End of training
Browse files- README.md +3 -2
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2442 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3.2
|
|
| 4 |
base_model: meta-llama/Llama-3.2-1B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: mcmaster-llama3-1b-full-pt
|
|
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# mcmaster-llama3-1b-full-pt
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 4 |
base_model: meta-llama/Llama-3.2-1B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: mcmaster-llama3-1b-full-pt
|
|
|
|
| 16 |
|
| 17 |
# mcmaster-llama3-1b-full-pt
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on the mcmaster dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.0319
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.998698642870422,
|
| 3 |
+
"eval_loss": 0.03189178928732872,
|
| 4 |
+
"eval_runtime": 212.4751,
|
| 5 |
+
"eval_samples_per_second": 22.501,
|
| 6 |
+
"eval_steps_per_second": 22.501,
|
| 7 |
+
"perplexity": 1.0324057818949677,
|
| 8 |
+
"total_flos": 5.14290499398402e+18,
|
| 9 |
+
"train_loss": 0.19145491501161208,
|
| 10 |
+
"train_runtime": 31931.328,
|
| 11 |
+
"train_samples_per_second": 6.737,
|
| 12 |
+
"train_steps_per_second": 0.105
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.998698642870422,
|
| 3 |
+
"eval_loss": 0.03189178928732872,
|
| 4 |
+
"eval_runtime": 212.4751,
|
| 5 |
+
"eval_samples_per_second": 22.501,
|
| 6 |
+
"eval_steps_per_second": 22.501,
|
| 7 |
+
"perplexity": 1.0324057818949677
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.998698642870422,
|
| 3 |
+
"total_flos": 5.14290499398402e+18,
|
| 4 |
+
"train_loss": 0.19145491501161208,
|
| 5 |
+
"train_runtime": 31931.328,
|
| 6 |
+
"train_samples_per_second": 6.737,
|
| 7 |
+
"train_steps_per_second": 0.105
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2442 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.998698642870422,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3360,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.014872652909462726,
|
| 13 |
+
"grad_norm": 1.0425002574920654,
|
| 14 |
+
"learning_rate": 8.928571428571428e-06,
|
| 15 |
+
"loss": 0.4346,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02974530581892545,
|
| 20 |
+
"grad_norm": 0.5861272811889648,
|
| 21 |
+
"learning_rate": 1.7857142857142855e-05,
|
| 22 |
+
"loss": 0.3527,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.044617958728388175,
|
| 27 |
+
"grad_norm": 0.5629558563232422,
|
| 28 |
+
"learning_rate": 2.6785714285714284e-05,
|
| 29 |
+
"loss": 0.2922,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0594906116378509,
|
| 34 |
+
"grad_norm": 0.49566933512687683,
|
| 35 |
+
"learning_rate": 3.571428571428571e-05,
|
| 36 |
+
"loss": 0.281,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07436326454731362,
|
| 41 |
+
"grad_norm": 0.45387113094329834,
|
| 42 |
+
"learning_rate": 4.4642857142857136e-05,
|
| 43 |
+
"loss": 0.2689,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08923591745677635,
|
| 48 |
+
"grad_norm": 0.43913352489471436,
|
| 49 |
+
"learning_rate": 5.357142857142857e-05,
|
| 50 |
+
"loss": 0.2506,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.10410857036623908,
|
| 55 |
+
"grad_norm": 0.7242547869682312,
|
| 56 |
+
"learning_rate": 6.25e-05,
|
| 57 |
+
"loss": 0.229,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.1189812232757018,
|
| 62 |
+
"grad_norm": 0.5109072923660278,
|
| 63 |
+
"learning_rate": 7.142857142857142e-05,
|
| 64 |
+
"loss": 0.2373,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.13385387618516453,
|
| 69 |
+
"grad_norm": 0.5291035175323486,
|
| 70 |
+
"learning_rate": 8.035714285714285e-05,
|
| 71 |
+
"loss": 0.2405,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.14872652909462725,
|
| 76 |
+
"grad_norm": 0.48036572337150574,
|
| 77 |
+
"learning_rate": 8.928571428571427e-05,
|
| 78 |
+
"loss": 0.2274,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.16359918200409,
|
| 83 |
+
"grad_norm": 0.3294093906879425,
|
| 84 |
+
"learning_rate": 9.82142857142857e-05,
|
| 85 |
+
"loss": 0.2038,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1784718349135527,
|
| 90 |
+
"grad_norm": 0.49968525767326355,
|
| 91 |
+
"learning_rate": 0.00010714285714285714,
|
| 92 |
+
"loss": 0.2084,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.19334448782301544,
|
| 97 |
+
"grad_norm": 0.32227209210395813,
|
| 98 |
+
"learning_rate": 0.00011607142857142857,
|
| 99 |
+
"loss": 0.1981,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.20821714073247816,
|
| 104 |
+
"grad_norm": 0.37266677618026733,
|
| 105 |
+
"learning_rate": 0.000125,
|
| 106 |
+
"loss": 0.2192,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.22308979364194087,
|
| 111 |
+
"grad_norm": 0.5228686928749084,
|
| 112 |
+
"learning_rate": 0.00013392857142857144,
|
| 113 |
+
"loss": 0.2014,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.2379624465514036,
|
| 118 |
+
"grad_norm": 0.4202245771884918,
|
| 119 |
+
"learning_rate": 0.00014285714285714284,
|
| 120 |
+
"loss": 0.1912,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.25283509946086635,
|
| 125 |
+
"grad_norm": 0.45801258087158203,
|
| 126 |
+
"learning_rate": 0.00015178571428571427,
|
| 127 |
+
"loss": 0.212,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.26770775237032907,
|
| 132 |
+
"grad_norm": 0.4326329827308655,
|
| 133 |
+
"learning_rate": 0.0001607142857142857,
|
| 134 |
+
"loss": 0.1973,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2825804052797918,
|
| 139 |
+
"grad_norm": 0.38971471786499023,
|
| 140 |
+
"learning_rate": 0.0001696428571428571,
|
| 141 |
+
"loss": 0.1907,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2974530581892545,
|
| 146 |
+
"grad_norm": 0.3728097975254059,
|
| 147 |
+
"learning_rate": 0.00017857142857142854,
|
| 148 |
+
"loss": 0.192,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3123257110987172,
|
| 153 |
+
"grad_norm": 0.34695690870285034,
|
| 154 |
+
"learning_rate": 0.00018749999999999998,
|
| 155 |
+
"loss": 0.1855,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.32719836400818,
|
| 160 |
+
"grad_norm": 0.41753408312797546,
|
| 161 |
+
"learning_rate": 0.0001964285714285714,
|
| 162 |
+
"loss": 0.1883,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3420710169176427,
|
| 167 |
+
"grad_norm": 0.27681878209114075,
|
| 168 |
+
"learning_rate": 0.00020535714285714284,
|
| 169 |
+
"loss": 0.1809,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.3569436698271054,
|
| 174 |
+
"grad_norm": 2.382871150970459,
|
| 175 |
+
"learning_rate": 0.00021428571428571427,
|
| 176 |
+
"loss": 0.1735,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3718163227365681,
|
| 181 |
+
"grad_norm": 160.2670440673828,
|
| 182 |
+
"learning_rate": 0.0002232142857142857,
|
| 183 |
+
"loss": 1.2159,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3866889756460309,
|
| 188 |
+
"grad_norm": 21.60050392150879,
|
| 189 |
+
"learning_rate": 0.00023214285714285714,
|
| 190 |
+
"loss": 5.4026,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4015616285554936,
|
| 195 |
+
"grad_norm": 13.928524017333984,
|
| 196 |
+
"learning_rate": 0.00024107142857142857,
|
| 197 |
+
"loss": 4.3573,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4164342814649563,
|
| 202 |
+
"grad_norm": 5.3707685470581055,
|
| 203 |
+
"learning_rate": 0.00025,
|
| 204 |
+
"loss": 3.2782,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.431306934374419,
|
| 209 |
+
"grad_norm": 5.556903839111328,
|
| 210 |
+
"learning_rate": 0.0002589285714285714,
|
| 211 |
+
"loss": 2.8033,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.44617958728388174,
|
| 216 |
+
"grad_norm": 2.512521505355835,
|
| 217 |
+
"learning_rate": 0.00026785714285714287,
|
| 218 |
+
"loss": 2.5486,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4610522401933445,
|
| 223 |
+
"grad_norm": 3.592169761657715,
|
| 224 |
+
"learning_rate": 0.0002767857142857143,
|
| 225 |
+
"loss": 2.2779,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.4759248931028072,
|
| 230 |
+
"grad_norm": 2.791459321975708,
|
| 231 |
+
"learning_rate": 0.0002857142857142857,
|
| 232 |
+
"loss": 2.1011,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.49079754601226994,
|
| 237 |
+
"grad_norm": 1.1407463550567627,
|
| 238 |
+
"learning_rate": 0.0002946428571428571,
|
| 239 |
+
"loss": 1.9929,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5056701989217327,
|
| 244 |
+
"grad_norm": 1.795841097831726,
|
| 245 |
+
"learning_rate": 0.0002999987048597728,
|
| 246 |
+
"loss": 1.8818,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5205428518311954,
|
| 251 |
+
"grad_norm": 1.4798821210861206,
|
| 252 |
+
"learning_rate": 0.00029998413478906613,
|
| 253 |
+
"loss": 1.772,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5354155047406581,
|
| 258 |
+
"grad_norm": 1.5337024927139282,
|
| 259 |
+
"learning_rate": 0.0002999533773001224,
|
| 260 |
+
"loss": 1.6782,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5502881576501208,
|
| 265 |
+
"grad_norm": 1.332065463066101,
|
| 266 |
+
"learning_rate": 0.00029990643571252174,
|
| 267 |
+
"loss": 1.6035,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5651608105595836,
|
| 272 |
+
"grad_norm": 1.0516103506088257,
|
| 273 |
+
"learning_rate": 0.00029984331509255415,
|
| 274 |
+
"loss": 1.5053,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.5800334634690463,
|
| 279 |
+
"grad_norm": 1.034192442893982,
|
| 280 |
+
"learning_rate": 0.00029976402225267247,
|
| 281 |
+
"loss": 1.3906,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.594906116378509,
|
| 286 |
+
"grad_norm": 1.2757515907287598,
|
| 287 |
+
"learning_rate": 0.0002996685657507577,
|
| 288 |
+
"loss": 1.2592,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6097787692879717,
|
| 293 |
+
"grad_norm": 0.8252782225608826,
|
| 294 |
+
"learning_rate": 0.000299556955889195,
|
| 295 |
+
"loss": 1.0907,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6246514221974344,
|
| 300 |
+
"grad_norm": 1.020588994026184,
|
| 301 |
+
"learning_rate": 0.0002994292047137618,
|
| 302 |
+
"loss": 0.9035,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6395240751068972,
|
| 307 |
+
"grad_norm": 0.5973761677742004,
|
| 308 |
+
"learning_rate": 0.0002992853260123278,
|
| 309 |
+
"loss": 0.7538,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.65439672801636,
|
| 314 |
+
"grad_norm": 0.6886543035507202,
|
| 315 |
+
"learning_rate": 0.0002991253353133668,
|
| 316 |
+
"loss": 0.6621,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.6692693809258227,
|
| 321 |
+
"grad_norm": 0.44221287965774536,
|
| 322 |
+
"learning_rate": 0.00029894924988428087,
|
| 323 |
+
"loss": 0.59,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.6841420338352854,
|
| 328 |
+
"grad_norm": 0.7888408899307251,
|
| 329 |
+
"learning_rate": 0.00029875708872953677,
|
| 330 |
+
"loss": 0.539,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.6990146867447481,
|
| 335 |
+
"grad_norm": 0.43110209703445435,
|
| 336 |
+
"learning_rate": 0.00029854887258861447,
|
| 337 |
+
"loss": 0.4903,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7138873396542108,
|
| 342 |
+
"grad_norm": 0.41334015130996704,
|
| 343 |
+
"learning_rate": 0.0002983246239337692,
|
| 344 |
+
"loss": 0.4488,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7287599925636735,
|
| 349 |
+
"grad_norm": 0.3482460379600525,
|
| 350 |
+
"learning_rate": 0.0002980843669676061,
|
| 351 |
+
"loss": 0.4165,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.7436326454731362,
|
| 356 |
+
"grad_norm": 0.3593901991844177,
|
| 357 |
+
"learning_rate": 0.0002978281276204675,
|
| 358 |
+
"loss": 0.3821,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.7436326454731362,
|
| 363 |
+
"eval_loss": 0.37597203254699707,
|
| 364 |
+
"eval_runtime": 212.4955,
|
| 365 |
+
"eval_samples_per_second": 22.499,
|
| 366 |
+
"eval_steps_per_second": 22.499,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.758505298382599,
|
| 371 |
+
"grad_norm": 0.4221905469894409,
|
| 372 |
+
"learning_rate": 0.00029755593354763516,
|
| 373 |
+
"loss": 0.3627,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.7733779512920618,
|
| 378 |
+
"grad_norm": 0.31105437874794006,
|
| 379 |
+
"learning_rate": 0.0002972678141263449,
|
| 380 |
+
"loss": 0.3346,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.7882506042015245,
|
| 385 |
+
"grad_norm": 0.2600822150707245,
|
| 386 |
+
"learning_rate": 0.000296963800452616,
|
| 387 |
+
"loss": 0.3217,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.8031232571109872,
|
| 392 |
+
"grad_norm": 0.21437157690525055,
|
| 393 |
+
"learning_rate": 0.0002966439253378957,
|
| 394 |
+
"loss": 0.3095,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.8179959100204499,
|
| 399 |
+
"grad_norm": 0.22641418874263763,
|
| 400 |
+
"learning_rate": 0.000296308223305517,
|
| 401 |
+
"loss": 0.2866,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.8328685629299126,
|
| 406 |
+
"grad_norm": 0.2200980931520462,
|
| 407 |
+
"learning_rate": 0.00029595673058697357,
|
| 408 |
+
"loss": 0.2579,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.8477412158393753,
|
| 413 |
+
"grad_norm": 0.21351036429405212,
|
| 414 |
+
"learning_rate": 0.0002955894851180086,
|
| 415 |
+
"loss": 0.2727,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.862613868748838,
|
| 420 |
+
"grad_norm": 0.2137759029865265,
|
| 421 |
+
"learning_rate": 0.0002952065265345211,
|
| 422 |
+
"loss": 0.2621,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.8774865216583008,
|
| 427 |
+
"grad_norm": 0.18923349678516388,
|
| 428 |
+
"learning_rate": 0.00029480789616828765,
|
| 429 |
+
"loss": 0.2647,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.8923591745677635,
|
| 434 |
+
"grad_norm": 0.1697588562965393,
|
| 435 |
+
"learning_rate": 0.00029439363704250176,
|
| 436 |
+
"loss": 0.2434,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9072318274772263,
|
| 441 |
+
"grad_norm": 0.15528830885887146,
|
| 442 |
+
"learning_rate": 0.0002939637938671306,
|
| 443 |
+
"loss": 0.2293,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.922104480386689,
|
| 448 |
+
"grad_norm": 0.43390974402427673,
|
| 449 |
+
"learning_rate": 0.0002935184130340893,
|
| 450 |
+
"loss": 0.228,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.9369771332961517,
|
| 455 |
+
"grad_norm": 0.2026420682668686,
|
| 456 |
+
"learning_rate": 0.000293057542612234,
|
| 457 |
+
"loss": 0.2355,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.9518497862056144,
|
| 462 |
+
"grad_norm": 0.16864228248596191,
|
| 463 |
+
"learning_rate": 0.00029258123234217435,
|
| 464 |
+
"loss": 0.2213,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.9667224391150772,
|
| 469 |
+
"grad_norm": 0.15947186946868896,
|
| 470 |
+
"learning_rate": 0.0002920895336309044,
|
| 471 |
+
"loss": 0.2079,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.9815950920245399,
|
| 476 |
+
"grad_norm": 0.21965055167675018,
|
| 477 |
+
"learning_rate": 0.0002915824995462551,
|
| 478 |
+
"loss": 0.2002,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.9964677449340026,
|
| 483 |
+
"grad_norm": 0.23223313689231873,
|
| 484 |
+
"learning_rate": 0.00029106018481116626,
|
| 485 |
+
"loss": 0.1983,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.0117122141662018,
|
| 490 |
+
"grad_norm": 0.26117920875549316,
|
| 491 |
+
"learning_rate": 0.00029052264579778063,
|
| 492 |
+
"loss": 0.2175,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.0265848670756645,
|
| 497 |
+
"grad_norm": 0.176736518740654,
|
| 498 |
+
"learning_rate": 0.00028996994052135996,
|
| 499 |
+
"loss": 0.1831,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.0414575199851273,
|
| 504 |
+
"grad_norm": 0.17873461544513702,
|
| 505 |
+
"learning_rate": 0.0002894021286340233,
|
| 506 |
+
"loss": 0.1784,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.05633017289459,
|
| 511 |
+
"grad_norm": 0.2646450996398926,
|
| 512 |
+
"learning_rate": 0.0002888192714183092,
|
| 513 |
+
"loss": 0.1784,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.0712028258040527,
|
| 518 |
+
"grad_norm": 0.16840551793575287,
|
| 519 |
+
"learning_rate": 0.00028822143178056114,
|
| 520 |
+
"loss": 0.1726,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.0860754787135156,
|
| 525 |
+
"grad_norm": 0.1423952877521515,
|
| 526 |
+
"learning_rate": 0.0002876086742441387,
|
| 527 |
+
"loss": 0.1608,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1009481316229783,
|
| 532 |
+
"grad_norm": 0.16237640380859375,
|
| 533 |
+
"learning_rate": 0.0002869810649424535,
|
| 534 |
+
"loss": 0.179,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.115820784532441,
|
| 539 |
+
"grad_norm": 0.158773735165596,
|
| 540 |
+
"learning_rate": 0.0002863386716118316,
|
| 541 |
+
"loss": 0.1742,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.1306934374419038,
|
| 546 |
+
"grad_norm": 0.17627516388893127,
|
| 547 |
+
"learning_rate": 0.0002856815635842029,
|
| 548 |
+
"loss": 0.1821,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.1455660903513665,
|
| 553 |
+
"grad_norm": 0.23613831400871277,
|
| 554 |
+
"learning_rate": 0.00028500981177961816,
|
| 555 |
+
"loss": 0.156,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.1604387432608292,
|
| 560 |
+
"grad_norm": 0.16501256823539734,
|
| 561 |
+
"learning_rate": 0.0002843234886985951,
|
| 562 |
+
"loss": 0.1517,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.175311396170292,
|
| 567 |
+
"grad_norm": 0.2365158647298813,
|
| 568 |
+
"learning_rate": 0.00028362266841429345,
|
| 569 |
+
"loss": 0.1391,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.1901840490797546,
|
| 574 |
+
"grad_norm": 0.17508777976036072,
|
| 575 |
+
"learning_rate": 0.00028290742656452014,
|
| 576 |
+
"loss": 0.1434,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.2050567019892173,
|
| 581 |
+
"grad_norm": 0.145797461271286,
|
| 582 |
+
"learning_rate": 0.0002821778403435663,
|
| 583 |
+
"loss": 0.1607,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.21992935489868,
|
| 588 |
+
"grad_norm": 0.15968403220176697,
|
| 589 |
+
"learning_rate": 0.00028143398849387577,
|
| 590 |
+
"loss": 0.1536,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.2348020078081428,
|
| 595 |
+
"grad_norm": 0.1553070992231369,
|
| 596 |
+
"learning_rate": 0.00028067595129754647,
|
| 597 |
+
"loss": 0.1481,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.2496746607176055,
|
| 602 |
+
"grad_norm": 0.1769135743379593,
|
| 603 |
+
"learning_rate": 0.0002799038105676658,
|
| 604 |
+
"loss": 0.1285,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.2645473136270682,
|
| 609 |
+
"grad_norm": 0.1639111191034317,
|
| 610 |
+
"learning_rate": 0.0002791176496394808,
|
| 611 |
+
"loss": 0.144,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.279419966536531,
|
| 616 |
+
"grad_norm": 0.19045153260231018,
|
| 617 |
+
"learning_rate": 0.00027831755336140416,
|
| 618 |
+
"loss": 0.1347,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.2942926194459936,
|
| 623 |
+
"grad_norm": 0.18079642951488495,
|
| 624 |
+
"learning_rate": 0.00027750360808585637,
|
| 625 |
+
"loss": 0.1254,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3091652723554563,
|
| 630 |
+
"grad_norm": 0.18368874490261078,
|
| 631 |
+
"learning_rate": 0.00027667590165994613,
|
| 632 |
+
"loss": 0.1289,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.324037925264919,
|
| 637 |
+
"grad_norm": 0.20005619525909424,
|
| 638 |
+
"learning_rate": 0.00027583452341598935,
|
| 639 |
+
"loss": 0.1246,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.338910578174382,
|
| 644 |
+
"grad_norm": 0.1317131668329239,
|
| 645 |
+
"learning_rate": 0.0002749795641618673,
|
| 646 |
+
"loss": 0.1238,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.3537832310838445,
|
| 651 |
+
"grad_norm": 0.15287995338439941,
|
| 652 |
+
"learning_rate": 0.00027411111617122656,
|
| 653 |
+
"loss": 0.1224,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.3686558839933074,
|
| 658 |
+
"grad_norm": 0.1613466739654541,
|
| 659 |
+
"learning_rate": 0.0002732292731735196,
|
| 660 |
+
"loss": 0.1178,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.3835285369027701,
|
| 665 |
+
"grad_norm": 0.1685304194688797,
|
| 666 |
+
"learning_rate": 0.000272334130343889,
|
| 667 |
+
"loss": 0.1201,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.3984011898122328,
|
| 672 |
+
"grad_norm": 0.19208119809627533,
|
| 673 |
+
"learning_rate": 0.0002714257842928956,
|
| 674 |
+
"loss": 0.1103,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.4132738427216955,
|
| 679 |
+
"grad_norm": 0.17899583280086517,
|
| 680 |
+
"learning_rate": 0.00027050433305609125,
|
| 681 |
+
"loss": 0.1128,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.4281464956311583,
|
| 686 |
+
"grad_norm": 0.19848547875881195,
|
| 687 |
+
"learning_rate": 0.0002695698760834384,
|
| 688 |
+
"loss": 0.1112,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.443019148540621,
|
| 693 |
+
"grad_norm": 0.1710231602191925,
|
| 694 |
+
"learning_rate": 0.0002686225142285762,
|
| 695 |
+
"loss": 0.1107,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.4578918014500837,
|
| 700 |
+
"grad_norm": 0.1552249938249588,
|
| 701 |
+
"learning_rate": 0.0002676623497379363,
|
| 702 |
+
"loss": 0.0984,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.4727644543595464,
|
| 707 |
+
"grad_norm": 0.1702568084001541,
|
| 708 |
+
"learning_rate": 0.0002666894862397072,
|
| 709 |
+
"loss": 0.1109,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.487637107269009,
|
| 714 |
+
"grad_norm": 0.12360525131225586,
|
| 715 |
+
"learning_rate": 0.00026570402873264996,
|
| 716 |
+
"loss": 0.1018,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.487637107269009,
|
| 721 |
+
"eval_loss": 0.10193677991628647,
|
| 722 |
+
"eval_runtime": 212.152,
|
| 723 |
+
"eval_samples_per_second": 22.536,
|
| 724 |
+
"eval_steps_per_second": 22.536,
|
| 725 |
+
"step": 1000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.5025097601784718,
|
| 729 |
+
"grad_norm": 0.14356306195259094,
|
| 730 |
+
"learning_rate": 0.0002647060835747659,
|
| 731 |
+
"loss": 0.101,
|
| 732 |
+
"step": 1010
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.5173824130879345,
|
| 736 |
+
"grad_norm": 0.12723973393440247,
|
| 737 |
+
"learning_rate": 0.00026369575847181795,
|
| 738 |
+
"loss": 0.095,
|
| 739 |
+
"step": 1020
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.5322550659973972,
|
| 743 |
+
"grad_norm": 0.12857410311698914,
|
| 744 |
+
"learning_rate": 0.0002626731624657058,
|
| 745 |
+
"loss": 0.0915,
|
| 746 |
+
"step": 1030
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.54712771890686,
|
| 750 |
+
"grad_norm": 0.1593610793352127,
|
| 751 |
+
"learning_rate": 0.0002616384059226977,
|
| 752 |
+
"loss": 0.0993,
|
| 753 |
+
"step": 1040
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.562000371816323,
|
| 757 |
+
"grad_norm": 0.11687605082988739,
|
| 758 |
+
"learning_rate": 0.0002605916005215186,
|
| 759 |
+
"loss": 0.0894,
|
| 760 |
+
"step": 1050
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.5768730247257854,
|
| 764 |
+
"grad_norm": 0.1873299479484558,
|
| 765 |
+
"learning_rate": 0.0002595328592412969,
|
| 766 |
+
"loss": 0.097,
|
| 767 |
+
"step": 1060
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.5917456776352483,
|
| 771 |
+
"grad_norm": 0.1516319364309311,
|
| 772 |
+
"learning_rate": 0.00025846229634937136,
|
| 773 |
+
"loss": 0.0931,
|
| 774 |
+
"step": 1070
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.6066183305447108,
|
| 778 |
+
"grad_norm": 0.1431397646665573,
|
| 779 |
+
"learning_rate": 0.0002573800273889577,
|
| 780 |
+
"loss": 0.0918,
|
| 781 |
+
"step": 1080
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.6214909834541738,
|
| 785 |
+
"grad_norm": 0.17975349724292755,
|
| 786 |
+
"learning_rate": 0.0002562861691666793,
|
| 787 |
+
"loss": 0.0892,
|
| 788 |
+
"step": 1090
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.6363636363636362,
|
| 792 |
+
"grad_norm": 0.1414797306060791,
|
| 793 |
+
"learning_rate": 0.0002551808397399597,
|
| 794 |
+
"loss": 0.0952,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.6512362892730992,
|
| 799 |
+
"grad_norm": 0.151850625872612,
|
| 800 |
+
"learning_rate": 0.0002540641584042812,
|
| 801 |
+
"loss": 0.1008,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.6661089421825617,
|
| 806 |
+
"grad_norm": 0.1266675442457199,
|
| 807 |
+
"learning_rate": 0.00025293624568031,
|
| 808 |
+
"loss": 0.0782,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.6809815950920246,
|
| 813 |
+
"grad_norm": 0.12076599150896072,
|
| 814 |
+
"learning_rate": 0.0002517972233008882,
|
| 815 |
+
"loss": 0.0772,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.6958542480014873,
|
| 820 |
+
"grad_norm": 0.125094935297966,
|
| 821 |
+
"learning_rate": 0.0002506472141978955,
|
| 822 |
+
"loss": 0.0837,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.71072690091095,
|
| 827 |
+
"grad_norm": 0.13272984325885773,
|
| 828 |
+
"learning_rate": 0.0002494863424889819,
|
| 829 |
+
"loss": 0.0736,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.7255995538204127,
|
| 834 |
+
"grad_norm": 0.16893050074577332,
|
| 835 |
+
"learning_rate": 0.00024831473346417153,
|
| 836 |
+
"loss": 0.0856,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.7404722067298755,
|
| 841 |
+
"grad_norm": 0.11702137440443039,
|
| 842 |
+
"learning_rate": 0.00024713251357234053,
|
| 843 |
+
"loss": 0.0799,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.7553448596393382,
|
| 848 |
+
"grad_norm": 0.13682794570922852,
|
| 849 |
+
"learning_rate": 0.00024593981040756997,
|
| 850 |
+
"loss": 0.089,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.7702175125488009,
|
| 855 |
+
"grad_norm": 0.13676613569259644,
|
| 856 |
+
"learning_rate": 0.0002447367526953746,
|
| 857 |
+
"loss": 0.0797,
|
| 858 |
+
"step": 1190
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.7850901654582636,
|
| 862 |
+
"grad_norm": 0.13324877619743347,
|
| 863 |
+
"learning_rate": 0.00024352347027881003,
|
| 864 |
+
"loss": 0.0792,
|
| 865 |
+
"step": 1200
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.7999628183677263,
|
| 869 |
+
"grad_norm": 0.11255478858947754,
|
| 870 |
+
"learning_rate": 0.00024230009410445893,
|
| 871 |
+
"loss": 0.0763,
|
| 872 |
+
"step": 1210
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.814835471277189,
|
| 876 |
+
"grad_norm": 0.10950371623039246,
|
| 877 |
+
"learning_rate": 0.0002410667562082985,
|
| 878 |
+
"loss": 0.0663,
|
| 879 |
+
"step": 1220
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.8297081241866517,
|
| 883 |
+
"grad_norm": 0.11777317523956299,
|
| 884 |
+
"learning_rate": 0.00023982358970145004,
|
| 885 |
+
"loss": 0.0694,
|
| 886 |
+
"step": 1230
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.8445807770961147,
|
| 890 |
+
"grad_norm": 0.1194106712937355,
|
| 891 |
+
"learning_rate": 0.00023857072875581244,
|
| 892 |
+
"loss": 0.0703,
|
| 893 |
+
"step": 1240
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.8594534300055772,
|
| 897 |
+
"grad_norm": 0.11233114451169968,
|
| 898 |
+
"learning_rate": 0.00023730830858958177,
|
| 899 |
+
"loss": 0.0655,
|
| 900 |
+
"step": 1250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.87432608291504,
|
| 904 |
+
"grad_norm": 0.11641702055931091,
|
| 905 |
+
"learning_rate": 0.00023603646545265687,
|
| 906 |
+
"loss": 0.0645,
|
| 907 |
+
"step": 1260
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.8891987358245026,
|
| 911 |
+
"grad_norm": 0.1414889097213745,
|
| 912 |
+
"learning_rate": 0.00023475533661193495,
|
| 913 |
+
"loss": 0.068,
|
| 914 |
+
"step": 1270
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.9040713887339655,
|
| 918 |
+
"grad_norm": 0.10632241517305374,
|
| 919 |
+
"learning_rate": 0.00023346506033649614,
|
| 920 |
+
"loss": 0.064,
|
| 921 |
+
"step": 1280
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.918944041643428,
|
| 925 |
+
"grad_norm": 0.10176625102758408,
|
| 926 |
+
"learning_rate": 0.0002321657758826807,
|
| 927 |
+
"loss": 0.062,
|
| 928 |
+
"step": 1290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.933816694552891,
|
| 932 |
+
"grad_norm": 0.09434150904417038,
|
| 933 |
+
"learning_rate": 0.00023085762347905943,
|
| 934 |
+
"loss": 0.0684,
|
| 935 |
+
"step": 1300
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.9486893474623534,
|
| 939 |
+
"grad_norm": 0.12967799603939056,
|
| 940 |
+
"learning_rate": 0.00022954074431129915,
|
| 941 |
+
"loss": 0.0605,
|
| 942 |
+
"step": 1310
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.9635620003718164,
|
| 946 |
+
"grad_norm": 0.1181391179561615,
|
| 947 |
+
"learning_rate": 0.0002282152805069247,
|
| 948 |
+
"loss": 0.0654,
|
| 949 |
+
"step": 1320
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.978434653281279,
|
| 953 |
+
"grad_norm": 0.10801093280315399,
|
| 954 |
+
"learning_rate": 0.00022688137511997977,
|
| 955 |
+
"loss": 0.07,
|
| 956 |
+
"step": 1330
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.9933073061907418,
|
| 960 |
+
"grad_norm": 0.11437591165304184,
|
| 961 |
+
"learning_rate": 0.00022553917211558713,
|
| 962 |
+
"loss": 0.0578,
|
| 963 |
+
"step": 1340
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 2.008551775422941,
|
| 967 |
+
"grad_norm": 0.11018254607915878,
|
| 968 |
+
"learning_rate": 0.0002241888163544111,
|
| 969 |
+
"loss": 0.0565,
|
| 970 |
+
"step": 1350
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 2.0234244283324037,
|
| 974 |
+
"grad_norm": 0.08331198990345001,
|
| 975 |
+
"learning_rate": 0.0002228304535770228,
|
| 976 |
+
"loss": 0.0399,
|
| 977 |
+
"step": 1360
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 2.0382970812418666,
|
| 981 |
+
"grad_norm": 0.09547814726829529,
|
| 982 |
+
"learning_rate": 0.00022146423038817102,
|
| 983 |
+
"loss": 0.0438,
|
| 984 |
+
"step": 1370
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.053169734151329,
|
| 988 |
+
"grad_norm": 0.10641171038150787,
|
| 989 |
+
"learning_rate": 0.00022009029424095928,
|
| 990 |
+
"loss": 0.0384,
|
| 991 |
+
"step": 1380
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 2.068042387060792,
|
| 995 |
+
"grad_norm": 0.10844069719314575,
|
| 996 |
+
"learning_rate": 0.0002187087934209318,
|
| 997 |
+
"loss": 0.044,
|
| 998 |
+
"step": 1390
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.0829150399702545,
|
| 1002 |
+
"grad_norm": 0.10333788394927979,
|
| 1003 |
+
"learning_rate": 0.00021731987703006933,
|
| 1004 |
+
"loss": 0.041,
|
| 1005 |
+
"step": 1400
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 2.0977876928797174,
|
| 1009 |
+
"grad_norm": 0.10635129362344742,
|
| 1010 |
+
"learning_rate": 0.0002159236949706967,
|
| 1011 |
+
"loss": 0.04,
|
| 1012 |
+
"step": 1410
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.11266034578918,
|
| 1016 |
+
"grad_norm": 0.09010270237922668,
|
| 1017 |
+
"learning_rate": 0.00021452039792930474,
|
| 1018 |
+
"loss": 0.0402,
|
| 1019 |
+
"step": 1420
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 2.127532998698643,
|
| 1023 |
+
"grad_norm": 0.09274252504110336,
|
| 1024 |
+
"learning_rate": 0.00021311013736028658,
|
| 1025 |
+
"loss": 0.0384,
|
| 1026 |
+
"step": 1430
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.1424056516081054,
|
| 1030 |
+
"grad_norm": 0.08550871163606644,
|
| 1031 |
+
"learning_rate": 0.00021169306546959174,
|
| 1032 |
+
"loss": 0.0428,
|
| 1033 |
+
"step": 1440
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.1572783045175683,
|
| 1037 |
+
"grad_norm": 0.10152186453342438,
|
| 1038 |
+
"learning_rate": 0.00021026933519829896,
|
| 1039 |
+
"loss": 0.0442,
|
| 1040 |
+
"step": 1450
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.1721509574270312,
|
| 1044 |
+
"grad_norm": 0.08528181910514832,
|
| 1045 |
+
"learning_rate": 0.00020883910020610957,
|
| 1046 |
+
"loss": 0.0375,
|
| 1047 |
+
"step": 1460
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 2.1870236103364937,
|
| 1051 |
+
"grad_norm": 0.09736708551645279,
|
| 1052 |
+
"learning_rate": 0.00020740251485476345,
|
| 1053 |
+
"loss": 0.0387,
|
| 1054 |
+
"step": 1470
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.2018962632459567,
|
| 1058 |
+
"grad_norm": 0.09133671224117279,
|
| 1059 |
+
"learning_rate": 0.00020595973419137908,
|
| 1060 |
+
"loss": 0.0373,
|
| 1061 |
+
"step": 1480
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 2.216768916155419,
|
| 1065 |
+
"grad_norm": 0.08406363427639008,
|
| 1066 |
+
"learning_rate": 0.00020451091393171964,
|
| 1067 |
+
"loss": 0.0381,
|
| 1068 |
+
"step": 1490
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.231641569064882,
|
| 1072 |
+
"grad_norm": 0.08503925055265427,
|
| 1073 |
+
"learning_rate": 0.00020305621044338718,
|
| 1074 |
+
"loss": 0.0376,
|
| 1075 |
+
"step": 1500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 2.231641569064882,
|
| 1079 |
+
"eval_loss": 0.051042910665273666,
|
| 1080 |
+
"eval_runtime": 212.5441,
|
| 1081 |
+
"eval_samples_per_second": 22.494,
|
| 1082 |
+
"eval_steps_per_second": 22.494,
|
| 1083 |
+
"step": 1500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 2.2465142219743446,
|
| 1087 |
+
"grad_norm": 0.09201103448867798,
|
| 1088 |
+
"learning_rate": 0.00020159578072894606,
|
| 1089 |
+
"loss": 0.0393,
|
| 1090 |
+
"step": 1510
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.2613868748838075,
|
| 1094 |
+
"grad_norm": 0.09499834477901459,
|
| 1095 |
+
"learning_rate": 0.00020012978240897814,
|
| 1096 |
+
"loss": 0.0346,
|
| 1097 |
+
"step": 1520
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 2.27625952779327,
|
| 1101 |
+
"grad_norm": 0.09396501630544662,
|
| 1102 |
+
"learning_rate": 0.00019865837370507106,
|
| 1103 |
+
"loss": 0.039,
|
| 1104 |
+
"step": 1530
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.291132180702733,
|
| 1108 |
+
"grad_norm": 0.08983522653579712,
|
| 1109 |
+
"learning_rate": 0.00019718171342274205,
|
| 1110 |
+
"loss": 0.0387,
|
| 1111 |
+
"step": 1540
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 2.3060048336121954,
|
| 1115 |
+
"grad_norm": 0.1118871420621872,
|
| 1116 |
+
"learning_rate": 0.00019569996093429814,
|
| 1117 |
+
"loss": 0.0379,
|
| 1118 |
+
"step": 1550
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.3208774865216584,
|
| 1122 |
+
"grad_norm": 0.08434595167636871,
|
| 1123 |
+
"learning_rate": 0.00019421327616163563,
|
| 1124 |
+
"loss": 0.0372,
|
| 1125 |
+
"step": 1560
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 2.335750139431121,
|
| 1129 |
+
"grad_norm": 0.0915694460272789,
|
| 1130 |
+
"learning_rate": 0.00019272181955898017,
|
| 1131 |
+
"loss": 0.036,
|
| 1132 |
+
"step": 1570
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.350622792340584,
|
| 1136 |
+
"grad_norm": 0.08459066599607468,
|
| 1137 |
+
"learning_rate": 0.0001912257520955692,
|
| 1138 |
+
"loss": 0.0363,
|
| 1139 |
+
"step": 1580
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 2.3654954452500463,
|
| 1143 |
+
"grad_norm": 0.09195558726787567,
|
| 1144 |
+
"learning_rate": 0.00018972523523827907,
|
| 1145 |
+
"loss": 0.0389,
|
| 1146 |
+
"step": 1590
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.3803680981595092,
|
| 1150 |
+
"grad_norm": 0.09830203652381897,
|
| 1151 |
+
"learning_rate": 0.0001882204309341982,
|
| 1152 |
+
"loss": 0.0373,
|
| 1153 |
+
"step": 1600
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 2.3952407510689717,
|
| 1157 |
+
"grad_norm": 0.08541320264339447,
|
| 1158 |
+
"learning_rate": 0.00018671150159314855,
|
| 1159 |
+
"loss": 0.0342,
|
| 1160 |
+
"step": 1610
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.4101134039784347,
|
| 1164 |
+
"grad_norm": 0.08817029744386673,
|
| 1165 |
+
"learning_rate": 0.00018519861007015729,
|
| 1166 |
+
"loss": 0.0371,
|
| 1167 |
+
"step": 1620
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 2.4249860568878976,
|
| 1171 |
+
"grad_norm": 0.08839129656553268,
|
| 1172 |
+
"learning_rate": 0.00018368191964788,
|
| 1173 |
+
"loss": 0.0355,
|
| 1174 |
+
"step": 1630
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 2.43985870979736,
|
| 1178 |
+
"grad_norm": 0.08589951694011688,
|
| 1179 |
+
"learning_rate": 0.00018216159401897812,
|
| 1180 |
+
"loss": 0.0339,
|
| 1181 |
+
"step": 1640
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 2.454731362706823,
|
| 1185 |
+
"grad_norm": 0.09998754411935806,
|
| 1186 |
+
"learning_rate": 0.00018063779726845203,
|
| 1187 |
+
"loss": 0.0339,
|
| 1188 |
+
"step": 1650
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.4696040156162855,
|
| 1192 |
+
"grad_norm": 0.08363664150238037,
|
| 1193 |
+
"learning_rate": 0.0001791106938559317,
|
| 1194 |
+
"loss": 0.0357,
|
| 1195 |
+
"step": 1660
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 2.4844766685257484,
|
| 1199 |
+
"grad_norm": 0.08930620551109314,
|
| 1200 |
+
"learning_rate": 0.00017758044859792705,
|
| 1201 |
+
"loss": 0.0347,
|
| 1202 |
+
"step": 1670
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.499349321435211,
|
| 1206 |
+
"grad_norm": 0.08270251750946045,
|
| 1207 |
+
"learning_rate": 0.00017604722665003956,
|
| 1208 |
+
"loss": 0.0332,
|
| 1209 |
+
"step": 1680
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 2.514221974344674,
|
| 1213 |
+
"grad_norm": 0.09085123986005783,
|
| 1214 |
+
"learning_rate": 0.00017451119348913744,
|
| 1215 |
+
"loss": 0.0357,
|
| 1216 |
+
"step": 1690
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 2.5290946272541364,
|
| 1220 |
+
"grad_norm": 0.0897296592593193,
|
| 1221 |
+
"learning_rate": 0.00017297251489549638,
|
| 1222 |
+
"loss": 0.0368,
|
| 1223 |
+
"step": 1700
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 2.5439672801635993,
|
| 1227 |
+
"grad_norm": 0.07172433286905289,
|
| 1228 |
+
"learning_rate": 0.000171431356934907,
|
| 1229 |
+
"loss": 0.0371,
|
| 1230 |
+
"step": 1710
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 2.558839933073062,
|
| 1234 |
+
"grad_norm": 0.0848449245095253,
|
| 1235 |
+
"learning_rate": 0.0001698878859407519,
|
| 1236 |
+
"loss": 0.032,
|
| 1237 |
+
"step": 1720
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 2.5737125859825247,
|
| 1241 |
+
"grad_norm": 0.08270355314016342,
|
| 1242 |
+
"learning_rate": 0.00016834226849605371,
|
| 1243 |
+
"loss": 0.0333,
|
| 1244 |
+
"step": 1730
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.588585238891987,
|
| 1248 |
+
"grad_norm": 0.07130729407072067,
|
| 1249 |
+
"learning_rate": 0.00016679467141549617,
|
| 1250 |
+
"loss": 0.0324,
|
| 1251 |
+
"step": 1740
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 2.60345789180145,
|
| 1255 |
+
"grad_norm": 0.07863139361143112,
|
| 1256 |
+
"learning_rate": 0.00016524526172742026,
|
| 1257 |
+
"loss": 0.0295,
|
| 1258 |
+
"step": 1750
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 2.6183305447109126,
|
| 1262 |
+
"grad_norm": 0.08600688725709915,
|
| 1263 |
+
"learning_rate": 0.00016369420665579725,
|
| 1264 |
+
"loss": 0.0342,
|
| 1265 |
+
"step": 1760
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 2.6332031976203756,
|
| 1269 |
+
"grad_norm": 0.10146727412939072,
|
| 1270 |
+
"learning_rate": 0.0001621416736021805,
|
| 1271 |
+
"loss": 0.032,
|
| 1272 |
+
"step": 1770
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 2.648075850529838,
|
| 1276 |
+
"grad_norm": 0.0812121257185936,
|
| 1277 |
+
"learning_rate": 0.00016058783012763844,
|
| 1278 |
+
"loss": 0.0341,
|
| 1279 |
+
"step": 1780
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 2.662948503439301,
|
| 1283 |
+
"grad_norm": 0.0973149985074997,
|
| 1284 |
+
"learning_rate": 0.00015903284393466987,
|
| 1285 |
+
"loss": 0.0313,
|
| 1286 |
+
"step": 1790
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.677821156348764,
|
| 1290 |
+
"grad_norm": 0.0835902988910675,
|
| 1291 |
+
"learning_rate": 0.00015747688284910457,
|
| 1292 |
+
"loss": 0.0298,
|
| 1293 |
+
"step": 1800
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 2.6926938092582264,
|
| 1297 |
+
"grad_norm": 0.07972200214862823,
|
| 1298 |
+
"learning_rate": 0.00015592011480198992,
|
| 1299 |
+
"loss": 0.0346,
|
| 1300 |
+
"step": 1810
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 2.707566462167689,
|
| 1304 |
+
"grad_norm": 0.07594762742519379,
|
| 1305 |
+
"learning_rate": 0.0001543627078114667,
|
| 1306 |
+
"loss": 0.0338,
|
| 1307 |
+
"step": 1820
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 2.722439115077152,
|
| 1311 |
+
"grad_norm": 0.07757771015167236,
|
| 1312 |
+
"learning_rate": 0.00015280482996463533,
|
| 1313 |
+
"loss": 0.0315,
|
| 1314 |
+
"step": 1830
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 2.737311767986615,
|
| 1318 |
+
"grad_norm": 0.06432707607746124,
|
| 1319 |
+
"learning_rate": 0.00015124664939941457,
|
| 1320 |
+
"loss": 0.0319,
|
| 1321 |
+
"step": 1840
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 2.7521844208960773,
|
| 1325 |
+
"grad_norm": 0.07696104794740677,
|
| 1326 |
+
"learning_rate": 0.00014968833428639474,
|
| 1327 |
+
"loss": 0.0301,
|
| 1328 |
+
"step": 1850
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.7670570738055402,
|
| 1332 |
+
"grad_norm": 0.07426641881465912,
|
| 1333 |
+
"learning_rate": 0.00014813005281068774,
|
| 1334 |
+
"loss": 0.0285,
|
| 1335 |
+
"step": 1860
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 2.7819297267150027,
|
| 1339 |
+
"grad_norm": 0.0765393078327179,
|
| 1340 |
+
"learning_rate": 0.00014657197315377495,
|
| 1341 |
+
"loss": 0.0313,
|
| 1342 |
+
"step": 1870
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 2.7968023796244657,
|
| 1346 |
+
"grad_norm": 0.07151610404253006,
|
| 1347 |
+
"learning_rate": 0.00014501426347535598,
|
| 1348 |
+
"loss": 0.03,
|
| 1349 |
+
"step": 1880
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 2.811675032533928,
|
| 1353 |
+
"grad_norm": 0.07834175229072571,
|
| 1354 |
+
"learning_rate": 0.0001434570918951996,
|
| 1355 |
+
"loss": 0.0286,
|
| 1356 |
+
"step": 1890
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 2.826547685443391,
|
| 1360 |
+
"grad_norm": 0.09932053834199905,
|
| 1361 |
+
"learning_rate": 0.00014190062647499892,
|
| 1362 |
+
"loss": 0.0307,
|
| 1363 |
+
"step": 1900
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 2.8414203383528536,
|
| 1367 |
+
"grad_norm": 0.08595503121614456,
|
| 1368 |
+
"learning_rate": 0.00014034503520023297,
|
| 1369 |
+
"loss": 0.0306,
|
| 1370 |
+
"step": 1910
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 2.8562929912623165,
|
| 1374 |
+
"grad_norm": 0.08349858224391937,
|
| 1375 |
+
"learning_rate": 0.00013879048596203636,
|
| 1376 |
+
"loss": 0.0306,
|
| 1377 |
+
"step": 1920
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 2.871165644171779,
|
| 1381 |
+
"grad_norm": 0.07905739545822144,
|
| 1382 |
+
"learning_rate": 0.0001372371465390794,
|
| 1383 |
+
"loss": 0.0305,
|
| 1384 |
+
"step": 1930
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 2.886038297081242,
|
| 1388 |
+
"grad_norm": 0.06820567697286606,
|
| 1389 |
+
"learning_rate": 0.0001356851845794598,
|
| 1390 |
+
"loss": 0.0276,
|
| 1391 |
+
"step": 1940
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 2.9009109499907044,
|
| 1395 |
+
"grad_norm": 0.07227708399295807,
|
| 1396 |
+
"learning_rate": 0.00013413476758260934,
|
| 1397 |
+
"loss": 0.0267,
|
| 1398 |
+
"step": 1950
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 2.9157836029001674,
|
| 1402 |
+
"grad_norm": 0.09035148471593857,
|
| 1403 |
+
"learning_rate": 0.00013258606288121542,
|
| 1404 |
+
"loss": 0.0287,
|
| 1405 |
+
"step": 1960
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 2.9306562558096303,
|
| 1409 |
+
"grad_norm": 0.08626757562160492,
|
| 1410 |
+
"learning_rate": 0.00013103923762316198,
|
| 1411 |
+
"loss": 0.0298,
|
| 1412 |
+
"step": 1970
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 2.945528908719093,
|
| 1416 |
+
"grad_norm": 0.0765102431178093,
|
| 1417 |
+
"learning_rate": 0.00012949445875348902,
|
| 1418 |
+
"loss": 0.0274,
|
| 1419 |
+
"step": 1980
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 2.9604015616285553,
|
| 1423 |
+
"grad_norm": 0.08610813319683075,
|
| 1424 |
+
"learning_rate": 0.00012795189299637483,
|
| 1425 |
+
"loss": 0.0283,
|
| 1426 |
+
"step": 1990
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.975274214538018,
|
| 1430 |
+
"grad_norm": 0.08020433783531189,
|
| 1431 |
+
"learning_rate": 0.00012641170683714222,
|
| 1432 |
+
"loss": 0.0267,
|
| 1433 |
+
"step": 2000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 2.975274214538018,
|
| 1437 |
+
"eval_loss": 0.0338360071182251,
|
| 1438 |
+
"eval_runtime": 212.7237,
|
| 1439 |
+
"eval_samples_per_second": 22.475,
|
| 1440 |
+
"eval_steps_per_second": 22.475,
|
| 1441 |
+
"step": 2000
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.990146867447481,
|
| 1445 |
+
"grad_norm": 0.06885667890310287,
|
| 1446 |
+
"learning_rate": 0.00012487406650428954,
|
| 1447 |
+
"loss": 0.0277,
|
| 1448 |
+
"step": 2010
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 3.00539133667968,
|
| 1452 |
+
"grad_norm": 0.07658534497022629,
|
| 1453 |
+
"learning_rate": 0.00012333913795155053,
|
| 1454 |
+
"loss": 0.0251,
|
| 1455 |
+
"step": 2020
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 3.020263989589143,
|
| 1459 |
+
"grad_norm": 0.06449634581804276,
|
| 1460 |
+
"learning_rate": 0.00012180708683998321,
|
| 1461 |
+
"loss": 0.0147,
|
| 1462 |
+
"step": 2030
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 3.0351366424986055,
|
| 1466 |
+
"grad_norm": 0.06312290579080582,
|
| 1467 |
+
"learning_rate": 0.00012027807852009038,
|
| 1468 |
+
"loss": 0.0157,
|
| 1469 |
+
"step": 2040
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 3.0500092954080684,
|
| 1473 |
+
"grad_norm": 0.07343071699142456,
|
| 1474 |
+
"learning_rate": 0.00011875227801397381,
|
| 1475 |
+
"loss": 0.0149,
|
| 1476 |
+
"step": 2050
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 3.064881948317531,
|
| 1480 |
+
"grad_norm": 0.06489036977291107,
|
| 1481 |
+
"learning_rate": 0.00011722984999752392,
|
| 1482 |
+
"loss": 0.0155,
|
| 1483 |
+
"step": 2060
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 3.079754601226994,
|
| 1487 |
+
"grad_norm": 0.06041651591658592,
|
| 1488 |
+
"learning_rate": 0.00011571095878264658,
|
| 1489 |
+
"loss": 0.0139,
|
| 1490 |
+
"step": 2070
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 3.094627254136457,
|
| 1494 |
+
"grad_norm": 0.07048339396715164,
|
| 1495 |
+
"learning_rate": 0.00011419576829952933,
|
| 1496 |
+
"loss": 0.014,
|
| 1497 |
+
"step": 2080
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 3.1094999070459193,
|
| 1501 |
+
"grad_norm": 0.05680292099714279,
|
| 1502 |
+
"learning_rate": 0.00011268444207894902,
|
| 1503 |
+
"loss": 0.0133,
|
| 1504 |
+
"step": 2090
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 3.124372559955382,
|
| 1508 |
+
"grad_norm": 0.0727318823337555,
|
| 1509 |
+
"learning_rate": 0.00011117714323462186,
|
| 1510 |
+
"loss": 0.0147,
|
| 1511 |
+
"step": 2100
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 3.1392452128648447,
|
| 1515 |
+
"grad_norm": 0.054686855524778366,
|
| 1516 |
+
"learning_rate": 0.00010967403444559963,
|
| 1517 |
+
"loss": 0.0143,
|
| 1518 |
+
"step": 2110
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 3.1541178657743076,
|
| 1522 |
+
"grad_norm": 0.05729954317212105,
|
| 1523 |
+
"learning_rate": 0.00010817527793871143,
|
| 1524 |
+
"loss": 0.0134,
|
| 1525 |
+
"step": 2120
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 3.16899051868377,
|
| 1529 |
+
"grad_norm": 0.08314567804336548,
|
| 1530 |
+
"learning_rate": 0.00010668103547105553,
|
| 1531 |
+
"loss": 0.0148,
|
| 1532 |
+
"step": 2130
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 3.183863171593233,
|
| 1536 |
+
"grad_norm": 0.05523039028048515,
|
| 1537 |
+
"learning_rate": 0.00010519146831254088,
|
| 1538 |
+
"loss": 0.0129,
|
| 1539 |
+
"step": 2140
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 3.1987358245026956,
|
| 1543 |
+
"grad_norm": 0.05546917766332626,
|
| 1544 |
+
"learning_rate": 0.00010370673722848183,
|
| 1545 |
+
"loss": 0.0139,
|
| 1546 |
+
"step": 2150
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 3.2136084774121585,
|
| 1550 |
+
"grad_norm": 0.05486704409122467,
|
| 1551 |
+
"learning_rate": 0.00010222700246224735,
|
| 1552 |
+
"loss": 0.0135,
|
| 1553 |
+
"step": 2160
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 3.228481130321621,
|
| 1557 |
+
"grad_norm": 0.05656208097934723,
|
| 1558 |
+
"learning_rate": 0.00010075242371796585,
|
| 1559 |
+
"loss": 0.0125,
|
| 1560 |
+
"step": 2170
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 3.243353783231084,
|
| 1564 |
+
"grad_norm": 0.053801827132701874,
|
| 1565 |
+
"learning_rate": 9.928316014328916e-05,
|
| 1566 |
+
"loss": 0.0141,
|
| 1567 |
+
"step": 2180
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 3.2582264361405464,
|
| 1571 |
+
"grad_norm": 0.061040911823511124,
|
| 1572 |
+
"learning_rate": 9.781937031221589e-05,
|
| 1573 |
+
"loss": 0.0136,
|
| 1574 |
+
"step": 2190
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 3.2730990890500093,
|
| 1578 |
+
"grad_norm": 0.05558522418141365,
|
| 1579 |
+
"learning_rate": 9.636121220797707e-05,
|
| 1580 |
+
"loss": 0.0138,
|
| 1581 |
+
"step": 2200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 3.287971741959472,
|
| 1585 |
+
"grad_norm": 0.055547308176755905,
|
| 1586 |
+
"learning_rate": 9.490884320598516e-05,
|
| 1587 |
+
"loss": 0.0136,
|
| 1588 |
+
"step": 2210
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 3.3028443948689348,
|
| 1592 |
+
"grad_norm": 0.061592597514390945,
|
| 1593 |
+
"learning_rate": 9.34624200568492e-05,
|
| 1594 |
+
"loss": 0.014,
|
| 1595 |
+
"step": 2220
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 3.3177170477783973,
|
| 1599 |
+
"grad_norm": 0.05287894979119301,
|
| 1600 |
+
"learning_rate": 9.202209886945698e-05,
|
| 1601 |
+
"loss": 0.0125,
|
| 1602 |
+
"step": 2230
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 3.33258970068786,
|
| 1606 |
+
"grad_norm": 0.06365808844566345,
|
| 1607 |
+
"learning_rate": 9.058803509412646e-05,
|
| 1608 |
+
"loss": 0.0139,
|
| 1609 |
+
"step": 2240
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 3.347462353597323,
|
| 1613 |
+
"grad_norm": 0.05474059656262398,
|
| 1614 |
+
"learning_rate": 8.916038350582876e-05,
|
| 1615 |
+
"loss": 0.0141,
|
| 1616 |
+
"step": 2250
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 3.3623350065067856,
|
| 1620 |
+
"grad_norm": 0.054872963577508926,
|
| 1621 |
+
"learning_rate": 8.773929818748315e-05,
|
| 1622 |
+
"loss": 0.0135,
|
| 1623 |
+
"step": 2260
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 3.3772076594162486,
|
| 1627 |
+
"grad_norm": 0.05935963988304138,
|
| 1628 |
+
"learning_rate": 8.632493251332793e-05,
|
| 1629 |
+
"loss": 0.0128,
|
| 1630 |
+
"step": 2270
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 3.392080312325711,
|
| 1634 |
+
"grad_norm": 0.06830602139234543,
|
| 1635 |
+
"learning_rate": 8.491743913236628e-05,
|
| 1636 |
+
"loss": 0.0133,
|
| 1637 |
+
"step": 2280
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 3.406952965235174,
|
| 1641 |
+
"grad_norm": 0.057178862392902374,
|
| 1642 |
+
"learning_rate": 8.351696995189218e-05,
|
| 1643 |
+
"loss": 0.0121,
|
| 1644 |
+
"step": 2290
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 3.4218256181446365,
|
| 1648 |
+
"grad_norm": 0.06827449798583984,
|
| 1649 |
+
"learning_rate": 8.212367612109464e-05,
|
| 1650 |
+
"loss": 0.0127,
|
| 1651 |
+
"step": 2300
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 3.4366982710540994,
|
| 1655 |
+
"grad_norm": 0.04981634393334389,
|
| 1656 |
+
"learning_rate": 8.073770801474495e-05,
|
| 1657 |
+
"loss": 0.0132,
|
| 1658 |
+
"step": 2310
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 3.451570923963562,
|
| 1662 |
+
"grad_norm": 0.052124422043561935,
|
| 1663 |
+
"learning_rate": 7.935921521696702e-05,
|
| 1664 |
+
"loss": 0.0129,
|
| 1665 |
+
"step": 2320
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 3.466443576873025,
|
| 1669 |
+
"grad_norm": 0.05991722270846367,
|
| 1670 |
+
"learning_rate": 7.798834650509306e-05,
|
| 1671 |
+
"loss": 0.0128,
|
| 1672 |
+
"step": 2330
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 3.4813162297824873,
|
| 1676 |
+
"grad_norm": 0.05946414917707443,
|
| 1677 |
+
"learning_rate": 7.662524983360665e-05,
|
| 1678 |
+
"loss": 0.0127,
|
| 1679 |
+
"step": 2340
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 3.4961888826919503,
|
| 1683 |
+
"grad_norm": 0.05650801584124565,
|
| 1684 |
+
"learning_rate": 7.527007231817389e-05,
|
| 1685 |
+
"loss": 0.0127,
|
| 1686 |
+
"step": 2350
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 3.5110615356014128,
|
| 1690 |
+
"grad_norm": 0.04841410368680954,
|
| 1691 |
+
"learning_rate": 7.392296021976614e-05,
|
| 1692 |
+
"loss": 0.0122,
|
| 1693 |
+
"step": 2360
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 3.5259341885108757,
|
| 1697 |
+
"grad_norm": 0.05933946743607521,
|
| 1698 |
+
"learning_rate": 7.258405892887398e-05,
|
| 1699 |
+
"loss": 0.0121,
|
| 1700 |
+
"step": 2370
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 3.540806841420338,
|
| 1704 |
+
"grad_norm": 0.05451497435569763,
|
| 1705 |
+
"learning_rate": 7.125351294981598e-05,
|
| 1706 |
+
"loss": 0.0127,
|
| 1707 |
+
"step": 2380
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 3.555679494329801,
|
| 1711 |
+
"grad_norm": 0.05574881285429001,
|
| 1712 |
+
"learning_rate": 6.993146588514225e-05,
|
| 1713 |
+
"loss": 0.0124,
|
| 1714 |
+
"step": 2390
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 3.5705521472392636,
|
| 1718 |
+
"grad_norm": 0.057919006794691086,
|
| 1719 |
+
"learning_rate": 6.86180604201361e-05,
|
| 1720 |
+
"loss": 0.0119,
|
| 1721 |
+
"step": 2400
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 3.5854248001487266,
|
| 1725 |
+
"grad_norm": 0.051368821412324905,
|
| 1726 |
+
"learning_rate": 6.731343830741433e-05,
|
| 1727 |
+
"loss": 0.0126,
|
| 1728 |
+
"step": 2410
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 3.6002974530581895,
|
| 1732 |
+
"grad_norm": 0.06351654976606369,
|
| 1733 |
+
"learning_rate": 6.6017740351628e-05,
|
| 1734 |
+
"loss": 0.0135,
|
| 1735 |
+
"step": 2420
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 3.615170105967652,
|
| 1739 |
+
"grad_norm": 0.053709421306848526,
|
| 1740 |
+
"learning_rate": 6.473110639426616e-05,
|
| 1741 |
+
"loss": 0.0122,
|
| 1742 |
+
"step": 2430
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 3.6300427588771145,
|
| 1746 |
+
"grad_norm": 0.061445701867341995,
|
| 1747 |
+
"learning_rate": 6.345367529856254e-05,
|
| 1748 |
+
"loss": 0.0132,
|
| 1749 |
+
"step": 2440
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 3.6449154117865774,
|
| 1753 |
+
"grad_norm": 0.0678747370839119,
|
| 1754 |
+
"learning_rate": 6.218558493450893e-05,
|
| 1755 |
+
"loss": 0.0125,
|
| 1756 |
+
"step": 2450
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 3.6597880646960403,
|
| 1760 |
+
"grad_norm": 0.05095114931464195,
|
| 1761 |
+
"learning_rate": 6.0926972163974775e-05,
|
| 1762 |
+
"loss": 0.012,
|
| 1763 |
+
"step": 2460
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 3.674660717605503,
|
| 1767 |
+
"grad_norm": 0.05740583688020706,
|
| 1768 |
+
"learning_rate": 5.9677972825936254e-05,
|
| 1769 |
+
"loss": 0.0125,
|
| 1770 |
+
"step": 2470
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 3.6895333705149658,
|
| 1774 |
+
"grad_norm": 0.05399662256240845,
|
| 1775 |
+
"learning_rate": 5.8438721721815536e-05,
|
| 1776 |
+
"loss": 0.0134,
|
| 1777 |
+
"step": 2480
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 3.7044060234244283,
|
| 1781 |
+
"grad_norm": 0.056056030094623566,
|
| 1782 |
+
"learning_rate": 5.720935260093177e-05,
|
| 1783 |
+
"loss": 0.0125,
|
| 1784 |
+
"step": 2490
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 3.719278676333891,
|
| 1788 |
+
"grad_norm": 0.046866290271282196,
|
| 1789 |
+
"learning_rate": 5.598999814606618e-05,
|
| 1790 |
+
"loss": 0.0118,
|
| 1791 |
+
"step": 2500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 3.719278676333891,
|
| 1795 |
+
"eval_loss": 0.031009526923298836,
|
| 1796 |
+
"eval_runtime": 212.3923,
|
| 1797 |
+
"eval_samples_per_second": 22.51,
|
| 1798 |
+
"eval_steps_per_second": 22.51,
|
| 1799 |
+
"step": 2500
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 3.7341513292433537,
|
| 1803 |
+
"grad_norm": 0.046400755643844604,
|
| 1804 |
+
"learning_rate": 5.4780789959141524e-05,
|
| 1805 |
+
"loss": 0.0122,
|
| 1806 |
+
"step": 2510
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 3.7490239821528166,
|
| 1810 |
+
"grad_norm": 0.05211547762155533,
|
| 1811 |
+
"learning_rate": 5.358185854701909e-05,
|
| 1812 |
+
"loss": 0.0122,
|
| 1813 |
+
"step": 2520
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 3.763896635062279,
|
| 1817 |
+
"grad_norm": 0.0429752878844738,
|
| 1818 |
+
"learning_rate": 5.239333330741298e-05,
|
| 1819 |
+
"loss": 0.0124,
|
| 1820 |
+
"step": 2530
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 3.778769287971742,
|
| 1824 |
+
"grad_norm": 0.05008607730269432,
|
| 1825 |
+
"learning_rate": 5.121534251492486e-05,
|
| 1826 |
+
"loss": 0.0125,
|
| 1827 |
+
"step": 2540
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 3.7936419408812045,
|
| 1831 |
+
"grad_norm": 0.046397943049669266,
|
| 1832 |
+
"learning_rate": 5.004801330719941e-05,
|
| 1833 |
+
"loss": 0.0111,
|
| 1834 |
+
"step": 2550
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 3.8085145937906675,
|
| 1838 |
+
"grad_norm": 0.05960022658109665,
|
| 1839 |
+
"learning_rate": 4.8891471671202675e-05,
|
| 1840 |
+
"loss": 0.0117,
|
| 1841 |
+
"step": 2560
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 3.82338724670013,
|
| 1845 |
+
"grad_norm": 0.04353282228112221,
|
| 1846 |
+
"learning_rate": 4.7745842429624795e-05,
|
| 1847 |
+
"loss": 0.0119,
|
| 1848 |
+
"step": 2570
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 3.838259899609593,
|
| 1852 |
+
"grad_norm": 0.05057670921087265,
|
| 1853 |
+
"learning_rate": 4.661124922740794e-05,
|
| 1854 |
+
"loss": 0.0116,
|
| 1855 |
+
"step": 2580
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 3.853132552519056,
|
| 1859 |
+
"grad_norm": 0.04886782541871071,
|
| 1860 |
+
"learning_rate": 4.548781451840179e-05,
|
| 1861 |
+
"loss": 0.0113,
|
| 1862 |
+
"step": 2590
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 3.8680052054285183,
|
| 1866 |
+
"grad_norm": 0.055182382464408875,
|
| 1867 |
+
"learning_rate": 4.437565955214723e-05,
|
| 1868 |
+
"loss": 0.0116,
|
| 1869 |
+
"step": 2600
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 3.882877858337981,
|
| 1873 |
+
"grad_norm": 0.048834457993507385,
|
| 1874 |
+
"learning_rate": 4.3274904360790505e-05,
|
| 1875 |
+
"loss": 0.0121,
|
| 1876 |
+
"step": 2610
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 3.8977505112474438,
|
| 1880 |
+
"grad_norm": 0.05025951564311981,
|
| 1881 |
+
"learning_rate": 4.218566774612802e-05,
|
| 1882 |
+
"loss": 0.0112,
|
| 1883 |
+
"step": 2620
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 3.9126231641569067,
|
| 1887 |
+
"grad_norm": 0.05054251477122307,
|
| 1888 |
+
"learning_rate": 4.1108067266784746e-05,
|
| 1889 |
+
"loss": 0.0112,
|
| 1890 |
+
"step": 2630
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 3.927495817066369,
|
| 1894 |
+
"grad_norm": 0.05326022952795029,
|
| 1895 |
+
"learning_rate": 4.004221922552608e-05,
|
| 1896 |
+
"loss": 0.0119,
|
| 1897 |
+
"step": 2640
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 3.9423684699758317,
|
| 1901 |
+
"grad_norm": 0.05668502673506737,
|
| 1902 |
+
"learning_rate": 3.898823865670579e-05,
|
| 1903 |
+
"loss": 0.0114,
|
| 1904 |
+
"step": 2650
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 3.9572411228852946,
|
| 1908 |
+
"grad_norm": 0.054235439747571945,
|
| 1909 |
+
"learning_rate": 3.794623931385062e-05,
|
| 1910 |
+
"loss": 0.0119,
|
| 1911 |
+
"step": 2660
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 3.9721137757947576,
|
| 1915 |
+
"grad_norm": 0.05231969431042671,
|
| 1916 |
+
"learning_rate": 3.6916333657383024e-05,
|
| 1917 |
+
"loss": 0.0108,
|
| 1918 |
+
"step": 2670
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 3.98698642870422,
|
| 1922 |
+
"grad_norm": 0.057500049471855164,
|
| 1923 |
+
"learning_rate": 3.5898632842483746e-05,
|
| 1924 |
+
"loss": 0.011,
|
| 1925 |
+
"step": 2680
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 4.002230897936419,
|
| 1929 |
+
"grad_norm": 0.04203633964061737,
|
| 1930 |
+
"learning_rate": 3.489324670709494e-05,
|
| 1931 |
+
"loss": 0.0113,
|
| 1932 |
+
"step": 2690
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 4.017103550845882,
|
| 1936 |
+
"grad_norm": 0.029648838564753532,
|
| 1937 |
+
"learning_rate": 3.390028376006589e-05,
|
| 1938 |
+
"loss": 0.0059,
|
| 1939 |
+
"step": 2700
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 4.031976203755345,
|
| 1943 |
+
"grad_norm": 0.03779765963554382,
|
| 1944 |
+
"learning_rate": 3.2919851169441625e-05,
|
| 1945 |
+
"loss": 0.006,
|
| 1946 |
+
"step": 2710
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 4.046848856664807,
|
| 1950 |
+
"grad_norm": 0.040116600692272186,
|
| 1951 |
+
"learning_rate": 3.195205475089667e-05,
|
| 1952 |
+
"loss": 0.0058,
|
| 1953 |
+
"step": 2720
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 4.06172150957427,
|
| 1957 |
+
"grad_norm": 0.030058899894356728,
|
| 1958 |
+
"learning_rate": 3.099699895631474e-05,
|
| 1959 |
+
"loss": 0.0056,
|
| 1960 |
+
"step": 2730
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 4.076594162483733,
|
| 1964 |
+
"grad_norm": 0.03675166517496109,
|
| 1965 |
+
"learning_rate": 3.0054786862515257e-05,
|
| 1966 |
+
"loss": 0.0058,
|
| 1967 |
+
"step": 2740
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 4.091466815393196,
|
| 1971 |
+
"grad_norm": 0.03470413759350777,
|
| 1972 |
+
"learning_rate": 2.912552016012879e-05,
|
| 1973 |
+
"loss": 0.0057,
|
| 1974 |
+
"step": 2750
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 4.106339468302658,
|
| 1978 |
+
"grad_norm": 0.03222460299730301,
|
| 1979 |
+
"learning_rate": 2.8209299142621522e-05,
|
| 1980 |
+
"loss": 0.0057,
|
| 1981 |
+
"step": 2760
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 4.121212121212121,
|
| 1985 |
+
"grad_norm": 0.036458127200603485,
|
| 1986 |
+
"learning_rate": 2.7306222695471173e-05,
|
| 1987 |
+
"loss": 0.0056,
|
| 1988 |
+
"step": 2770
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 4.136084774121584,
|
| 1992 |
+
"grad_norm": 0.035760316997766495,
|
| 1993 |
+
"learning_rate": 2.641638828549425e-05,
|
| 1994 |
+
"loss": 0.0055,
|
| 1995 |
+
"step": 2780
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 4.150957427031047,
|
| 1999 |
+
"grad_norm": 0.04281270503997803,
|
| 2000 |
+
"learning_rate": 2.5539891950326875e-05,
|
| 2001 |
+
"loss": 0.0056,
|
| 2002 |
+
"step": 2790
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 4.165830079940509,
|
| 2006 |
+
"grad_norm": 0.030339548364281654,
|
| 2007 |
+
"learning_rate": 2.4676828288059558e-05,
|
| 2008 |
+
"loss": 0.0057,
|
| 2009 |
+
"step": 2800
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 4.180702732849972,
|
| 2013 |
+
"grad_norm": 0.03753247857093811,
|
| 2014 |
+
"learning_rate": 2.382729044702748e-05,
|
| 2015 |
+
"loss": 0.0058,
|
| 2016 |
+
"step": 2810
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 4.195575385759435,
|
| 2020 |
+
"grad_norm": 0.035988811403512955,
|
| 2021 |
+
"learning_rate": 2.299137011575738e-05,
|
| 2022 |
+
"loss": 0.0055,
|
| 2023 |
+
"step": 2820
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 4.210448038668898,
|
| 2027 |
+
"grad_norm": 0.0344134196639061,
|
| 2028 |
+
"learning_rate": 2.2169157513071566e-05,
|
| 2029 |
+
"loss": 0.0057,
|
| 2030 |
+
"step": 2830
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 4.22532069157836,
|
| 2034 |
+
"grad_norm": 0.03696177527308464,
|
| 2035 |
+
"learning_rate": 2.136074137835107e-05,
|
| 2036 |
+
"loss": 0.0056,
|
| 2037 |
+
"step": 2840
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 4.240193344487823,
|
| 2041 |
+
"grad_norm": 0.03733756020665169,
|
| 2042 |
+
"learning_rate": 2.056620896195804e-05,
|
| 2043 |
+
"loss": 0.0057,
|
| 2044 |
+
"step": 2850
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 4.255065997397286,
|
| 2048 |
+
"grad_norm": 0.03630942478775978,
|
| 2049 |
+
"learning_rate": 1.978564601581919e-05,
|
| 2050 |
+
"loss": 0.0056,
|
| 2051 |
+
"step": 2860
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 4.269938650306749,
|
| 2055 |
+
"grad_norm": 0.03577449545264244,
|
| 2056 |
+
"learning_rate": 1.9019136784170635e-05,
|
| 2057 |
+
"loss": 0.0055,
|
| 2058 |
+
"step": 2870
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 4.284811303216211,
|
| 2062 |
+
"grad_norm": 0.03209745138883591,
|
| 2063 |
+
"learning_rate": 1.82667639944657e-05,
|
| 2064 |
+
"loss": 0.0054,
|
| 2065 |
+
"step": 2880
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 4.299683956125674,
|
| 2069 |
+
"grad_norm": 0.03668665885925293,
|
| 2070 |
+
"learning_rate": 1.752860884844646e-05,
|
| 2071 |
+
"loss": 0.0055,
|
| 2072 |
+
"step": 2890
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 4.314556609035137,
|
| 2076 |
+
"grad_norm": 0.03498975560069084,
|
| 2077 |
+
"learning_rate": 1.680475101337959e-05,
|
| 2078 |
+
"loss": 0.0055,
|
| 2079 |
+
"step": 2900
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 4.3294292619445995,
|
| 2083 |
+
"grad_norm": 0.04088146984577179,
|
| 2084 |
+
"learning_rate": 1.60952686134583e-05,
|
| 2085 |
+
"loss": 0.0055,
|
| 2086 |
+
"step": 2910
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 4.3443019148540625,
|
| 2090 |
+
"grad_norm": 0.035557616502046585,
|
| 2091 |
+
"learning_rate": 1.5400238221370413e-05,
|
| 2092 |
+
"loss": 0.0056,
|
| 2093 |
+
"step": 2920
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 4.3591745677635245,
|
| 2097 |
+
"grad_norm": 0.03443196415901184,
|
| 2098 |
+
"learning_rate": 1.4719734850034277e-05,
|
| 2099 |
+
"loss": 0.0056,
|
| 2100 |
+
"step": 2930
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 4.3740472206729875,
|
| 2104 |
+
"grad_norm": 0.03481742739677429,
|
| 2105 |
+
"learning_rate": 1.4053831944502508e-05,
|
| 2106 |
+
"loss": 0.0057,
|
| 2107 |
+
"step": 2940
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 4.38891987358245,
|
| 2111 |
+
"grad_norm": 0.03648516163229942,
|
| 2112 |
+
"learning_rate": 1.340260137403557e-05,
|
| 2113 |
+
"loss": 0.0053,
|
| 2114 |
+
"step": 2950
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 4.403792526491913,
|
| 2118 |
+
"grad_norm": 0.03400832787156105,
|
| 2119 |
+
"learning_rate": 1.2766113424344814e-05,
|
| 2120 |
+
"loss": 0.0055,
|
| 2121 |
+
"step": 2960
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 4.418665179401375,
|
| 2125 |
+
"grad_norm": 0.03558880090713501,
|
| 2126 |
+
"learning_rate": 1.21444367900069e-05,
|
| 2127 |
+
"loss": 0.0055,
|
| 2128 |
+
"step": 2970
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 4.433537832310838,
|
| 2132 |
+
"grad_norm": 0.035319775342941284,
|
| 2133 |
+
"learning_rate": 1.1537638567049729e-05,
|
| 2134 |
+
"loss": 0.0055,
|
| 2135 |
+
"step": 2980
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 4.448410485220301,
|
| 2139 |
+
"grad_norm": 0.03432595729827881,
|
| 2140 |
+
"learning_rate": 1.0945784245710848e-05,
|
| 2141 |
+
"loss": 0.0054,
|
| 2142 |
+
"step": 2990
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 4.463283138129764,
|
| 2146 |
+
"grad_norm": 0.03571225702762604,
|
| 2147 |
+
"learning_rate": 1.036893770336938e-05,
|
| 2148 |
+
"loss": 0.0055,
|
| 2149 |
+
"step": 3000
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 4.463283138129764,
|
| 2153 |
+
"eval_loss": 0.03200867399573326,
|
| 2154 |
+
"eval_runtime": 212.5457,
|
| 2155 |
+
"eval_samples_per_second": 22.494,
|
| 2156 |
+
"eval_steps_per_second": 22.494,
|
| 2157 |
+
"step": 3000
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 4.478155791039226,
|
| 2161 |
+
"grad_norm": 0.040391724556684494,
|
| 2162 |
+
"learning_rate": 9.807161197651742e-06,
|
| 2163 |
+
"loss": 0.0056,
|
| 2164 |
+
"step": 3010
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 4.493028443948689,
|
| 2168 |
+
"grad_norm": 0.03410281240940094,
|
| 2169 |
+
"learning_rate": 9.260515359712517e-06,
|
| 2170 |
+
"loss": 0.0055,
|
| 2171 |
+
"step": 3020
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 4.507901096858152,
|
| 2175 |
+
"grad_norm": 0.03447275608778,
|
| 2176 |
+
"learning_rate": 8.729059187690479e-06,
|
| 2177 |
+
"loss": 0.0054,
|
| 2178 |
+
"step": 3030
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 4.522773749767615,
|
| 2182 |
+
"grad_norm": 0.032652657479047775,
|
| 2183 |
+
"learning_rate": 8.212850040341273e-06,
|
| 2184 |
+
"loss": 0.0055,
|
| 2185 |
+
"step": 3040
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 4.537646402677078,
|
| 2189 |
+
"grad_norm": 0.035828616470098495,
|
| 2190 |
+
"learning_rate": 7.711943630846684e-06,
|
| 2191 |
+
"loss": 0.0053,
|
| 2192 |
+
"step": 3050
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 4.55251905558654,
|
| 2196 |
+
"grad_norm": 0.03351854532957077,
|
| 2197 |
+
"learning_rate": 7.226394020801645e-06,
|
| 2198 |
+
"loss": 0.0054,
|
| 2199 |
+
"step": 3060
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 4.567391708496003,
|
| 2203 |
+
"grad_norm": 0.03872072696685791,
|
| 2204 |
+
"learning_rate": 6.7562536143796254e-06,
|
| 2205 |
+
"loss": 0.0056,
|
| 2206 |
+
"step": 3070
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 4.582264361405466,
|
| 2210 |
+
"grad_norm": 0.03518550843000412,
|
| 2211 |
+
"learning_rate": 6.301573152676664e-06,
|
| 2212 |
+
"loss": 0.0054,
|
| 2213 |
+
"step": 3080
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 4.597137014314928,
|
| 2217 |
+
"grad_norm": 0.0351685993373394,
|
| 2218 |
+
"learning_rate": 5.862401708235076e-06,
|
| 2219 |
+
"loss": 0.0052,
|
| 2220 |
+
"step": 3090
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 4.612009667224391,
|
| 2224 |
+
"grad_norm": 0.0348668210208416,
|
| 2225 |
+
"learning_rate": 5.438786679747081e-06,
|
| 2226 |
+
"loss": 0.0055,
|
| 2227 |
+
"step": 3100
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 4.626882320133854,
|
| 2231 |
+
"grad_norm": 0.03660331293940544,
|
| 2232 |
+
"learning_rate": 5.030773786939319e-06,
|
| 2233 |
+
"loss": 0.0055,
|
| 2234 |
+
"step": 3110
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 4.641754973043317,
|
| 2238 |
+
"grad_norm": 0.04046601429581642,
|
| 2239 |
+
"learning_rate": 4.638407065638322e-06,
|
| 2240 |
+
"loss": 0.0054,
|
| 2241 |
+
"step": 3120
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 4.65662762595278,
|
| 2245 |
+
"grad_norm": 0.03230154886841774,
|
| 2246 |
+
"learning_rate": 4.261728863017827e-06,
|
| 2247 |
+
"loss": 0.0054,
|
| 2248 |
+
"step": 3130
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 4.671500278862242,
|
| 2252 |
+
"grad_norm": 0.034297142177820206,
|
| 2253 |
+
"learning_rate": 3.900779833028472e-06,
|
| 2254 |
+
"loss": 0.0054,
|
| 2255 |
+
"step": 3140
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 4.686372931771705,
|
| 2259 |
+
"grad_norm": 0.03240946680307388,
|
| 2260 |
+
"learning_rate": 3.5555989320099952e-06,
|
| 2261 |
+
"loss": 0.0053,
|
| 2262 |
+
"step": 3150
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 4.701245584681168,
|
| 2266 |
+
"grad_norm": 0.04137023165822029,
|
| 2267 |
+
"learning_rate": 3.2262234144868116e-06,
|
| 2268 |
+
"loss": 0.0054,
|
| 2269 |
+
"step": 3160
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 4.7161182375906305,
|
| 2273 |
+
"grad_norm": 0.030783316120505333,
|
| 2274 |
+
"learning_rate": 2.912688829147214e-06,
|
| 2275 |
+
"loss": 0.0052,
|
| 2276 |
+
"step": 3170
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 4.730990890500093,
|
| 2280 |
+
"grad_norm": 0.03588159382343292,
|
| 2281 |
+
"learning_rate": 2.6150290150067588e-06,
|
| 2282 |
+
"loss": 0.0055,
|
| 2283 |
+
"step": 3180
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 4.7458635434095555,
|
| 2287 |
+
"grad_norm": 0.03300805762410164,
|
| 2288 |
+
"learning_rate": 2.3332760977559873e-06,
|
| 2289 |
+
"loss": 0.0053,
|
| 2290 |
+
"step": 3190
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 4.7607361963190185,
|
| 2294 |
+
"grad_norm": 0.03986676409840584,
|
| 2295 |
+
"learning_rate": 2.0674604862932654e-06,
|
| 2296 |
+
"loss": 0.0055,
|
| 2297 |
+
"step": 3200
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 4.775608849228481,
|
| 2301 |
+
"grad_norm": 0.03252493590116501,
|
| 2302 |
+
"learning_rate": 1.8176108694427927e-06,
|
| 2303 |
+
"loss": 0.0052,
|
| 2304 |
+
"step": 3210
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 4.790481502137943,
|
| 2308 |
+
"grad_norm": 0.03938417136669159,
|
| 2309 |
+
"learning_rate": 1.583754212858329e-06,
|
| 2310 |
+
"loss": 0.0054,
|
| 2311 |
+
"step": 3220
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 4.805354155047406,
|
| 2315 |
+
"grad_norm": 0.03552339971065521,
|
| 2316 |
+
"learning_rate": 1.3659157561127732e-06,
|
| 2317 |
+
"loss": 0.0057,
|
| 2318 |
+
"step": 3230
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 4.820226807956869,
|
| 2322 |
+
"grad_norm": 0.03480495885014534,
|
| 2323 |
+
"learning_rate": 1.1641190099741904e-06,
|
| 2324 |
+
"loss": 0.0053,
|
| 2325 |
+
"step": 3240
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 4.835099460866332,
|
| 2329 |
+
"grad_norm": 0.03451026231050491,
|
| 2330 |
+
"learning_rate": 9.783857538683603e-07,
|
| 2331 |
+
"loss": 0.0053,
|
| 2332 |
+
"step": 3250
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 4.849972113775795,
|
| 2336 |
+
"grad_norm": 0.033308371901512146,
|
| 2337 |
+
"learning_rate": 8.087360335281235e-07,
|
| 2338 |
+
"loss": 0.0055,
|
| 2339 |
+
"step": 3260
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 4.864844766685257,
|
| 2343 |
+
"grad_norm": 0.035610370337963104,
|
| 2344 |
+
"learning_rate": 6.551881588299279e-07,
|
| 2345 |
+
"loss": 0.0054,
|
| 2346 |
+
"step": 3270
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 4.87971741959472,
|
| 2350 |
+
"grad_norm": 0.030910024419426918,
|
| 2351 |
+
"learning_rate": 5.177587018176777e-07,
|
| 2352 |
+
"loss": 0.0054,
|
| 2353 |
+
"step": 3280
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 4.894590072504183,
|
| 2357 |
+
"grad_norm": 0.034942276775836945,
|
| 2358 |
+
"learning_rate": 3.964624949141626e-07,
|
| 2359 |
+
"loss": 0.0054,
|
| 2360 |
+
"step": 3290
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 4.909462725413646,
|
| 2364 |
+
"grad_norm": 0.03491232544183731,
|
| 2365 |
+
"learning_rate": 2.913126293202228e-07,
|
| 2366 |
+
"loss": 0.0053,
|
| 2367 |
+
"step": 3300
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 4.924335378323108,
|
| 2371 |
+
"grad_norm": 0.0331818163394928,
|
| 2372 |
+
"learning_rate": 2.0232045360184523e-07,
|
| 2373 |
+
"loss": 0.0051,
|
| 2374 |
+
"step": 3310
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 4.939208031232571,
|
| 2378 |
+
"grad_norm": 0.034393060952425,
|
| 2379 |
+
"learning_rate": 1.2949557246537678e-07,
|
| 2380 |
+
"loss": 0.0053,
|
| 2381 |
+
"step": 3320
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 4.954080684142034,
|
| 2385 |
+
"grad_norm": 0.03940508887171745,
|
| 2386 |
+
"learning_rate": 7.284584572085361e-08,
|
| 2387 |
+
"loss": 0.0052,
|
| 2388 |
+
"step": 3330
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 4.968953337051497,
|
| 2392 |
+
"grad_norm": 0.03125544637441635,
|
| 2393 |
+
"learning_rate": 3.237738743372964e-08,
|
| 2394 |
+
"loss": 0.0052,
|
| 2395 |
+
"step": 3340
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 4.983825989960959,
|
| 2399 |
+
"grad_norm": 0.03558258339762688,
|
| 2400 |
+
"learning_rate": 8.094565265054365e-09,
|
| 2401 |
+
"loss": 0.0054,
|
| 2402 |
+
"step": 3350
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 4.998698642870422,
|
| 2406 |
+
"grad_norm": 0.03360743075609207,
|
| 2407 |
+
"learning_rate": 0.0,
|
| 2408 |
+
"loss": 0.0054,
|
| 2409 |
+
"step": 3360
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 4.998698642870422,
|
| 2413 |
+
"step": 3360,
|
| 2414 |
+
"total_flos": 5.14290499398402e+18,
|
| 2415 |
+
"train_loss": 0.19145491501161208,
|
| 2416 |
+
"train_runtime": 31931.328,
|
| 2417 |
+
"train_samples_per_second": 6.737,
|
| 2418 |
+
"train_steps_per_second": 0.105
|
| 2419 |
+
}
|
| 2420 |
+
],
|
| 2421 |
+
"logging_steps": 10,
|
| 2422 |
+
"max_steps": 3360,
|
| 2423 |
+
"num_input_tokens_seen": 0,
|
| 2424 |
+
"num_train_epochs": 5,
|
| 2425 |
+
"save_steps": 500,
|
| 2426 |
+
"stateful_callbacks": {
|
| 2427 |
+
"TrainerControl": {
|
| 2428 |
+
"args": {
|
| 2429 |
+
"should_epoch_stop": false,
|
| 2430 |
+
"should_evaluate": false,
|
| 2431 |
+
"should_log": false,
|
| 2432 |
+
"should_save": true,
|
| 2433 |
+
"should_training_stop": true
|
| 2434 |
+
},
|
| 2435 |
+
"attributes": {}
|
| 2436 |
+
}
|
| 2437 |
+
},
|
| 2438 |
+
"total_flos": 5.14290499398402e+18,
|
| 2439 |
+
"train_batch_size": 8,
|
| 2440 |
+
"trial_name": null,
|
| 2441 |
+
"trial_params": null
|
| 2442 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|