": 0.213,
|
| 1880 |
+
"step": 2650
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 2.5688073394495414,
|
| 1884 |
+
"grad_norm": 1.6118010065233517,
|
| 1885 |
+
"learning_rate": 5e-06,
|
| 1886 |
+
"loss": 0.2148,
|
| 1887 |
+
"step": 2660
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 2.5784645098985997,
|
| 1891 |
+
"grad_norm": 1.7015205827064956,
|
| 1892 |
+
"learning_rate": 5e-06,
|
| 1893 |
+
"loss": 0.219,
|
| 1894 |
+
"step": 2670
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 2.5881216803476583,
|
| 1898 |
+
"grad_norm": 1.688200340928073,
|
| 1899 |
+
"learning_rate": 5e-06,
|
| 1900 |
+
"loss": 0.2145,
|
| 1901 |
+
"step": 2680
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 2.5977788507967166,
|
| 1905 |
+
"grad_norm": 1.628479976082271,
|
| 1906 |
+
"learning_rate": 5e-06,
|
| 1907 |
+
"loss": 0.2173,
|
| 1908 |
+
"step": 2690
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 2.607436021245775,
|
| 1912 |
+
"grad_norm": 1.6284654190992232,
|
| 1913 |
+
"learning_rate": 5e-06,
|
| 1914 |
+
"loss": 0.2166,
|
| 1915 |
+
"step": 2700
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 2.6170931916948335,
|
| 1919 |
+
"grad_norm": 1.7027701458933544,
|
| 1920 |
+
"learning_rate": 5e-06,
|
| 1921 |
+
"loss": 0.2154,
|
| 1922 |
+
"step": 2710
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 2.6267503621438917,
|
| 1926 |
+
"grad_norm": 1.6155718880952934,
|
| 1927 |
+
"learning_rate": 5e-06,
|
| 1928 |
+
"loss": 0.2154,
|
| 1929 |
+
"step": 2720
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 2.6364075325929504,
|
| 1933 |
+
"grad_norm": 1.6351349351929865,
|
| 1934 |
+
"learning_rate": 5e-06,
|
| 1935 |
+
"loss": 0.2161,
|
| 1936 |
+
"step": 2730
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 2.6460647030420086,
|
| 1940 |
+
"grad_norm": 1.7361275613850788,
|
| 1941 |
+
"learning_rate": 5e-06,
|
| 1942 |
+
"loss": 0.2183,
|
| 1943 |
+
"step": 2740
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 2.6557218734910673,
|
| 1947 |
+
"grad_norm": 1.9309759687598864,
|
| 1948 |
+
"learning_rate": 5e-06,
|
| 1949 |
+
"loss": 0.2183,
|
| 1950 |
+
"step": 2750
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 2.6653790439401255,
|
| 1954 |
+
"grad_norm": 1.7778715949832347,
|
| 1955 |
+
"learning_rate": 5e-06,
|
| 1956 |
+
"loss": 0.2164,
|
| 1957 |
+
"step": 2760
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 2.675036214389184,
|
| 1961 |
+
"grad_norm": 1.689951172416981,
|
| 1962 |
+
"learning_rate": 5e-06,
|
| 1963 |
+
"loss": 0.2181,
|
| 1964 |
+
"step": 2770
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.6846933848382424,
|
| 1968 |
+
"grad_norm": 1.9355370897589272,
|
| 1969 |
+
"learning_rate": 5e-06,
|
| 1970 |
+
"loss": 0.2205,
|
| 1971 |
+
"step": 2780
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 2.6943505552873006,
|
| 1975 |
+
"grad_norm": 1.8460056442038406,
|
| 1976 |
+
"learning_rate": 5e-06,
|
| 1977 |
+
"loss": 0.2153,
|
| 1978 |
+
"step": 2790
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.7040077257363593,
|
| 1982 |
+
"grad_norm": 1.857187825531335,
|
| 1983 |
+
"learning_rate": 5e-06,
|
| 1984 |
+
"loss": 0.2194,
|
| 1985 |
+
"step": 2800
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 2.7136648961854175,
|
| 1989 |
+
"grad_norm": 1.7784483414745131,
|
| 1990 |
+
"learning_rate": 5e-06,
|
| 1991 |
+
"loss": 0.2155,
|
| 1992 |
+
"step": 2810
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 2.723322066634476,
|
| 1996 |
+
"grad_norm": 1.6570596632282855,
|
| 1997 |
+
"learning_rate": 5e-06,
|
| 1998 |
+
"loss": 0.2204,
|
| 1999 |
+
"step": 2820
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 2.7329792370835344,
|
| 2003 |
+
"grad_norm": 1.8006740077758066,
|
| 2004 |
+
"learning_rate": 5e-06,
|
| 2005 |
+
"loss": 0.2185,
|
| 2006 |
+
"step": 2830
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.742636407532593,
|
| 2010 |
+
"grad_norm": 1.6910488555694076,
|
| 2011 |
+
"learning_rate": 5e-06,
|
| 2012 |
+
"loss": 0.2215,
|
| 2013 |
+
"step": 2840
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 2.7522935779816513,
|
| 2017 |
+
"grad_norm": 1.6824061864717337,
|
| 2018 |
+
"learning_rate": 5e-06,
|
| 2019 |
+
"loss": 0.2168,
|
| 2020 |
+
"step": 2850
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 2.76195074843071,
|
| 2024 |
+
"grad_norm": 1.667248411860486,
|
| 2025 |
+
"learning_rate": 5e-06,
|
| 2026 |
+
"loss": 0.2171,
|
| 2027 |
+
"step": 2860
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 2.771607918879768,
|
| 2031 |
+
"grad_norm": 1.6621742506671506,
|
| 2032 |
+
"learning_rate": 5e-06,
|
| 2033 |
+
"loss": 0.2172,
|
| 2034 |
+
"step": 2870
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 2.7812650893288264,
|
| 2038 |
+
"grad_norm": 1.951777361512585,
|
| 2039 |
+
"learning_rate": 5e-06,
|
| 2040 |
+
"loss": 0.2169,
|
| 2041 |
+
"step": 2880
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 2.790922259777885,
|
| 2045 |
+
"grad_norm": 1.6376961183273515,
|
| 2046 |
+
"learning_rate": 5e-06,
|
| 2047 |
+
"loss": 0.2184,
|
| 2048 |
+
"step": 2890
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 2.8005794302269438,
|
| 2052 |
+
"grad_norm": 1.7499302221356705,
|
| 2053 |
+
"learning_rate": 5e-06,
|
| 2054 |
+
"loss": 0.2172,
|
| 2055 |
+
"step": 2900
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 2.810236600676002,
|
| 2059 |
+
"grad_norm": 1.6961296830905148,
|
| 2060 |
+
"learning_rate": 5e-06,
|
| 2061 |
+
"loss": 0.2191,
|
| 2062 |
+
"step": 2910
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 2.8198937711250602,
|
| 2066 |
+
"grad_norm": 1.7673008414488,
|
| 2067 |
+
"learning_rate": 5e-06,
|
| 2068 |
+
"loss": 0.2203,
|
| 2069 |
+
"step": 2920
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 2.829550941574119,
|
| 2073 |
+
"grad_norm": 1.756708895885833,
|
| 2074 |
+
"learning_rate": 5e-06,
|
| 2075 |
+
"loss": 0.2192,
|
| 2076 |
+
"step": 2930
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 2.839208112023177,
|
| 2080 |
+
"grad_norm": 1.7064534219401073,
|
| 2081 |
+
"learning_rate": 5e-06,
|
| 2082 |
+
"loss": 0.2194,
|
| 2083 |
+
"step": 2940
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 2.848865282472236,
|
| 2087 |
+
"grad_norm": 1.854536004459549,
|
| 2088 |
+
"learning_rate": 5e-06,
|
| 2089 |
+
"loss": 0.2207,
|
| 2090 |
+
"step": 2950
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.858522452921294,
|
| 2094 |
+
"grad_norm": 1.7590173226380559,
|
| 2095 |
+
"learning_rate": 5e-06,
|
| 2096 |
+
"loss": 0.2234,
|
| 2097 |
+
"step": 2960
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 2.8681796233703523,
|
| 2101 |
+
"grad_norm": 1.6888816007203014,
|
| 2102 |
+
"learning_rate": 5e-06,
|
| 2103 |
+
"loss": 0.2206,
|
| 2104 |
+
"step": 2970
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 2.877836793819411,
|
| 2108 |
+
"grad_norm": 1.8848356855139519,
|
| 2109 |
+
"learning_rate": 5e-06,
|
| 2110 |
+
"loss": 0.2199,
|
| 2111 |
+
"step": 2980
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.8874939642684696,
|
| 2115 |
+
"grad_norm": 1.7081747642661718,
|
| 2116 |
+
"learning_rate": 5e-06,
|
| 2117 |
+
"loss": 0.2181,
|
| 2118 |
+
"step": 2990
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 2.897151134717528,
|
| 2122 |
+
"grad_norm": 1.8139898022279473,
|
| 2123 |
+
"learning_rate": 5e-06,
|
| 2124 |
+
"loss": 0.221,
|
| 2125 |
+
"step": 3000
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 2.906808305166586,
|
| 2129 |
+
"grad_norm": 1.8123670196961439,
|
| 2130 |
+
"learning_rate": 5e-06,
|
| 2131 |
+
"loss": 0.2176,
|
| 2132 |
+
"step": 3010
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 2.9164654756156447,
|
| 2136 |
+
"grad_norm": 1.6673754695154868,
|
| 2137 |
+
"learning_rate": 5e-06,
|
| 2138 |
+
"loss": 0.2232,
|
| 2139 |
+
"step": 3020
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 2.926122646064703,
|
| 2143 |
+
"grad_norm": 1.6571101014785201,
|
| 2144 |
+
"learning_rate": 5e-06,
|
| 2145 |
+
"loss": 0.2203,
|
| 2146 |
+
"step": 3030
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 2.9357798165137616,
|
| 2150 |
+
"grad_norm": 1.7173214731808049,
|
| 2151 |
+
"learning_rate": 5e-06,
|
| 2152 |
+
"loss": 0.2215,
|
| 2153 |
+
"step": 3040
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 2.94543698696282,
|
| 2157 |
+
"grad_norm": 1.74112205509197,
|
| 2158 |
+
"learning_rate": 5e-06,
|
| 2159 |
+
"loss": 0.2193,
|
| 2160 |
+
"step": 3050
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 2.955094157411878,
|
| 2164 |
+
"grad_norm": 1.6720045527666803,
|
| 2165 |
+
"learning_rate": 5e-06,
|
| 2166 |
+
"loss": 0.2215,
|
| 2167 |
+
"step": 3060
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 2.9647513278609368,
|
| 2171 |
+
"grad_norm": 1.699005956548853,
|
| 2172 |
+
"learning_rate": 5e-06,
|
| 2173 |
+
"loss": 0.2208,
|
| 2174 |
+
"step": 3070
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 2.9744084983099954,
|
| 2178 |
+
"grad_norm": 1.65145518846451,
|
| 2179 |
+
"learning_rate": 5e-06,
|
| 2180 |
+
"loss": 0.2229,
|
| 2181 |
+
"step": 3080
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 2.9840656687590537,
|
| 2185 |
+
"grad_norm": 1.7864091472536987,
|
| 2186 |
+
"learning_rate": 5e-06,
|
| 2187 |
+
"loss": 0.2221,
|
| 2188 |
+
"step": 3090
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 2.993722839208112,
|
| 2192 |
+
"grad_norm": 1.7192522447102805,
|
| 2193 |
+
"learning_rate": 5e-06,
|
| 2194 |
+
"loss": 0.2213,
|
| 2195 |
+
"step": 3100
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 2.998551424432641,
|
| 2199 |
+
"eval_loss": 0.4467960000038147,
|
| 2200 |
+
"eval_runtime": 177.447,
|
| 2201 |
+
"eval_samples_per_second": 157.247,
|
| 2202 |
+
"eval_steps_per_second": 0.614,
|
| 2203 |
+
"step": 3105
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 2.998551424432641,
|
| 2207 |
+
"step": 3105,
|
| 2208 |
+
"total_flos": 5200153128468480.0,
|
| 2209 |
+
"train_loss": 0.3273588392085691,
|
| 2210 |
+
"train_runtime": 26340.0491,
|
| 2211 |
+
"train_samples_per_second": 60.381,
|
| 2212 |
+
"train_steps_per_second": 0.118
|
| 2213 |
+
}
|
| 2214 |
+
],
|
| 2215 |
+
"logging_steps": 10,
|
| 2216 |
+
"max_steps": 3105,
|
| 2217 |
+
"num_input_tokens_seen": 0,
|
| 2218 |
+
"num_train_epochs": 3,
|
| 2219 |
+
"save_steps": 500,
|
| 2220 |
+
"stateful_callbacks": {
|
| 2221 |
+
"TrainerControl": {
|
| 2222 |
+
"args": {
|
| 2223 |
+
"should_epoch_stop": false,
|
| 2224 |
+
"should_evaluate": false,
|
| 2225 |
+
"should_log": false,
|
| 2226 |
+
"should_save": true,
|
| 2227 |
+
"should_training_stop": true
|
| 2228 |
+
},
|
| 2229 |
+
"attributes": {}
|
| 2230 |
+
}
|
| 2231 |
+
},
|
| 2232 |
+
"total_flos": 5200153128468480.0,
|
| 2233 |
+
"train_batch_size": 8,
|
| 2234 |
+
"trial_name": null,
|
| 2235 |
+
"trial_params": null
|
| 2236 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|