End of training
Browse files- README.md +3 -1
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +4262 -0
README.md
CHANGED
|
@@ -3,6 +3,8 @@ license: apache-2.0
|
|
| 3 |
base_model: google/long-t5-tglobal-xl
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
model-index:
|
| 7 |
- name: longt5_xl_sfd_memsum_40
|
| 8 |
results: []
|
|
@@ -13,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# longt5_xl_sfd_memsum_40
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [google/long-t5-tglobal-xl](https://huggingface.co/google/long-t5-tglobal-xl) on
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
- Loss: 5.2679
|
| 19 |
|
|
|
|
| 3 |
base_model: google/long-t5-tglobal-xl
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- learn3r/summ_screen_memsum_oracle
|
| 8 |
model-index:
|
| 9 |
- name: longt5_xl_sfd_memsum_40
|
| 10 |
results: []
|
|
|
|
| 15 |
|
| 16 |
# longt5_xl_sfd_memsum_40
|
| 17 |
|
| 18 |
+
This model is a fine-tuned version of [google/long-t5-tglobal-xl](https://huggingface.co/google/long-t5-tglobal-xl) on the learn3r/summ_screen_memsum_oracle dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 5.2679
|
| 21 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 38.96,
|
| 3 |
+
"eval_loss": 5.267875671386719,
|
| 4 |
+
"eval_runtime": 14.0885,
|
| 5 |
+
"eval_samples": 338,
|
| 6 |
+
"eval_samples_per_second": 23.991,
|
| 7 |
+
"eval_steps_per_second": 3.052,
|
| 8 |
+
"train_loss": 0.32971865670822026,
|
| 9 |
+
"train_runtime": 25027.7089,
|
| 10 |
+
"train_samples": 3673,
|
| 11 |
+
"train_samples_per_second": 5.87,
|
| 12 |
+
"train_steps_per_second": 0.045
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 38.96,
|
| 3 |
+
"eval_loss": 5.267875671386719,
|
| 4 |
+
"eval_runtime": 14.0885,
|
| 5 |
+
"eval_samples": 338,
|
| 6 |
+
"eval_samples_per_second": 23.991,
|
| 7 |
+
"eval_steps_per_second": 3.052
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 38.96,
|
| 3 |
+
"train_loss": 0.32971865670822026,
|
| 4 |
+
"train_runtime": 25027.7089,
|
| 5 |
+
"train_samples": 3673,
|
| 6 |
+
"train_samples_per_second": 5.87,
|
| 7 |
+
"train_steps_per_second": 0.045
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,4262 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 2.2837178707122803,
|
| 3 |
+
"best_model_checkpoint": "/exports/eddie/scratch/s1970716/models/longt5_xl_sfd_memsum_40/checkpoint-57",
|
| 4 |
+
"epoch": 38.95652173913044,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1120,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.07,
|
| 13 |
+
"grad_norm": 1.5414308309555054,
|
| 14 |
+
"learning_rate": 0.001,
|
| 15 |
+
"loss": 3.1508,
|
| 16 |
+
"step": 2
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.14,
|
| 20 |
+
"grad_norm": 1.1113317012786865,
|
| 21 |
+
"learning_rate": 0.001,
|
| 22 |
+
"loss": 3.2124,
|
| 23 |
+
"step": 4
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.21,
|
| 27 |
+
"grad_norm": 0.7978044152259827,
|
| 28 |
+
"learning_rate": 0.001,
|
| 29 |
+
"loss": 2.9466,
|
| 30 |
+
"step": 6
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.28,
|
| 34 |
+
"grad_norm": 0.6611266136169434,
|
| 35 |
+
"learning_rate": 0.001,
|
| 36 |
+
"loss": 2.8914,
|
| 37 |
+
"step": 8
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.35,
|
| 41 |
+
"grad_norm": 0.6737671494483948,
|
| 42 |
+
"learning_rate": 0.001,
|
| 43 |
+
"loss": 2.7507,
|
| 44 |
+
"step": 10
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.42,
|
| 48 |
+
"grad_norm": 0.41579362750053406,
|
| 49 |
+
"learning_rate": 0.001,
|
| 50 |
+
"loss": 2.6653,
|
| 51 |
+
"step": 12
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.49,
|
| 55 |
+
"grad_norm": 0.37621861696243286,
|
| 56 |
+
"learning_rate": 0.001,
|
| 57 |
+
"loss": 2.579,
|
| 58 |
+
"step": 14
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.56,
|
| 62 |
+
"grad_norm": 0.581123411655426,
|
| 63 |
+
"learning_rate": 0.001,
|
| 64 |
+
"loss": 2.631,
|
| 65 |
+
"step": 16
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.63,
|
| 69 |
+
"grad_norm": 0.6653873920440674,
|
| 70 |
+
"learning_rate": 0.001,
|
| 71 |
+
"loss": 2.6978,
|
| 72 |
+
"step": 18
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.7,
|
| 76 |
+
"grad_norm": 0.45471808314323425,
|
| 77 |
+
"learning_rate": 0.001,
|
| 78 |
+
"loss": 2.6541,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.77,
|
| 83 |
+
"grad_norm": 0.38716837763786316,
|
| 84 |
+
"learning_rate": 0.001,
|
| 85 |
+
"loss": 2.5325,
|
| 86 |
+
"step": 22
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.83,
|
| 90 |
+
"grad_norm": 0.40614280104637146,
|
| 91 |
+
"learning_rate": 0.001,
|
| 92 |
+
"loss": 2.5584,
|
| 93 |
+
"step": 24
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.9,
|
| 97 |
+
"grad_norm": 0.3157476484775543,
|
| 98 |
+
"learning_rate": 0.001,
|
| 99 |
+
"loss": 2.5054,
|
| 100 |
+
"step": 26
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.97,
|
| 104 |
+
"grad_norm": 0.3763532042503357,
|
| 105 |
+
"learning_rate": 0.001,
|
| 106 |
+
"loss": 2.5238,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.97,
|
| 111 |
+
"eval_loss": 2.314685106277466,
|
| 112 |
+
"eval_runtime": 14.2455,
|
| 113 |
+
"eval_samples_per_second": 23.727,
|
| 114 |
+
"eval_steps_per_second": 3.018,
|
| 115 |
+
"step": 28
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.04,
|
| 119 |
+
"grad_norm": 0.3202365040779114,
|
| 120 |
+
"learning_rate": 0.001,
|
| 121 |
+
"loss": 2.3018,
|
| 122 |
+
"step": 30
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.11,
|
| 126 |
+
"grad_norm": 0.4395117461681366,
|
| 127 |
+
"learning_rate": 0.001,
|
| 128 |
+
"loss": 2.1284,
|
| 129 |
+
"step": 32
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.18,
|
| 133 |
+
"grad_norm": 0.47938525676727295,
|
| 134 |
+
"learning_rate": 0.001,
|
| 135 |
+
"loss": 2.1782,
|
| 136 |
+
"step": 34
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.25,
|
| 140 |
+
"grad_norm": 1.0489932298660278,
|
| 141 |
+
"learning_rate": 0.001,
|
| 142 |
+
"loss": 2.1666,
|
| 143 |
+
"step": 36
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.32,
|
| 147 |
+
"grad_norm": 0.34978240728378296,
|
| 148 |
+
"learning_rate": 0.001,
|
| 149 |
+
"loss": 2.1598,
|
| 150 |
+
"step": 38
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.39,
|
| 154 |
+
"grad_norm": 0.28448227047920227,
|
| 155 |
+
"learning_rate": 0.001,
|
| 156 |
+
"loss": 2.023,
|
| 157 |
+
"step": 40
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.46,
|
| 161 |
+
"grad_norm": 0.3449482023715973,
|
| 162 |
+
"learning_rate": 0.001,
|
| 163 |
+
"loss": 2.0545,
|
| 164 |
+
"step": 42
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.53,
|
| 168 |
+
"grad_norm": 0.6303105354309082,
|
| 169 |
+
"learning_rate": 0.001,
|
| 170 |
+
"loss": 2.1264,
|
| 171 |
+
"step": 44
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.6,
|
| 175 |
+
"grad_norm": 0.3750413656234741,
|
| 176 |
+
"learning_rate": 0.001,
|
| 177 |
+
"loss": 2.1176,
|
| 178 |
+
"step": 46
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.67,
|
| 182 |
+
"grad_norm": 0.44775936007499695,
|
| 183 |
+
"learning_rate": 0.001,
|
| 184 |
+
"loss": 2.1624,
|
| 185 |
+
"step": 48
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.74,
|
| 189 |
+
"grad_norm": 0.34849441051483154,
|
| 190 |
+
"learning_rate": 0.001,
|
| 191 |
+
"loss": 2.0809,
|
| 192 |
+
"step": 50
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.81,
|
| 196 |
+
"grad_norm": 0.28750795125961304,
|
| 197 |
+
"learning_rate": 0.001,
|
| 198 |
+
"loss": 2.0913,
|
| 199 |
+
"step": 52
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.88,
|
| 203 |
+
"grad_norm": 0.2710261940956116,
|
| 204 |
+
"learning_rate": 0.001,
|
| 205 |
+
"loss": 2.0855,
|
| 206 |
+
"step": 54
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.95,
|
| 210 |
+
"grad_norm": 0.28888803720474243,
|
| 211 |
+
"learning_rate": 0.001,
|
| 212 |
+
"loss": 2.1298,
|
| 213 |
+
"step": 56
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.98,
|
| 217 |
+
"eval_loss": 2.2837178707122803,
|
| 218 |
+
"eval_runtime": 14.2329,
|
| 219 |
+
"eval_samples_per_second": 23.748,
|
| 220 |
+
"eval_steps_per_second": 3.021,
|
| 221 |
+
"step": 57
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.02,
|
| 225 |
+
"grad_norm": 0.2953917682170868,
|
| 226 |
+
"learning_rate": 0.001,
|
| 227 |
+
"loss": 1.9808,
|
| 228 |
+
"step": 58
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.09,
|
| 232 |
+
"grad_norm": 2.0295698642730713,
|
| 233 |
+
"learning_rate": 0.001,
|
| 234 |
+
"loss": 1.7208,
|
| 235 |
+
"step": 60
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.16,
|
| 239 |
+
"grad_norm": 0.4198295772075653,
|
| 240 |
+
"learning_rate": 0.001,
|
| 241 |
+
"loss": 1.7108,
|
| 242 |
+
"step": 62
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.23,
|
| 246 |
+
"grad_norm": 0.3806304335594177,
|
| 247 |
+
"learning_rate": 0.001,
|
| 248 |
+
"loss": 1.677,
|
| 249 |
+
"step": 64
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.3,
|
| 253 |
+
"grad_norm": 0.36914414167404175,
|
| 254 |
+
"learning_rate": 0.001,
|
| 255 |
+
"loss": 1.6717,
|
| 256 |
+
"step": 66
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.37,
|
| 260 |
+
"grad_norm": 0.34834641218185425,
|
| 261 |
+
"learning_rate": 0.001,
|
| 262 |
+
"loss": 1.6516,
|
| 263 |
+
"step": 68
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.43,
|
| 267 |
+
"grad_norm": 0.3333311080932617,
|
| 268 |
+
"learning_rate": 0.001,
|
| 269 |
+
"loss": 1.6677,
|
| 270 |
+
"step": 70
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.5,
|
| 274 |
+
"grad_norm": 0.41852736473083496,
|
| 275 |
+
"learning_rate": 0.001,
|
| 276 |
+
"loss": 1.6135,
|
| 277 |
+
"step": 72
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.57,
|
| 281 |
+
"grad_norm": 0.40690872073173523,
|
| 282 |
+
"learning_rate": 0.001,
|
| 283 |
+
"loss": 1.6447,
|
| 284 |
+
"step": 74
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.64,
|
| 288 |
+
"grad_norm": 0.4816026985645294,
|
| 289 |
+
"learning_rate": 0.001,
|
| 290 |
+
"loss": 1.7096,
|
| 291 |
+
"step": 76
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.71,
|
| 295 |
+
"grad_norm": 0.8278970122337341,
|
| 296 |
+
"learning_rate": 0.001,
|
| 297 |
+
"loss": 1.6721,
|
| 298 |
+
"step": 78
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.78,
|
| 302 |
+
"grad_norm": 0.27179405093193054,
|
| 303 |
+
"learning_rate": 0.001,
|
| 304 |
+
"loss": 1.6351,
|
| 305 |
+
"step": 80
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.85,
|
| 309 |
+
"grad_norm": 0.2935534417629242,
|
| 310 |
+
"learning_rate": 0.001,
|
| 311 |
+
"loss": 1.6402,
|
| 312 |
+
"step": 82
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.92,
|
| 316 |
+
"grad_norm": 0.3557533919811249,
|
| 317 |
+
"learning_rate": 0.001,
|
| 318 |
+
"loss": 1.6755,
|
| 319 |
+
"step": 84
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.99,
|
| 323 |
+
"grad_norm": 0.3663090169429779,
|
| 324 |
+
"learning_rate": 0.001,
|
| 325 |
+
"loss": 1.7525,
|
| 326 |
+
"step": 86
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.99,
|
| 330 |
+
"eval_loss": 2.333530902862549,
|
| 331 |
+
"eval_runtime": 14.2215,
|
| 332 |
+
"eval_samples_per_second": 23.767,
|
| 333 |
+
"eval_steps_per_second": 3.024,
|
| 334 |
+
"step": 86
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 3.06,
|
| 338 |
+
"grad_norm": 0.37869980931282043,
|
| 339 |
+
"learning_rate": 0.001,
|
| 340 |
+
"loss": 1.3103,
|
| 341 |
+
"step": 88
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 3.13,
|
| 345 |
+
"grad_norm": 0.32068952918052673,
|
| 346 |
+
"learning_rate": 0.001,
|
| 347 |
+
"loss": 1.2763,
|
| 348 |
+
"step": 90
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 3.2,
|
| 352 |
+
"grad_norm": 0.33284562826156616,
|
| 353 |
+
"learning_rate": 0.001,
|
| 354 |
+
"loss": 1.2235,
|
| 355 |
+
"step": 92
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 3.27,
|
| 359 |
+
"grad_norm": 0.4027257561683655,
|
| 360 |
+
"learning_rate": 0.001,
|
| 361 |
+
"loss": 1.2606,
|
| 362 |
+
"step": 94
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.34,
|
| 366 |
+
"grad_norm": 0.8390576839447021,
|
| 367 |
+
"learning_rate": 0.001,
|
| 368 |
+
"loss": 1.2426,
|
| 369 |
+
"step": 96
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 3.41,
|
| 373 |
+
"grad_norm": 0.3938799798488617,
|
| 374 |
+
"learning_rate": 0.001,
|
| 375 |
+
"loss": 1.2499,
|
| 376 |
+
"step": 98
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 3.48,
|
| 380 |
+
"grad_norm": 0.39702850580215454,
|
| 381 |
+
"learning_rate": 0.001,
|
| 382 |
+
"loss": 1.3165,
|
| 383 |
+
"step": 100
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 3.55,
|
| 387 |
+
"grad_norm": 0.4196929633617401,
|
| 388 |
+
"learning_rate": 0.001,
|
| 389 |
+
"loss": 1.2884,
|
| 390 |
+
"step": 102
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 3.62,
|
| 394 |
+
"grad_norm": 0.4435763657093048,
|
| 395 |
+
"learning_rate": 0.001,
|
| 396 |
+
"loss": 1.3407,
|
| 397 |
+
"step": 104
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 3.69,
|
| 401 |
+
"grad_norm": 0.3989652395248413,
|
| 402 |
+
"learning_rate": 0.001,
|
| 403 |
+
"loss": 1.3119,
|
| 404 |
+
"step": 106
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.76,
|
| 408 |
+
"grad_norm": 0.35742542147636414,
|
| 409 |
+
"learning_rate": 0.001,
|
| 410 |
+
"loss": 1.2882,
|
| 411 |
+
"step": 108
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 3.83,
|
| 415 |
+
"grad_norm": 0.3512905538082123,
|
| 416 |
+
"learning_rate": 0.001,
|
| 417 |
+
"loss": 1.276,
|
| 418 |
+
"step": 110
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 3.9,
|
| 422 |
+
"grad_norm": 0.35384494066238403,
|
| 423 |
+
"learning_rate": 0.001,
|
| 424 |
+
"loss": 1.31,
|
| 425 |
+
"step": 112
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 3.97,
|
| 429 |
+
"grad_norm": 0.38055282831192017,
|
| 430 |
+
"learning_rate": 0.001,
|
| 431 |
+
"loss": 1.2954,
|
| 432 |
+
"step": 114
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 4.0,
|
| 436 |
+
"eval_loss": 2.4995365142822266,
|
| 437 |
+
"eval_runtime": 14.2261,
|
| 438 |
+
"eval_samples_per_second": 23.759,
|
| 439 |
+
"eval_steps_per_second": 3.023,
|
| 440 |
+
"step": 115
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 4.03,
|
| 444 |
+
"grad_norm": 0.33426040410995483,
|
| 445 |
+
"learning_rate": 0.001,
|
| 446 |
+
"loss": 1.0702,
|
| 447 |
+
"step": 116
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 4.1,
|
| 451 |
+
"grad_norm": 0.3915540277957916,
|
| 452 |
+
"learning_rate": 0.001,
|
| 453 |
+
"loss": 0.93,
|
| 454 |
+
"step": 118
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 4.17,
|
| 458 |
+
"grad_norm": 0.3885715901851654,
|
| 459 |
+
"learning_rate": 0.001,
|
| 460 |
+
"loss": 0.8923,
|
| 461 |
+
"step": 120
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 4.24,
|
| 465 |
+
"grad_norm": 0.3450148105621338,
|
| 466 |
+
"learning_rate": 0.001,
|
| 467 |
+
"loss": 0.9367,
|
| 468 |
+
"step": 122
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 4.31,
|
| 472 |
+
"grad_norm": 0.34307780861854553,
|
| 473 |
+
"learning_rate": 0.001,
|
| 474 |
+
"loss": 0.9471,
|
| 475 |
+
"step": 124
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 4.38,
|
| 479 |
+
"grad_norm": 0.3358236253261566,
|
| 480 |
+
"learning_rate": 0.001,
|
| 481 |
+
"loss": 0.9597,
|
| 482 |
+
"step": 126
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.45,
|
| 486 |
+
"grad_norm": 0.32632964849472046,
|
| 487 |
+
"learning_rate": 0.001,
|
| 488 |
+
"loss": 0.9678,
|
| 489 |
+
"step": 128
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.52,
|
| 493 |
+
"grad_norm": 0.3462707996368408,
|
| 494 |
+
"learning_rate": 0.001,
|
| 495 |
+
"loss": 0.975,
|
| 496 |
+
"step": 130
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.59,
|
| 500 |
+
"grad_norm": 0.4402617812156677,
|
| 501 |
+
"learning_rate": 0.001,
|
| 502 |
+
"loss": 0.9908,
|
| 503 |
+
"step": 132
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 4.66,
|
| 507 |
+
"grad_norm": 0.4284449815750122,
|
| 508 |
+
"learning_rate": 0.001,
|
| 509 |
+
"loss": 0.9721,
|
| 510 |
+
"step": 134
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 4.73,
|
| 514 |
+
"grad_norm": 0.3820173144340515,
|
| 515 |
+
"learning_rate": 0.001,
|
| 516 |
+
"loss": 0.983,
|
| 517 |
+
"step": 136
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 4.8,
|
| 521 |
+
"grad_norm": 0.5288845300674438,
|
| 522 |
+
"learning_rate": 0.001,
|
| 523 |
+
"loss": 0.9376,
|
| 524 |
+
"step": 138
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.87,
|
| 528 |
+
"grad_norm": 0.489346981048584,
|
| 529 |
+
"learning_rate": 0.001,
|
| 530 |
+
"loss": 0.9801,
|
| 531 |
+
"step": 140
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.94,
|
| 535 |
+
"grad_norm": 0.4548284709453583,
|
| 536 |
+
"learning_rate": 0.001,
|
| 537 |
+
"loss": 1.0518,
|
| 538 |
+
"step": 142
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.97,
|
| 542 |
+
"eval_loss": 2.832628011703491,
|
| 543 |
+
"eval_runtime": 14.24,
|
| 544 |
+
"eval_samples_per_second": 23.736,
|
| 545 |
+
"eval_steps_per_second": 3.02,
|
| 546 |
+
"step": 143
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 5.01,
|
| 550 |
+
"grad_norm": 0.4093753397464752,
|
| 551 |
+
"learning_rate": 0.001,
|
| 552 |
+
"loss": 0.9381,
|
| 553 |
+
"step": 144
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 5.08,
|
| 557 |
+
"grad_norm": 0.5986877679824829,
|
| 558 |
+
"learning_rate": 0.001,
|
| 559 |
+
"loss": 0.6649,
|
| 560 |
+
"step": 146
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 5.15,
|
| 564 |
+
"grad_norm": 0.4789908230304718,
|
| 565 |
+
"learning_rate": 0.001,
|
| 566 |
+
"loss": 0.6925,
|
| 567 |
+
"step": 148
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 5.22,
|
| 571 |
+
"grad_norm": 0.4392305612564087,
|
| 572 |
+
"learning_rate": 0.001,
|
| 573 |
+
"loss": 0.6901,
|
| 574 |
+
"step": 150
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 5.29,
|
| 578 |
+
"grad_norm": 0.49618610739707947,
|
| 579 |
+
"learning_rate": 0.001,
|
| 580 |
+
"loss": 0.7173,
|
| 581 |
+
"step": 152
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 5.36,
|
| 585 |
+
"grad_norm": 0.513729453086853,
|
| 586 |
+
"learning_rate": 0.001,
|
| 587 |
+
"loss": 0.6951,
|
| 588 |
+
"step": 154
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 5.43,
|
| 592 |
+
"grad_norm": 0.4622480571269989,
|
| 593 |
+
"learning_rate": 0.001,
|
| 594 |
+
"loss": 0.6757,
|
| 595 |
+
"step": 156
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 5.5,
|
| 599 |
+
"grad_norm": 0.4192960560321808,
|
| 600 |
+
"learning_rate": 0.001,
|
| 601 |
+
"loss": 0.7125,
|
| 602 |
+
"step": 158
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 5.57,
|
| 606 |
+
"grad_norm": 0.3887766897678375,
|
| 607 |
+
"learning_rate": 0.001,
|
| 608 |
+
"loss": 0.6905,
|
| 609 |
+
"step": 160
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 5.63,
|
| 613 |
+
"grad_norm": 0.3633786141872406,
|
| 614 |
+
"learning_rate": 0.001,
|
| 615 |
+
"loss": 0.73,
|
| 616 |
+
"step": 162
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 5.7,
|
| 620 |
+
"grad_norm": 0.4010884761810303,
|
| 621 |
+
"learning_rate": 0.001,
|
| 622 |
+
"loss": 0.7473,
|
| 623 |
+
"step": 164
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 5.77,
|
| 627 |
+
"grad_norm": 0.4541136920452118,
|
| 628 |
+
"learning_rate": 0.001,
|
| 629 |
+
"loss": 0.7659,
|
| 630 |
+
"step": 166
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 5.84,
|
| 634 |
+
"grad_norm": 0.4243777394294739,
|
| 635 |
+
"learning_rate": 0.001,
|
| 636 |
+
"loss": 0.7742,
|
| 637 |
+
"step": 168
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 5.91,
|
| 641 |
+
"grad_norm": 0.40389642119407654,
|
| 642 |
+
"learning_rate": 0.001,
|
| 643 |
+
"loss": 0.7312,
|
| 644 |
+
"step": 170
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 5.98,
|
| 648 |
+
"grad_norm": 0.39572134613990784,
|
| 649 |
+
"learning_rate": 0.001,
|
| 650 |
+
"loss": 0.7083,
|
| 651 |
+
"step": 172
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 5.98,
|
| 655 |
+
"eval_loss": 2.9095335006713867,
|
| 656 |
+
"eval_runtime": 14.2201,
|
| 657 |
+
"eval_samples_per_second": 23.769,
|
| 658 |
+
"eval_steps_per_second": 3.024,
|
| 659 |
+
"step": 172
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 6.05,
|
| 663 |
+
"grad_norm": 0.34924399852752686,
|
| 664 |
+
"learning_rate": 0.001,
|
| 665 |
+
"loss": 0.5277,
|
| 666 |
+
"step": 174
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 6.12,
|
| 670 |
+
"grad_norm": 0.42162197828292847,
|
| 671 |
+
"learning_rate": 0.001,
|
| 672 |
+
"loss": 0.5141,
|
| 673 |
+
"step": 176
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 6.19,
|
| 677 |
+
"grad_norm": 0.43131959438323975,
|
| 678 |
+
"learning_rate": 0.001,
|
| 679 |
+
"loss": 0.5295,
|
| 680 |
+
"step": 178
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 6.26,
|
| 684 |
+
"grad_norm": 0.3414030075073242,
|
| 685 |
+
"learning_rate": 0.001,
|
| 686 |
+
"loss": 0.5293,
|
| 687 |
+
"step": 180
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 6.33,
|
| 691 |
+
"grad_norm": 0.2947409451007843,
|
| 692 |
+
"learning_rate": 0.001,
|
| 693 |
+
"loss": 0.5094,
|
| 694 |
+
"step": 182
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 6.4,
|
| 698 |
+
"grad_norm": 0.3493882119655609,
|
| 699 |
+
"learning_rate": 0.001,
|
| 700 |
+
"loss": 0.5103,
|
| 701 |
+
"step": 184
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 6.47,
|
| 705 |
+
"grad_norm": 0.32564929127693176,
|
| 706 |
+
"learning_rate": 0.001,
|
| 707 |
+
"loss": 0.4955,
|
| 708 |
+
"step": 186
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 6.54,
|
| 712 |
+
"grad_norm": 0.3580136299133301,
|
| 713 |
+
"learning_rate": 0.001,
|
| 714 |
+
"loss": 0.4923,
|
| 715 |
+
"step": 188
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 6.61,
|
| 719 |
+
"grad_norm": 0.40878280997276306,
|
| 720 |
+
"learning_rate": 0.001,
|
| 721 |
+
"loss": 0.5312,
|
| 722 |
+
"step": 190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 6.68,
|
| 726 |
+
"grad_norm": 0.42687222361564636,
|
| 727 |
+
"learning_rate": 0.001,
|
| 728 |
+
"loss": 0.5977,
|
| 729 |
+
"step": 192
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 6.75,
|
| 733 |
+
"grad_norm": 0.3438556492328644,
|
| 734 |
+
"learning_rate": 0.001,
|
| 735 |
+
"loss": 0.5811,
|
| 736 |
+
"step": 194
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 6.82,
|
| 740 |
+
"grad_norm": 0.3497261703014374,
|
| 741 |
+
"learning_rate": 0.001,
|
| 742 |
+
"loss": 0.542,
|
| 743 |
+
"step": 196
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 6.89,
|
| 747 |
+
"grad_norm": 0.5538045167922974,
|
| 748 |
+
"learning_rate": 0.001,
|
| 749 |
+
"loss": 0.5392,
|
| 750 |
+
"step": 198
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 6.96,
|
| 754 |
+
"grad_norm": 0.40168508887290955,
|
| 755 |
+
"learning_rate": 0.001,
|
| 756 |
+
"loss": 0.5124,
|
| 757 |
+
"step": 200
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 6.99,
|
| 761 |
+
"eval_loss": 3.4108099937438965,
|
| 762 |
+
"eval_runtime": 14.2162,
|
| 763 |
+
"eval_samples_per_second": 23.776,
|
| 764 |
+
"eval_steps_per_second": 3.025,
|
| 765 |
+
"step": 201
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 7.03,
|
| 769 |
+
"grad_norm": 0.45531949400901794,
|
| 770 |
+
"learning_rate": 0.001,
|
| 771 |
+
"loss": 0.4621,
|
| 772 |
+
"step": 202
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 7.1,
|
| 776 |
+
"grad_norm": 0.420217901468277,
|
| 777 |
+
"learning_rate": 0.001,
|
| 778 |
+
"loss": 0.3526,
|
| 779 |
+
"step": 204
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 7.17,
|
| 783 |
+
"grad_norm": 0.39295145869255066,
|
| 784 |
+
"learning_rate": 0.001,
|
| 785 |
+
"loss": 0.3522,
|
| 786 |
+
"step": 206
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 7.23,
|
| 790 |
+
"grad_norm": 0.42002934217453003,
|
| 791 |
+
"learning_rate": 0.001,
|
| 792 |
+
"loss": 0.37,
|
| 793 |
+
"step": 208
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 7.3,
|
| 797 |
+
"grad_norm": 0.3919942378997803,
|
| 798 |
+
"learning_rate": 0.001,
|
| 799 |
+
"loss": 0.4123,
|
| 800 |
+
"step": 210
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 7.37,
|
| 804 |
+
"grad_norm": 0.34964632987976074,
|
| 805 |
+
"learning_rate": 0.001,
|
| 806 |
+
"loss": 0.4155,
|
| 807 |
+
"step": 212
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 7.44,
|
| 811 |
+
"grad_norm": 0.3752292990684509,
|
| 812 |
+
"learning_rate": 0.001,
|
| 813 |
+
"loss": 0.392,
|
| 814 |
+
"step": 214
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 7.51,
|
| 818 |
+
"grad_norm": 0.31345558166503906,
|
| 819 |
+
"learning_rate": 0.001,
|
| 820 |
+
"loss": 0.3973,
|
| 821 |
+
"step": 216
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 7.58,
|
| 825 |
+
"grad_norm": 0.32503190636634827,
|
| 826 |
+
"learning_rate": 0.001,
|
| 827 |
+
"loss": 0.3788,
|
| 828 |
+
"step": 218
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 7.65,
|
| 832 |
+
"grad_norm": 0.40439680218696594,
|
| 833 |
+
"learning_rate": 0.001,
|
| 834 |
+
"loss": 0.3997,
|
| 835 |
+
"step": 220
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 7.72,
|
| 839 |
+
"grad_norm": 0.5034092664718628,
|
| 840 |
+
"learning_rate": 0.001,
|
| 841 |
+
"loss": 0.3934,
|
| 842 |
+
"step": 222
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 7.79,
|
| 846 |
+
"grad_norm": 0.4480283260345459,
|
| 847 |
+
"learning_rate": 0.001,
|
| 848 |
+
"loss": 0.4095,
|
| 849 |
+
"step": 224
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 7.86,
|
| 853 |
+
"grad_norm": 0.44010230898857117,
|
| 854 |
+
"learning_rate": 0.001,
|
| 855 |
+
"loss": 0.3963,
|
| 856 |
+
"step": 226
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 7.93,
|
| 860 |
+
"grad_norm": 0.4241428077220917,
|
| 861 |
+
"learning_rate": 0.001,
|
| 862 |
+
"loss": 0.4081,
|
| 863 |
+
"step": 228
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 8.0,
|
| 867 |
+
"grad_norm": 1.4331179857254028,
|
| 868 |
+
"learning_rate": 0.001,
|
| 869 |
+
"loss": 0.4503,
|
| 870 |
+
"step": 230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 8.0,
|
| 874 |
+
"eval_loss": 3.445927858352661,
|
| 875 |
+
"eval_runtime": 14.2138,
|
| 876 |
+
"eval_samples_per_second": 23.78,
|
| 877 |
+
"eval_steps_per_second": 3.025,
|
| 878 |
+
"step": 230
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 8.07,
|
| 882 |
+
"grad_norm": 0.26638248562812805,
|
| 883 |
+
"learning_rate": 0.001,
|
| 884 |
+
"loss": 0.2491,
|
| 885 |
+
"step": 232
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 8.14,
|
| 889 |
+
"grad_norm": 0.30395275354385376,
|
| 890 |
+
"learning_rate": 0.001,
|
| 891 |
+
"loss": 0.2501,
|
| 892 |
+
"step": 234
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 8.21,
|
| 896 |
+
"grad_norm": 0.36178329586982727,
|
| 897 |
+
"learning_rate": 0.001,
|
| 898 |
+
"loss": 0.2818,
|
| 899 |
+
"step": 236
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 8.28,
|
| 903 |
+
"grad_norm": 0.33713603019714355,
|
| 904 |
+
"learning_rate": 0.001,
|
| 905 |
+
"loss": 0.3005,
|
| 906 |
+
"step": 238
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 8.35,
|
| 910 |
+
"grad_norm": 0.2999133765697479,
|
| 911 |
+
"learning_rate": 0.001,
|
| 912 |
+
"loss": 0.2863,
|
| 913 |
+
"step": 240
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 8.42,
|
| 917 |
+
"grad_norm": 0.4105747938156128,
|
| 918 |
+
"learning_rate": 0.001,
|
| 919 |
+
"loss": 0.2946,
|
| 920 |
+
"step": 242
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 8.49,
|
| 924 |
+
"grad_norm": 0.3367477357387543,
|
| 925 |
+
"learning_rate": 0.001,
|
| 926 |
+
"loss": 0.3052,
|
| 927 |
+
"step": 244
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 8.56,
|
| 931 |
+
"grad_norm": 0.3373625874519348,
|
| 932 |
+
"learning_rate": 0.001,
|
| 933 |
+
"loss": 0.3166,
|
| 934 |
+
"step": 246
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 8.63,
|
| 938 |
+
"grad_norm": 0.2816782295703888,
|
| 939 |
+
"learning_rate": 0.001,
|
| 940 |
+
"loss": 0.2819,
|
| 941 |
+
"step": 248
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 8.7,
|
| 945 |
+
"grad_norm": 0.30990245938301086,
|
| 946 |
+
"learning_rate": 0.001,
|
| 947 |
+
"loss": 0.3023,
|
| 948 |
+
"step": 250
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 8.77,
|
| 952 |
+
"grad_norm": 0.317903608083725,
|
| 953 |
+
"learning_rate": 0.001,
|
| 954 |
+
"loss": 0.3128,
|
| 955 |
+
"step": 252
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 8.83,
|
| 959 |
+
"grad_norm": 0.3024023473262787,
|
| 960 |
+
"learning_rate": 0.001,
|
| 961 |
+
"loss": 0.3167,
|
| 962 |
+
"step": 254
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 8.9,
|
| 966 |
+
"grad_norm": 0.3257545828819275,
|
| 967 |
+
"learning_rate": 0.001,
|
| 968 |
+
"loss": 0.3121,
|
| 969 |
+
"step": 256
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 8.97,
|
| 973 |
+
"grad_norm": 0.34182751178741455,
|
| 974 |
+
"learning_rate": 0.001,
|
| 975 |
+
"loss": 0.3145,
|
| 976 |
+
"step": 258
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 8.97,
|
| 980 |
+
"eval_loss": 3.52162766456604,
|
| 981 |
+
"eval_runtime": 14.1996,
|
| 982 |
+
"eval_samples_per_second": 23.804,
|
| 983 |
+
"eval_steps_per_second": 3.028,
|
| 984 |
+
"step": 258
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 9.04,
|
| 988 |
+
"grad_norm": 0.2942521274089813,
|
| 989 |
+
"learning_rate": 0.001,
|
| 990 |
+
"loss": 0.2462,
|
| 991 |
+
"step": 260
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 9.11,
|
| 995 |
+
"grad_norm": 0.2955419719219208,
|
| 996 |
+
"learning_rate": 0.001,
|
| 997 |
+
"loss": 0.205,
|
| 998 |
+
"step": 262
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 9.18,
|
| 1002 |
+
"grad_norm": 0.29875773191452026,
|
| 1003 |
+
"learning_rate": 0.001,
|
| 1004 |
+
"loss": 0.2185,
|
| 1005 |
+
"step": 264
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 9.25,
|
| 1009 |
+
"grad_norm": 0.9682367444038391,
|
| 1010 |
+
"learning_rate": 0.001,
|
| 1011 |
+
"loss": 0.5828,
|
| 1012 |
+
"step": 266
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 9.32,
|
| 1016 |
+
"grad_norm": 0.32209742069244385,
|
| 1017 |
+
"learning_rate": 0.001,
|
| 1018 |
+
"loss": 0.2255,
|
| 1019 |
+
"step": 268
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 9.39,
|
| 1023 |
+
"grad_norm": 0.2810392379760742,
|
| 1024 |
+
"learning_rate": 0.001,
|
| 1025 |
+
"loss": 0.2157,
|
| 1026 |
+
"step": 270
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 9.46,
|
| 1030 |
+
"grad_norm": 0.26448503136634827,
|
| 1031 |
+
"learning_rate": 0.001,
|
| 1032 |
+
"loss": 0.2136,
|
| 1033 |
+
"step": 272
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 9.53,
|
| 1037 |
+
"grad_norm": 0.2602413296699524,
|
| 1038 |
+
"learning_rate": 0.001,
|
| 1039 |
+
"loss": 0.2264,
|
| 1040 |
+
"step": 274
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 9.6,
|
| 1044 |
+
"grad_norm": 0.30358317494392395,
|
| 1045 |
+
"learning_rate": 0.001,
|
| 1046 |
+
"loss": 0.2349,
|
| 1047 |
+
"step": 276
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 9.67,
|
| 1051 |
+
"grad_norm": 0.2755320966243744,
|
| 1052 |
+
"learning_rate": 0.001,
|
| 1053 |
+
"loss": 0.2346,
|
| 1054 |
+
"step": 278
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 9.74,
|
| 1058 |
+
"grad_norm": 0.257522314786911,
|
| 1059 |
+
"learning_rate": 0.001,
|
| 1060 |
+
"loss": 0.2312,
|
| 1061 |
+
"step": 280
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 9.81,
|
| 1065 |
+
"grad_norm": 0.3185376524925232,
|
| 1066 |
+
"learning_rate": 0.001,
|
| 1067 |
+
"loss": 0.2519,
|
| 1068 |
+
"step": 282
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 9.88,
|
| 1072 |
+
"grad_norm": 0.352522611618042,
|
| 1073 |
+
"learning_rate": 0.001,
|
| 1074 |
+
"loss": 0.2678,
|
| 1075 |
+
"step": 284
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 9.95,
|
| 1079 |
+
"grad_norm": 0.3061072826385498,
|
| 1080 |
+
"learning_rate": 0.001,
|
| 1081 |
+
"loss": 0.2573,
|
| 1082 |
+
"step": 286
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 9.98,
|
| 1086 |
+
"eval_loss": 4.012668609619141,
|
| 1087 |
+
"eval_runtime": 14.2083,
|
| 1088 |
+
"eval_samples_per_second": 23.789,
|
| 1089 |
+
"eval_steps_per_second": 3.026,
|
| 1090 |
+
"step": 287
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 10.02,
|
| 1094 |
+
"grad_norm": 0.34469902515411377,
|
| 1095 |
+
"learning_rate": 0.001,
|
| 1096 |
+
"loss": 0.2394,
|
| 1097 |
+
"step": 288
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 10.09,
|
| 1101 |
+
"grad_norm": 0.26758918166160583,
|
| 1102 |
+
"learning_rate": 0.001,
|
| 1103 |
+
"loss": 0.1749,
|
| 1104 |
+
"step": 290
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 10.16,
|
| 1108 |
+
"grad_norm": 0.24024738371372223,
|
| 1109 |
+
"learning_rate": 0.001,
|
| 1110 |
+
"loss": 0.1687,
|
| 1111 |
+
"step": 292
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 10.23,
|
| 1115 |
+
"grad_norm": 0.2440188080072403,
|
| 1116 |
+
"learning_rate": 0.001,
|
| 1117 |
+
"loss": 0.1634,
|
| 1118 |
+
"step": 294
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 10.3,
|
| 1122 |
+
"grad_norm": 0.2564130425453186,
|
| 1123 |
+
"learning_rate": 0.001,
|
| 1124 |
+
"loss": 0.1759,
|
| 1125 |
+
"step": 296
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 10.37,
|
| 1129 |
+
"grad_norm": 0.27317214012145996,
|
| 1130 |
+
"learning_rate": 0.001,
|
| 1131 |
+
"loss": 0.19,
|
| 1132 |
+
"step": 298
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 10.43,
|
| 1136 |
+
"grad_norm": 0.3196158707141876,
|
| 1137 |
+
"learning_rate": 0.001,
|
| 1138 |
+
"loss": 0.1903,
|
| 1139 |
+
"step": 300
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 10.5,
|
| 1143 |
+
"grad_norm": 0.3399882912635803,
|
| 1144 |
+
"learning_rate": 0.001,
|
| 1145 |
+
"loss": 0.2016,
|
| 1146 |
+
"step": 302
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 10.57,
|
| 1150 |
+
"grad_norm": 0.28635892271995544,
|
| 1151 |
+
"learning_rate": 0.001,
|
| 1152 |
+
"loss": 0.189,
|
| 1153 |
+
"step": 304
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 10.64,
|
| 1157 |
+
"grad_norm": 0.2602458596229553,
|
| 1158 |
+
"learning_rate": 0.001,
|
| 1159 |
+
"loss": 0.1792,
|
| 1160 |
+
"step": 306
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 10.71,
|
| 1164 |
+
"grad_norm": 0.29866456985473633,
|
| 1165 |
+
"learning_rate": 0.001,
|
| 1166 |
+
"loss": 0.1974,
|
| 1167 |
+
"step": 308
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 10.78,
|
| 1171 |
+
"grad_norm": 0.2757103443145752,
|
| 1172 |
+
"learning_rate": 0.001,
|
| 1173 |
+
"loss": 0.2025,
|
| 1174 |
+
"step": 310
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 10.85,
|
| 1178 |
+
"grad_norm": 0.2643924355506897,
|
| 1179 |
+
"learning_rate": 0.001,
|
| 1180 |
+
"loss": 0.2106,
|
| 1181 |
+
"step": 312
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 10.92,
|
| 1185 |
+
"grad_norm": 0.27770230174064636,
|
| 1186 |
+
"learning_rate": 0.001,
|
| 1187 |
+
"loss": 0.1999,
|
| 1188 |
+
"step": 314
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 10.99,
|
| 1192 |
+
"grad_norm": 0.2809506356716156,
|
| 1193 |
+
"learning_rate": 0.001,
|
| 1194 |
+
"loss": 0.213,
|
| 1195 |
+
"step": 316
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 10.99,
|
| 1199 |
+
"eval_loss": 3.9713540077209473,
|
| 1200 |
+
"eval_runtime": 14.2091,
|
| 1201 |
+
"eval_samples_per_second": 23.788,
|
| 1202 |
+
"eval_steps_per_second": 3.026,
|
| 1203 |
+
"step": 316
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 11.06,
|
| 1207 |
+
"grad_norm": 0.19931960105895996,
|
| 1208 |
+
"learning_rate": 0.001,
|
| 1209 |
+
"loss": 0.1467,
|
| 1210 |
+
"step": 318
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 11.13,
|
| 1214 |
+
"grad_norm": 0.20454713702201843,
|
| 1215 |
+
"learning_rate": 0.001,
|
| 1216 |
+
"loss": 0.1357,
|
| 1217 |
+
"step": 320
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 11.2,
|
| 1221 |
+
"grad_norm": 0.22833150625228882,
|
| 1222 |
+
"learning_rate": 0.001,
|
| 1223 |
+
"loss": 0.1357,
|
| 1224 |
+
"step": 322
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 11.27,
|
| 1228 |
+
"grad_norm": 0.45052453875541687,
|
| 1229 |
+
"learning_rate": 0.001,
|
| 1230 |
+
"loss": 0.1401,
|
| 1231 |
+
"step": 324
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 11.34,
|
| 1235 |
+
"grad_norm": 0.2621672451496124,
|
| 1236 |
+
"learning_rate": 0.001,
|
| 1237 |
+
"loss": 0.1402,
|
| 1238 |
+
"step": 326
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 11.41,
|
| 1242 |
+
"grad_norm": 0.2570265829563141,
|
| 1243 |
+
"learning_rate": 0.001,
|
| 1244 |
+
"loss": 0.1743,
|
| 1245 |
+
"step": 328
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 11.48,
|
| 1249 |
+
"grad_norm": 0.2333604246377945,
|
| 1250 |
+
"learning_rate": 0.001,
|
| 1251 |
+
"loss": 0.1572,
|
| 1252 |
+
"step": 330
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 11.55,
|
| 1256 |
+
"grad_norm": 0.2325439304113388,
|
| 1257 |
+
"learning_rate": 0.001,
|
| 1258 |
+
"loss": 0.1505,
|
| 1259 |
+
"step": 332
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 11.62,
|
| 1263 |
+
"grad_norm": 0.2524455487728119,
|
| 1264 |
+
"learning_rate": 0.001,
|
| 1265 |
+
"loss": 0.1574,
|
| 1266 |
+
"step": 334
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 11.69,
|
| 1270 |
+
"grad_norm": 0.2821509838104248,
|
| 1271 |
+
"learning_rate": 0.001,
|
| 1272 |
+
"loss": 0.162,
|
| 1273 |
+
"step": 336
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 11.76,
|
| 1277 |
+
"grad_norm": 0.2627921402454376,
|
| 1278 |
+
"learning_rate": 0.001,
|
| 1279 |
+
"loss": 0.1823,
|
| 1280 |
+
"step": 338
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 11.83,
|
| 1284 |
+
"grad_norm": 0.20858174562454224,
|
| 1285 |
+
"learning_rate": 0.001,
|
| 1286 |
+
"loss": 0.1581,
|
| 1287 |
+
"step": 340
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 11.9,
|
| 1291 |
+
"grad_norm": 0.24444954097270966,
|
| 1292 |
+
"learning_rate": 0.001,
|
| 1293 |
+
"loss": 0.1528,
|
| 1294 |
+
"step": 342
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 11.97,
|
| 1298 |
+
"grad_norm": 0.23985397815704346,
|
| 1299 |
+
"learning_rate": 0.001,
|
| 1300 |
+
"loss": 0.1682,
|
| 1301 |
+
"step": 344
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 12.0,
|
| 1305 |
+
"eval_loss": 3.842698097229004,
|
| 1306 |
+
"eval_runtime": 14.2204,
|
| 1307 |
+
"eval_samples_per_second": 23.769,
|
| 1308 |
+
"eval_steps_per_second": 3.024,
|
| 1309 |
+
"step": 345
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 12.03,
|
| 1313 |
+
"grad_norm": 0.20599780976772308,
|
| 1314 |
+
"learning_rate": 0.001,
|
| 1315 |
+
"loss": 0.1472,
|
| 1316 |
+
"step": 346
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 12.1,
|
| 1320 |
+
"grad_norm": 0.20421436429023743,
|
| 1321 |
+
"learning_rate": 0.001,
|
| 1322 |
+
"loss": 0.1136,
|
| 1323 |
+
"step": 348
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 12.17,
|
| 1327 |
+
"grad_norm": 0.21933647990226746,
|
| 1328 |
+
"learning_rate": 0.001,
|
| 1329 |
+
"loss": 0.1137,
|
| 1330 |
+
"step": 350
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 12.24,
|
| 1334 |
+
"grad_norm": 0.2488287091255188,
|
| 1335 |
+
"learning_rate": 0.001,
|
| 1336 |
+
"loss": 0.1223,
|
| 1337 |
+
"step": 352
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 12.31,
|
| 1341 |
+
"grad_norm": 0.21850407123565674,
|
| 1342 |
+
"learning_rate": 0.001,
|
| 1343 |
+
"loss": 0.1266,
|
| 1344 |
+
"step": 354
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 12.38,
|
| 1348 |
+
"grad_norm": 0.219721257686615,
|
| 1349 |
+
"learning_rate": 0.001,
|
| 1350 |
+
"loss": 0.1266,
|
| 1351 |
+
"step": 356
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 12.45,
|
| 1355 |
+
"grad_norm": 0.44860273599624634,
|
| 1356 |
+
"learning_rate": 0.001,
|
| 1357 |
+
"loss": 0.1271,
|
| 1358 |
+
"step": 358
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 12.52,
|
| 1362 |
+
"grad_norm": 0.20995499193668365,
|
| 1363 |
+
"learning_rate": 0.001,
|
| 1364 |
+
"loss": 0.1201,
|
| 1365 |
+
"step": 360
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 12.59,
|
| 1369 |
+
"grad_norm": 0.2180211991071701,
|
| 1370 |
+
"learning_rate": 0.001,
|
| 1371 |
+
"loss": 0.1325,
|
| 1372 |
+
"step": 362
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 12.66,
|
| 1376 |
+
"grad_norm": 0.23486308753490448,
|
| 1377 |
+
"learning_rate": 0.001,
|
| 1378 |
+
"loss": 0.1316,
|
| 1379 |
+
"step": 364
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 12.73,
|
| 1383 |
+
"grad_norm": 0.26361599564552307,
|
| 1384 |
+
"learning_rate": 0.001,
|
| 1385 |
+
"loss": 0.1417,
|
| 1386 |
+
"step": 366
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 12.8,
|
| 1390 |
+
"grad_norm": 0.27585896849632263,
|
| 1391 |
+
"learning_rate": 0.001,
|
| 1392 |
+
"loss": 0.1514,
|
| 1393 |
+
"step": 368
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 12.87,
|
| 1397 |
+
"grad_norm": 0.23249991238117218,
|
| 1398 |
+
"learning_rate": 0.001,
|
| 1399 |
+
"loss": 0.1387,
|
| 1400 |
+
"step": 370
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 12.94,
|
| 1404 |
+
"grad_norm": 0.2378188967704773,
|
| 1405 |
+
"learning_rate": 0.001,
|
| 1406 |
+
"loss": 0.1396,
|
| 1407 |
+
"step": 372
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 12.97,
|
| 1411 |
+
"eval_loss": 4.202455520629883,
|
| 1412 |
+
"eval_runtime": 14.2176,
|
| 1413 |
+
"eval_samples_per_second": 23.773,
|
| 1414 |
+
"eval_steps_per_second": 3.024,
|
| 1415 |
+
"step": 373
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 13.01,
|
| 1419 |
+
"grad_norm": 0.25618231296539307,
|
| 1420 |
+
"learning_rate": 0.001,
|
| 1421 |
+
"loss": 0.1403,
|
| 1422 |
+
"step": 374
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 13.08,
|
| 1426 |
+
"grad_norm": 0.19359038770198822,
|
| 1427 |
+
"learning_rate": 0.001,
|
| 1428 |
+
"loss": 0.1048,
|
| 1429 |
+
"step": 376
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 13.15,
|
| 1433 |
+
"grad_norm": 0.20456379652023315,
|
| 1434 |
+
"learning_rate": 0.001,
|
| 1435 |
+
"loss": 0.102,
|
| 1436 |
+
"step": 378
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 13.22,
|
| 1440 |
+
"grad_norm": 0.20489871501922607,
|
| 1441 |
+
"learning_rate": 0.001,
|
| 1442 |
+
"loss": 0.1039,
|
| 1443 |
+
"step": 380
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 13.29,
|
| 1447 |
+
"grad_norm": 0.20058828592300415,
|
| 1448 |
+
"learning_rate": 0.001,
|
| 1449 |
+
"loss": 0.1066,
|
| 1450 |
+
"step": 382
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 13.36,
|
| 1454 |
+
"grad_norm": 0.20939727127552032,
|
| 1455 |
+
"learning_rate": 0.001,
|
| 1456 |
+
"loss": 0.1017,
|
| 1457 |
+
"step": 384
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 13.43,
|
| 1461 |
+
"grad_norm": 0.20905180275440216,
|
| 1462 |
+
"learning_rate": 0.001,
|
| 1463 |
+
"loss": 0.1074,
|
| 1464 |
+
"step": 386
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 13.5,
|
| 1468 |
+
"grad_norm": 0.2082565426826477,
|
| 1469 |
+
"learning_rate": 0.001,
|
| 1470 |
+
"loss": 0.1122,
|
| 1471 |
+
"step": 388
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 13.57,
|
| 1475 |
+
"grad_norm": 0.20671486854553223,
|
| 1476 |
+
"learning_rate": 0.001,
|
| 1477 |
+
"loss": 0.1137,
|
| 1478 |
+
"step": 390
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 13.63,
|
| 1482 |
+
"grad_norm": 0.20714151859283447,
|
| 1483 |
+
"learning_rate": 0.001,
|
| 1484 |
+
"loss": 0.1085,
|
| 1485 |
+
"step": 392
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 13.7,
|
| 1489 |
+
"grad_norm": 0.21117734909057617,
|
| 1490 |
+
"learning_rate": 0.001,
|
| 1491 |
+
"loss": 0.1225,
|
| 1492 |
+
"step": 394
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 13.77,
|
| 1496 |
+
"grad_norm": 0.20374175906181335,
|
| 1497 |
+
"learning_rate": 0.001,
|
| 1498 |
+
"loss": 0.1135,
|
| 1499 |
+
"step": 396
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 13.84,
|
| 1503 |
+
"grad_norm": 0.2223602831363678,
|
| 1504 |
+
"learning_rate": 0.001,
|
| 1505 |
+
"loss": 0.1188,
|
| 1506 |
+
"step": 398
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 13.91,
|
| 1510 |
+
"grad_norm": 0.2405446618795395,
|
| 1511 |
+
"learning_rate": 0.001,
|
| 1512 |
+
"loss": 0.1329,
|
| 1513 |
+
"step": 400
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 13.98,
|
| 1517 |
+
"grad_norm": 0.25538983941078186,
|
| 1518 |
+
"learning_rate": 0.001,
|
| 1519 |
+
"loss": 0.1363,
|
| 1520 |
+
"step": 402
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 13.98,
|
| 1524 |
+
"eval_loss": 4.401235103607178,
|
| 1525 |
+
"eval_runtime": 14.2163,
|
| 1526 |
+
"eval_samples_per_second": 23.776,
|
| 1527 |
+
"eval_steps_per_second": 3.025,
|
| 1528 |
+
"step": 402
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 14.05,
|
| 1532 |
+
"grad_norm": 0.20007029175758362,
|
| 1533 |
+
"learning_rate": 0.001,
|
| 1534 |
+
"loss": 0.0995,
|
| 1535 |
+
"step": 404
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 14.12,
|
| 1539 |
+
"grad_norm": 0.17897407710552216,
|
| 1540 |
+
"learning_rate": 0.001,
|
| 1541 |
+
"loss": 0.0927,
|
| 1542 |
+
"step": 406
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 14.19,
|
| 1546 |
+
"grad_norm": 0.18746937811374664,
|
| 1547 |
+
"learning_rate": 0.001,
|
| 1548 |
+
"loss": 0.0898,
|
| 1549 |
+
"step": 408
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 14.26,
|
| 1553 |
+
"grad_norm": 0.1855156570672989,
|
| 1554 |
+
"learning_rate": 0.001,
|
| 1555 |
+
"loss": 0.0913,
|
| 1556 |
+
"step": 410
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 14.33,
|
| 1560 |
+
"grad_norm": 0.1712156981229782,
|
| 1561 |
+
"learning_rate": 0.001,
|
| 1562 |
+
"loss": 0.0891,
|
| 1563 |
+
"step": 412
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 14.4,
|
| 1567 |
+
"grad_norm": 0.193171888589859,
|
| 1568 |
+
"learning_rate": 0.001,
|
| 1569 |
+
"loss": 0.0871,
|
| 1570 |
+
"step": 414
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 14.47,
|
| 1574 |
+
"grad_norm": 0.19764076173305511,
|
| 1575 |
+
"learning_rate": 0.001,
|
| 1576 |
+
"loss": 0.0892,
|
| 1577 |
+
"step": 416
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 14.54,
|
| 1581 |
+
"grad_norm": 0.20630787312984467,
|
| 1582 |
+
"learning_rate": 0.001,
|
| 1583 |
+
"loss": 0.0984,
|
| 1584 |
+
"step": 418
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 14.61,
|
| 1588 |
+
"grad_norm": 0.18784281611442566,
|
| 1589 |
+
"learning_rate": 0.001,
|
| 1590 |
+
"loss": 0.1002,
|
| 1591 |
+
"step": 420
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 14.68,
|
| 1595 |
+
"grad_norm": 0.2134091705083847,
|
| 1596 |
+
"learning_rate": 0.001,
|
| 1597 |
+
"loss": 0.0994,
|
| 1598 |
+
"step": 422
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 14.75,
|
| 1602 |
+
"grad_norm": 0.20559023320674896,
|
| 1603 |
+
"learning_rate": 0.001,
|
| 1604 |
+
"loss": 0.1036,
|
| 1605 |
+
"step": 424
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 14.82,
|
| 1609 |
+
"grad_norm": 0.22063368558883667,
|
| 1610 |
+
"learning_rate": 0.001,
|
| 1611 |
+
"loss": 0.1008,
|
| 1612 |
+
"step": 426
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 14.89,
|
| 1616 |
+
"grad_norm": 0.21584555506706238,
|
| 1617 |
+
"learning_rate": 0.001,
|
| 1618 |
+
"loss": 0.1137,
|
| 1619 |
+
"step": 428
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 14.96,
|
| 1623 |
+
"grad_norm": 0.20185333490371704,
|
| 1624 |
+
"learning_rate": 0.001,
|
| 1625 |
+
"loss": 0.1148,
|
| 1626 |
+
"step": 430
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 14.99,
|
| 1630 |
+
"eval_loss": 4.717379093170166,
|
| 1631 |
+
"eval_runtime": 14.2112,
|
| 1632 |
+
"eval_samples_per_second": 23.784,
|
| 1633 |
+
"eval_steps_per_second": 3.026,
|
| 1634 |
+
"step": 431
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 15.03,
|
| 1638 |
+
"grad_norm": 0.22225050628185272,
|
| 1639 |
+
"learning_rate": 0.001,
|
| 1640 |
+
"loss": 0.0942,
|
| 1641 |
+
"step": 432
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 15.1,
|
| 1645 |
+
"grad_norm": 0.2037811130285263,
|
| 1646 |
+
"learning_rate": 0.001,
|
| 1647 |
+
"loss": 0.0776,
|
| 1648 |
+
"step": 434
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 15.17,
|
| 1652 |
+
"grad_norm": 0.1820273995399475,
|
| 1653 |
+
"learning_rate": 0.001,
|
| 1654 |
+
"loss": 0.0849,
|
| 1655 |
+
"step": 436
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 15.23,
|
| 1659 |
+
"grad_norm": 0.18018700182437897,
|
| 1660 |
+
"learning_rate": 0.001,
|
| 1661 |
+
"loss": 0.0744,
|
| 1662 |
+
"step": 438
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 15.3,
|
| 1666 |
+
"grad_norm": 0.21271073818206787,
|
| 1667 |
+
"learning_rate": 0.001,
|
| 1668 |
+
"loss": 0.0836,
|
| 1669 |
+
"step": 440
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 15.37,
|
| 1673 |
+
"grad_norm": 0.19622232019901276,
|
| 1674 |
+
"learning_rate": 0.001,
|
| 1675 |
+
"loss": 0.0841,
|
| 1676 |
+
"step": 442
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 15.44,
|
| 1680 |
+
"grad_norm": 0.19101227819919586,
|
| 1681 |
+
"learning_rate": 0.001,
|
| 1682 |
+
"loss": 0.085,
|
| 1683 |
+
"step": 444
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 15.51,
|
| 1687 |
+
"grad_norm": 0.18909965455532074,
|
| 1688 |
+
"learning_rate": 0.001,
|
| 1689 |
+
"loss": 0.093,
|
| 1690 |
+
"step": 446
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 15.58,
|
| 1694 |
+
"grad_norm": 0.2003660649061203,
|
| 1695 |
+
"learning_rate": 0.001,
|
| 1696 |
+
"loss": 0.0877,
|
| 1697 |
+
"step": 448
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 15.65,
|
| 1701 |
+
"grad_norm": 0.2101544439792633,
|
| 1702 |
+
"learning_rate": 0.001,
|
| 1703 |
+
"loss": 0.0908,
|
| 1704 |
+
"step": 450
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 15.72,
|
| 1708 |
+
"grad_norm": 0.20344135165214539,
|
| 1709 |
+
"learning_rate": 0.001,
|
| 1710 |
+
"loss": 0.0943,
|
| 1711 |
+
"step": 452
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 15.79,
|
| 1715 |
+
"grad_norm": 0.28948044776916504,
|
| 1716 |
+
"learning_rate": 0.001,
|
| 1717 |
+
"loss": 0.1094,
|
| 1718 |
+
"step": 454
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 15.86,
|
| 1722 |
+
"grad_norm": 0.29758819937705994,
|
| 1723 |
+
"learning_rate": 0.001,
|
| 1724 |
+
"loss": 0.1247,
|
| 1725 |
+
"step": 456
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 15.93,
|
| 1729 |
+
"grad_norm": 0.24994027614593506,
|
| 1730 |
+
"learning_rate": 0.001,
|
| 1731 |
+
"loss": 0.1106,
|
| 1732 |
+
"step": 458
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 16.0,
|
| 1736 |
+
"grad_norm": 0.2185268998146057,
|
| 1737 |
+
"learning_rate": 0.001,
|
| 1738 |
+
"loss": 0.0907,
|
| 1739 |
+
"step": 460
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 16.0,
|
| 1743 |
+
"eval_loss": 4.497994899749756,
|
| 1744 |
+
"eval_runtime": 14.2114,
|
| 1745 |
+
"eval_samples_per_second": 23.784,
|
| 1746 |
+
"eval_steps_per_second": 3.026,
|
| 1747 |
+
"step": 460
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 16.07,
|
| 1751 |
+
"grad_norm": 0.1781373769044876,
|
| 1752 |
+
"learning_rate": 0.001,
|
| 1753 |
+
"loss": 0.0698,
|
| 1754 |
+
"step": 462
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 16.14,
|
| 1758 |
+
"grad_norm": 0.20170117914676666,
|
| 1759 |
+
"learning_rate": 0.001,
|
| 1760 |
+
"loss": 0.0763,
|
| 1761 |
+
"step": 464
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 16.21,
|
| 1765 |
+
"grad_norm": 0.20545266568660736,
|
| 1766 |
+
"learning_rate": 0.001,
|
| 1767 |
+
"loss": 0.0858,
|
| 1768 |
+
"step": 466
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 16.28,
|
| 1772 |
+
"grad_norm": 0.18228977918624878,
|
| 1773 |
+
"learning_rate": 0.001,
|
| 1774 |
+
"loss": 0.0802,
|
| 1775 |
+
"step": 468
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 16.35,
|
| 1779 |
+
"grad_norm": 0.17187969386577606,
|
| 1780 |
+
"learning_rate": 0.001,
|
| 1781 |
+
"loss": 0.0803,
|
| 1782 |
+
"step": 470
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 16.42,
|
| 1786 |
+
"grad_norm": 0.17641624808311462,
|
| 1787 |
+
"learning_rate": 0.001,
|
| 1788 |
+
"loss": 0.0796,
|
| 1789 |
+
"step": 472
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 16.49,
|
| 1793 |
+
"grad_norm": 0.19069235026836395,
|
| 1794 |
+
"learning_rate": 0.001,
|
| 1795 |
+
"loss": 0.0798,
|
| 1796 |
+
"step": 474
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 16.56,
|
| 1800 |
+
"grad_norm": 0.1968899667263031,
|
| 1801 |
+
"learning_rate": 0.001,
|
| 1802 |
+
"loss": 0.081,
|
| 1803 |
+
"step": 476
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 16.63,
|
| 1807 |
+
"grad_norm": 0.196567103266716,
|
| 1808 |
+
"learning_rate": 0.001,
|
| 1809 |
+
"loss": 0.0891,
|
| 1810 |
+
"step": 478
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 16.7,
|
| 1814 |
+
"grad_norm": 0.18010011315345764,
|
| 1815 |
+
"learning_rate": 0.001,
|
| 1816 |
+
"loss": 0.0795,
|
| 1817 |
+
"step": 480
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 16.77,
|
| 1821 |
+
"grad_norm": 0.31678447127342224,
|
| 1822 |
+
"learning_rate": 0.001,
|
| 1823 |
+
"loss": 0.0819,
|
| 1824 |
+
"step": 482
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 16.83,
|
| 1828 |
+
"grad_norm": 0.20116592943668365,
|
| 1829 |
+
"learning_rate": 0.001,
|
| 1830 |
+
"loss": 0.0873,
|
| 1831 |
+
"step": 484
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 16.9,
|
| 1835 |
+
"grad_norm": 0.19425255060195923,
|
| 1836 |
+
"learning_rate": 0.001,
|
| 1837 |
+
"loss": 0.0898,
|
| 1838 |
+
"step": 486
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 16.97,
|
| 1842 |
+
"grad_norm": 0.21240267157554626,
|
| 1843 |
+
"learning_rate": 0.001,
|
| 1844 |
+
"loss": 0.0942,
|
| 1845 |
+
"step": 488
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 16.97,
|
| 1849 |
+
"eval_loss": 4.70243501663208,
|
| 1850 |
+
"eval_runtime": 14.2157,
|
| 1851 |
+
"eval_samples_per_second": 23.777,
|
| 1852 |
+
"eval_steps_per_second": 3.025,
|
| 1853 |
+
"step": 488
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 17.04,
|
| 1857 |
+
"grad_norm": 0.16432495415210724,
|
| 1858 |
+
"learning_rate": 0.001,
|
| 1859 |
+
"loss": 0.0762,
|
| 1860 |
+
"step": 490
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 17.11,
|
| 1864 |
+
"grad_norm": 0.19645223021507263,
|
| 1865 |
+
"learning_rate": 0.001,
|
| 1866 |
+
"loss": 0.0638,
|
| 1867 |
+
"step": 492
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 17.18,
|
| 1871 |
+
"grad_norm": 0.1714513599872589,
|
| 1872 |
+
"learning_rate": 0.001,
|
| 1873 |
+
"loss": 0.0669,
|
| 1874 |
+
"step": 494
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 17.25,
|
| 1878 |
+
"grad_norm": 0.16747209429740906,
|
| 1879 |
+
"learning_rate": 0.001,
|
| 1880 |
+
"loss": 0.064,
|
| 1881 |
+
"step": 496
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 17.32,
|
| 1885 |
+
"grad_norm": 0.17507390677928925,
|
| 1886 |
+
"learning_rate": 0.001,
|
| 1887 |
+
"loss": 0.0675,
|
| 1888 |
+
"step": 498
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 17.39,
|
| 1892 |
+
"grad_norm": 0.1752365380525589,
|
| 1893 |
+
"learning_rate": 0.001,
|
| 1894 |
+
"loss": 0.0716,
|
| 1895 |
+
"step": 500
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 17.46,
|
| 1899 |
+
"grad_norm": 0.1760423481464386,
|
| 1900 |
+
"learning_rate": 0.001,
|
| 1901 |
+
"loss": 0.0736,
|
| 1902 |
+
"step": 502
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 17.53,
|
| 1906 |
+
"grad_norm": 0.17969931662082672,
|
| 1907 |
+
"learning_rate": 0.001,
|
| 1908 |
+
"loss": 0.0725,
|
| 1909 |
+
"step": 504
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 17.6,
|
| 1913 |
+
"grad_norm": 0.20572717487812042,
|
| 1914 |
+
"learning_rate": 0.001,
|
| 1915 |
+
"loss": 0.0751,
|
| 1916 |
+
"step": 506
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 17.67,
|
| 1920 |
+
"grad_norm": 0.20842702686786652,
|
| 1921 |
+
"learning_rate": 0.001,
|
| 1922 |
+
"loss": 0.0796,
|
| 1923 |
+
"step": 508
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 17.74,
|
| 1927 |
+
"grad_norm": 0.18587857484817505,
|
| 1928 |
+
"learning_rate": 0.001,
|
| 1929 |
+
"loss": 0.0757,
|
| 1930 |
+
"step": 510
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 17.81,
|
| 1934 |
+
"grad_norm": 0.17455922067165375,
|
| 1935 |
+
"learning_rate": 0.001,
|
| 1936 |
+
"loss": 0.0792,
|
| 1937 |
+
"step": 512
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 17.88,
|
| 1941 |
+
"grad_norm": 0.181496262550354,
|
| 1942 |
+
"learning_rate": 0.001,
|
| 1943 |
+
"loss": 0.0711,
|
| 1944 |
+
"step": 514
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 17.95,
|
| 1948 |
+
"grad_norm": 0.1918957382440567,
|
| 1949 |
+
"learning_rate": 0.001,
|
| 1950 |
+
"loss": 0.0765,
|
| 1951 |
+
"step": 516
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 17.98,
|
| 1955 |
+
"eval_loss": 4.348242282867432,
|
| 1956 |
+
"eval_runtime": 14.2133,
|
| 1957 |
+
"eval_samples_per_second": 23.78,
|
| 1958 |
+
"eval_steps_per_second": 3.025,
|
| 1959 |
+
"step": 517
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 18.02,
|
| 1963 |
+
"grad_norm": 0.23215439915657043,
|
| 1964 |
+
"learning_rate": 0.001,
|
| 1965 |
+
"loss": 0.0854,
|
| 1966 |
+
"step": 518
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 18.09,
|
| 1970 |
+
"grad_norm": 0.16362574696540833,
|
| 1971 |
+
"learning_rate": 0.001,
|
| 1972 |
+
"loss": 0.0636,
|
| 1973 |
+
"step": 520
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 18.16,
|
| 1977 |
+
"grad_norm": 0.18604953587055206,
|
| 1978 |
+
"learning_rate": 0.001,
|
| 1979 |
+
"loss": 0.0632,
|
| 1980 |
+
"step": 522
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 18.23,
|
| 1984 |
+
"grad_norm": 0.20401595532894135,
|
| 1985 |
+
"learning_rate": 0.001,
|
| 1986 |
+
"loss": 0.077,
|
| 1987 |
+
"step": 524
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 18.3,
|
| 1991 |
+
"grad_norm": 0.1713324785232544,
|
| 1992 |
+
"learning_rate": 0.001,
|
| 1993 |
+
"loss": 0.0743,
|
| 1994 |
+
"step": 526
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 18.37,
|
| 1998 |
+
"grad_norm": 0.1794094294309616,
|
| 1999 |
+
"learning_rate": 0.001,
|
| 2000 |
+
"loss": 0.0693,
|
| 2001 |
+
"step": 528
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 18.43,
|
| 2005 |
+
"grad_norm": 0.1892070323228836,
|
| 2006 |
+
"learning_rate": 0.001,
|
| 2007 |
+
"loss": 0.0663,
|
| 2008 |
+
"step": 530
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 18.5,
|
| 2012 |
+
"grad_norm": 0.1843532919883728,
|
| 2013 |
+
"learning_rate": 0.001,
|
| 2014 |
+
"loss": 0.0721,
|
| 2015 |
+
"step": 532
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 18.57,
|
| 2019 |
+
"grad_norm": 0.17645902931690216,
|
| 2020 |
+
"learning_rate": 0.001,
|
| 2021 |
+
"loss": 0.0746,
|
| 2022 |
+
"step": 534
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 18.64,
|
| 2026 |
+
"grad_norm": 0.16752366721630096,
|
| 2027 |
+
"learning_rate": 0.001,
|
| 2028 |
+
"loss": 0.0712,
|
| 2029 |
+
"step": 536
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 18.71,
|
| 2033 |
+
"grad_norm": 0.16078142821788788,
|
| 2034 |
+
"learning_rate": 0.001,
|
| 2035 |
+
"loss": 0.0729,
|
| 2036 |
+
"step": 538
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 18.78,
|
| 2040 |
+
"grad_norm": 0.1775059700012207,
|
| 2041 |
+
"learning_rate": 0.001,
|
| 2042 |
+
"loss": 0.0745,
|
| 2043 |
+
"step": 540
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 18.85,
|
| 2047 |
+
"grad_norm": 0.18710005283355713,
|
| 2048 |
+
"learning_rate": 0.001,
|
| 2049 |
+
"loss": 0.0743,
|
| 2050 |
+
"step": 542
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 18.92,
|
| 2054 |
+
"grad_norm": 0.20217396318912506,
|
| 2055 |
+
"learning_rate": 0.001,
|
| 2056 |
+
"loss": 0.0761,
|
| 2057 |
+
"step": 544
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 18.99,
|
| 2061 |
+
"grad_norm": 0.19417209923267365,
|
| 2062 |
+
"learning_rate": 0.001,
|
| 2063 |
+
"loss": 0.0799,
|
| 2064 |
+
"step": 546
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 18.99,
|
| 2068 |
+
"eval_loss": 4.538631916046143,
|
| 2069 |
+
"eval_runtime": 14.2104,
|
| 2070 |
+
"eval_samples_per_second": 23.785,
|
| 2071 |
+
"eval_steps_per_second": 3.026,
|
| 2072 |
+
"step": 546
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 19.06,
|
| 2076 |
+
"grad_norm": 0.16049686074256897,
|
| 2077 |
+
"learning_rate": 0.001,
|
| 2078 |
+
"loss": 0.0633,
|
| 2079 |
+
"step": 548
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 19.13,
|
| 2083 |
+
"grad_norm": 0.15315844118595123,
|
| 2084 |
+
"learning_rate": 0.001,
|
| 2085 |
+
"loss": 0.0614,
|
| 2086 |
+
"step": 550
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 19.2,
|
| 2090 |
+
"grad_norm": 0.16441303491592407,
|
| 2091 |
+
"learning_rate": 0.001,
|
| 2092 |
+
"loss": 0.0658,
|
| 2093 |
+
"step": 552
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 19.27,
|
| 2097 |
+
"grad_norm": 0.16347071528434753,
|
| 2098 |
+
"learning_rate": 0.001,
|
| 2099 |
+
"loss": 0.0551,
|
| 2100 |
+
"step": 554
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 19.34,
|
| 2104 |
+
"grad_norm": 0.16774040460586548,
|
| 2105 |
+
"learning_rate": 0.001,
|
| 2106 |
+
"loss": 0.0698,
|
| 2107 |
+
"step": 556
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 19.41,
|
| 2111 |
+
"grad_norm": 0.15695886313915253,
|
| 2112 |
+
"learning_rate": 0.001,
|
| 2113 |
+
"loss": 0.063,
|
| 2114 |
+
"step": 558
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 19.48,
|
| 2118 |
+
"grad_norm": 0.1639881581068039,
|
| 2119 |
+
"learning_rate": 0.001,
|
| 2120 |
+
"loss": 0.0649,
|
| 2121 |
+
"step": 560
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 19.55,
|
| 2125 |
+
"grad_norm": 0.1653515100479126,
|
| 2126 |
+
"learning_rate": 0.001,
|
| 2127 |
+
"loss": 0.0592,
|
| 2128 |
+
"step": 562
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 19.62,
|
| 2132 |
+
"grad_norm": 0.1697942465543747,
|
| 2133 |
+
"learning_rate": 0.001,
|
| 2134 |
+
"loss": 0.0619,
|
| 2135 |
+
"step": 564
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 19.69,
|
| 2139 |
+
"grad_norm": 0.18345074355602264,
|
| 2140 |
+
"learning_rate": 0.001,
|
| 2141 |
+
"loss": 0.0674,
|
| 2142 |
+
"step": 566
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 19.76,
|
| 2146 |
+
"grad_norm": 0.17060339450836182,
|
| 2147 |
+
"learning_rate": 0.001,
|
| 2148 |
+
"loss": 0.0603,
|
| 2149 |
+
"step": 568
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 19.83,
|
| 2153 |
+
"grad_norm": 0.18731752038002014,
|
| 2154 |
+
"learning_rate": 0.001,
|
| 2155 |
+
"loss": 0.0668,
|
| 2156 |
+
"step": 570
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 19.9,
|
| 2160 |
+
"grad_norm": 0.1727019101381302,
|
| 2161 |
+
"learning_rate": 0.001,
|
| 2162 |
+
"loss": 0.0717,
|
| 2163 |
+
"step": 572
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 19.97,
|
| 2167 |
+
"grad_norm": 0.18209876120090485,
|
| 2168 |
+
"learning_rate": 0.001,
|
| 2169 |
+
"loss": 0.073,
|
| 2170 |
+
"step": 574
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 20.0,
|
| 2174 |
+
"eval_loss": 4.588906764984131,
|
| 2175 |
+
"eval_runtime": 14.2111,
|
| 2176 |
+
"eval_samples_per_second": 23.784,
|
| 2177 |
+
"eval_steps_per_second": 3.026,
|
| 2178 |
+
"step": 575
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 20.03,
|
| 2182 |
+
"grad_norm": 0.15836042165756226,
|
| 2183 |
+
"learning_rate": 0.001,
|
| 2184 |
+
"loss": 0.0655,
|
| 2185 |
+
"step": 576
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 20.1,
|
| 2189 |
+
"grad_norm": 0.1645493507385254,
|
| 2190 |
+
"learning_rate": 0.001,
|
| 2191 |
+
"loss": 0.0609,
|
| 2192 |
+
"step": 578
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 20.17,
|
| 2196 |
+
"grad_norm": 0.14691436290740967,
|
| 2197 |
+
"learning_rate": 0.001,
|
| 2198 |
+
"loss": 0.0533,
|
| 2199 |
+
"step": 580
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 20.24,
|
| 2203 |
+
"grad_norm": 0.15472760796546936,
|
| 2204 |
+
"learning_rate": 0.001,
|
| 2205 |
+
"loss": 0.0596,
|
| 2206 |
+
"step": 582
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 20.31,
|
| 2210 |
+
"grad_norm": 0.14597873389720917,
|
| 2211 |
+
"learning_rate": 0.001,
|
| 2212 |
+
"loss": 0.0548,
|
| 2213 |
+
"step": 584
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 20.38,
|
| 2217 |
+
"grad_norm": 0.14974112808704376,
|
| 2218 |
+
"learning_rate": 0.001,
|
| 2219 |
+
"loss": 0.057,
|
| 2220 |
+
"step": 586
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 20.45,
|
| 2224 |
+
"grad_norm": 0.16359882056713104,
|
| 2225 |
+
"learning_rate": 0.001,
|
| 2226 |
+
"loss": 0.0597,
|
| 2227 |
+
"step": 588
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 20.52,
|
| 2231 |
+
"grad_norm": 0.15623408555984497,
|
| 2232 |
+
"learning_rate": 0.001,
|
| 2233 |
+
"loss": 0.0581,
|
| 2234 |
+
"step": 590
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 20.59,
|
| 2238 |
+
"grad_norm": 0.16072715818881989,
|
| 2239 |
+
"learning_rate": 0.001,
|
| 2240 |
+
"loss": 0.0577,
|
| 2241 |
+
"step": 592
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 20.66,
|
| 2245 |
+
"grad_norm": 0.17041383683681488,
|
| 2246 |
+
"learning_rate": 0.001,
|
| 2247 |
+
"loss": 0.0601,
|
| 2248 |
+
"step": 594
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 20.73,
|
| 2252 |
+
"grad_norm": 0.182576522231102,
|
| 2253 |
+
"learning_rate": 0.001,
|
| 2254 |
+
"loss": 0.0614,
|
| 2255 |
+
"step": 596
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 20.8,
|
| 2259 |
+
"grad_norm": 0.17546887695789337,
|
| 2260 |
+
"learning_rate": 0.001,
|
| 2261 |
+
"loss": 0.0658,
|
| 2262 |
+
"step": 598
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 20.87,
|
| 2266 |
+
"grad_norm": 0.19615040719509125,
|
| 2267 |
+
"learning_rate": 0.001,
|
| 2268 |
+
"loss": 0.061,
|
| 2269 |
+
"step": 600
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 20.94,
|
| 2273 |
+
"grad_norm": 0.2895524799823761,
|
| 2274 |
+
"learning_rate": 0.001,
|
| 2275 |
+
"loss": 0.0825,
|
| 2276 |
+
"step": 602
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 20.97,
|
| 2280 |
+
"eval_loss": 4.681673049926758,
|
| 2281 |
+
"eval_runtime": 14.2047,
|
| 2282 |
+
"eval_samples_per_second": 23.795,
|
| 2283 |
+
"eval_steps_per_second": 3.027,
|
| 2284 |
+
"step": 603
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 21.01,
|
| 2288 |
+
"grad_norm": 0.3042687177658081,
|
| 2289 |
+
"learning_rate": 0.001,
|
| 2290 |
+
"loss": 0.102,
|
| 2291 |
+
"step": 604
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 21.08,
|
| 2295 |
+
"grad_norm": 0.14991188049316406,
|
| 2296 |
+
"learning_rate": 0.001,
|
| 2297 |
+
"loss": 0.0585,
|
| 2298 |
+
"step": 606
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 21.15,
|
| 2302 |
+
"grad_norm": 0.13689693808555603,
|
| 2303 |
+
"learning_rate": 0.001,
|
| 2304 |
+
"loss": 0.0546,
|
| 2305 |
+
"step": 608
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 21.22,
|
| 2309 |
+
"grad_norm": 0.14480513334274292,
|
| 2310 |
+
"learning_rate": 0.001,
|
| 2311 |
+
"loss": 0.0533,
|
| 2312 |
+
"step": 610
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 21.29,
|
| 2316 |
+
"grad_norm": 0.148710235953331,
|
| 2317 |
+
"learning_rate": 0.001,
|
| 2318 |
+
"loss": 0.0543,
|
| 2319 |
+
"step": 612
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 21.36,
|
| 2323 |
+
"grad_norm": 0.14531894028186798,
|
| 2324 |
+
"learning_rate": 0.001,
|
| 2325 |
+
"loss": 0.0528,
|
| 2326 |
+
"step": 614
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 21.43,
|
| 2330 |
+
"grad_norm": 0.14758220314979553,
|
| 2331 |
+
"learning_rate": 0.001,
|
| 2332 |
+
"loss": 0.0536,
|
| 2333 |
+
"step": 616
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 21.5,
|
| 2337 |
+
"grad_norm": 0.1422356367111206,
|
| 2338 |
+
"learning_rate": 0.001,
|
| 2339 |
+
"loss": 0.0537,
|
| 2340 |
+
"step": 618
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 21.57,
|
| 2344 |
+
"grad_norm": 0.16927896440029144,
|
| 2345 |
+
"learning_rate": 0.001,
|
| 2346 |
+
"loss": 0.0551,
|
| 2347 |
+
"step": 620
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 21.63,
|
| 2351 |
+
"grad_norm": 0.18368251621723175,
|
| 2352 |
+
"learning_rate": 0.001,
|
| 2353 |
+
"loss": 0.0589,
|
| 2354 |
+
"step": 622
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 21.7,
|
| 2358 |
+
"grad_norm": 0.16162264347076416,
|
| 2359 |
+
"learning_rate": 0.001,
|
| 2360 |
+
"loss": 0.0639,
|
| 2361 |
+
"step": 624
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 21.77,
|
| 2365 |
+
"grad_norm": 0.1610013097524643,
|
| 2366 |
+
"learning_rate": 0.001,
|
| 2367 |
+
"loss": 0.0604,
|
| 2368 |
+
"step": 626
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 21.84,
|
| 2372 |
+
"grad_norm": 0.15656866133213043,
|
| 2373 |
+
"learning_rate": 0.001,
|
| 2374 |
+
"loss": 0.0603,
|
| 2375 |
+
"step": 628
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 21.91,
|
| 2379 |
+
"grad_norm": 0.15664397180080414,
|
| 2380 |
+
"learning_rate": 0.001,
|
| 2381 |
+
"loss": 0.0589,
|
| 2382 |
+
"step": 630
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 21.98,
|
| 2386 |
+
"grad_norm": 0.1484573483467102,
|
| 2387 |
+
"learning_rate": 0.001,
|
| 2388 |
+
"loss": 0.0616,
|
| 2389 |
+
"step": 632
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 21.98,
|
| 2393 |
+
"eval_loss": 5.026252746582031,
|
| 2394 |
+
"eval_runtime": 14.2151,
|
| 2395 |
+
"eval_samples_per_second": 23.778,
|
| 2396 |
+
"eval_steps_per_second": 3.025,
|
| 2397 |
+
"step": 632
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 22.05,
|
| 2401 |
+
"grad_norm": 0.16003428399562836,
|
| 2402 |
+
"learning_rate": 0.001,
|
| 2403 |
+
"loss": 0.0579,
|
| 2404 |
+
"step": 634
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 22.12,
|
| 2408 |
+
"grad_norm": 0.14746786653995514,
|
| 2409 |
+
"learning_rate": 0.001,
|
| 2410 |
+
"loss": 0.0503,
|
| 2411 |
+
"step": 636
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 22.19,
|
| 2415 |
+
"grad_norm": 0.1616966724395752,
|
| 2416 |
+
"learning_rate": 0.001,
|
| 2417 |
+
"loss": 0.0518,
|
| 2418 |
+
"step": 638
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 22.26,
|
| 2422 |
+
"grad_norm": 0.14375324547290802,
|
| 2423 |
+
"learning_rate": 0.001,
|
| 2424 |
+
"loss": 0.0501,
|
| 2425 |
+
"step": 640
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 22.33,
|
| 2429 |
+
"grad_norm": 0.14273105561733246,
|
| 2430 |
+
"learning_rate": 0.001,
|
| 2431 |
+
"loss": 0.0543,
|
| 2432 |
+
"step": 642
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 22.4,
|
| 2436 |
+
"grad_norm": 0.14281895756721497,
|
| 2437 |
+
"learning_rate": 0.001,
|
| 2438 |
+
"loss": 0.0572,
|
| 2439 |
+
"step": 644
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 22.47,
|
| 2443 |
+
"grad_norm": 0.1508861482143402,
|
| 2444 |
+
"learning_rate": 0.001,
|
| 2445 |
+
"loss": 0.0538,
|
| 2446 |
+
"step": 646
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 22.54,
|
| 2450 |
+
"grad_norm": 0.1677195429801941,
|
| 2451 |
+
"learning_rate": 0.001,
|
| 2452 |
+
"loss": 0.052,
|
| 2453 |
+
"step": 648
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 22.61,
|
| 2457 |
+
"grad_norm": 0.16616381704807281,
|
| 2458 |
+
"learning_rate": 0.001,
|
| 2459 |
+
"loss": 0.0569,
|
| 2460 |
+
"step": 650
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 22.68,
|
| 2464 |
+
"grad_norm": 0.159612774848938,
|
| 2465 |
+
"learning_rate": 0.001,
|
| 2466 |
+
"loss": 0.0585,
|
| 2467 |
+
"step": 652
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 22.75,
|
| 2471 |
+
"grad_norm": 0.14692574739456177,
|
| 2472 |
+
"learning_rate": 0.001,
|
| 2473 |
+
"loss": 0.0559,
|
| 2474 |
+
"step": 654
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 22.82,
|
| 2478 |
+
"grad_norm": 0.14404280483722687,
|
| 2479 |
+
"learning_rate": 0.001,
|
| 2480 |
+
"loss": 0.0614,
|
| 2481 |
+
"step": 656
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 22.89,
|
| 2485 |
+
"grad_norm": 0.16935443878173828,
|
| 2486 |
+
"learning_rate": 0.001,
|
| 2487 |
+
"loss": 0.0531,
|
| 2488 |
+
"step": 658
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 22.96,
|
| 2492 |
+
"grad_norm": 0.16444338858127594,
|
| 2493 |
+
"learning_rate": 0.001,
|
| 2494 |
+
"loss": 0.0677,
|
| 2495 |
+
"step": 660
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 22.99,
|
| 2499 |
+
"eval_loss": 4.580361843109131,
|
| 2500 |
+
"eval_runtime": 14.2193,
|
| 2501 |
+
"eval_samples_per_second": 23.77,
|
| 2502 |
+
"eval_steps_per_second": 3.024,
|
| 2503 |
+
"step": 661
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 23.03,
|
| 2507 |
+
"grad_norm": 0.1523355394601822,
|
| 2508 |
+
"learning_rate": 0.001,
|
| 2509 |
+
"loss": 0.056,
|
| 2510 |
+
"step": 662
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 23.1,
|
| 2514 |
+
"grad_norm": 0.13197599351406097,
|
| 2515 |
+
"learning_rate": 0.001,
|
| 2516 |
+
"loss": 0.0505,
|
| 2517 |
+
"step": 664
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 23.17,
|
| 2521 |
+
"grad_norm": 0.1436968594789505,
|
| 2522 |
+
"learning_rate": 0.001,
|
| 2523 |
+
"loss": 0.0459,
|
| 2524 |
+
"step": 666
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 23.23,
|
| 2528 |
+
"grad_norm": 0.15447324514389038,
|
| 2529 |
+
"learning_rate": 0.001,
|
| 2530 |
+
"loss": 0.0484,
|
| 2531 |
+
"step": 668
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 23.3,
|
| 2535 |
+
"grad_norm": 0.1536988466978073,
|
| 2536 |
+
"learning_rate": 0.001,
|
| 2537 |
+
"loss": 0.0523,
|
| 2538 |
+
"step": 670
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 23.37,
|
| 2542 |
+
"grad_norm": 0.14964550733566284,
|
| 2543 |
+
"learning_rate": 0.001,
|
| 2544 |
+
"loss": 0.0523,
|
| 2545 |
+
"step": 672
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 23.44,
|
| 2549 |
+
"grad_norm": 0.14874936640262604,
|
| 2550 |
+
"learning_rate": 0.001,
|
| 2551 |
+
"loss": 0.053,
|
| 2552 |
+
"step": 674
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 23.51,
|
| 2556 |
+
"grad_norm": 0.14731401205062866,
|
| 2557 |
+
"learning_rate": 0.001,
|
| 2558 |
+
"loss": 0.0534,
|
| 2559 |
+
"step": 676
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 23.58,
|
| 2563 |
+
"grad_norm": 0.15508398413658142,
|
| 2564 |
+
"learning_rate": 0.001,
|
| 2565 |
+
"loss": 0.0538,
|
| 2566 |
+
"step": 678
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 23.65,
|
| 2570 |
+
"grad_norm": 0.16380004584789276,
|
| 2571 |
+
"learning_rate": 0.001,
|
| 2572 |
+
"loss": 0.0539,
|
| 2573 |
+
"step": 680
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 23.72,
|
| 2577 |
+
"grad_norm": 0.38461077213287354,
|
| 2578 |
+
"learning_rate": 0.001,
|
| 2579 |
+
"loss": 0.0609,
|
| 2580 |
+
"step": 682
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 23.79,
|
| 2584 |
+
"grad_norm": 0.1609809696674347,
|
| 2585 |
+
"learning_rate": 0.001,
|
| 2586 |
+
"loss": 0.0553,
|
| 2587 |
+
"step": 684
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 23.86,
|
| 2591 |
+
"grad_norm": 0.1482686847448349,
|
| 2592 |
+
"learning_rate": 0.001,
|
| 2593 |
+
"loss": 0.0576,
|
| 2594 |
+
"step": 686
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 23.93,
|
| 2598 |
+
"grad_norm": 0.16069763898849487,
|
| 2599 |
+
"learning_rate": 0.001,
|
| 2600 |
+
"loss": 0.0569,
|
| 2601 |
+
"step": 688
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 24.0,
|
| 2605 |
+
"grad_norm": 0.16612602770328522,
|
| 2606 |
+
"learning_rate": 0.001,
|
| 2607 |
+
"loss": 0.0571,
|
| 2608 |
+
"step": 690
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 24.0,
|
| 2612 |
+
"eval_loss": 4.839931488037109,
|
| 2613 |
+
"eval_runtime": 14.2054,
|
| 2614 |
+
"eval_samples_per_second": 23.794,
|
| 2615 |
+
"eval_steps_per_second": 3.027,
|
| 2616 |
+
"step": 690
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 24.07,
|
| 2620 |
+
"grad_norm": 0.14117293059825897,
|
| 2621 |
+
"learning_rate": 0.001,
|
| 2622 |
+
"loss": 0.0473,
|
| 2623 |
+
"step": 692
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 24.14,
|
| 2627 |
+
"grad_norm": 0.13660480082035065,
|
| 2628 |
+
"learning_rate": 0.001,
|
| 2629 |
+
"loss": 0.0462,
|
| 2630 |
+
"step": 694
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 24.21,
|
| 2634 |
+
"grad_norm": 0.13649675250053406,
|
| 2635 |
+
"learning_rate": 0.001,
|
| 2636 |
+
"loss": 0.0436,
|
| 2637 |
+
"step": 696
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 24.28,
|
| 2641 |
+
"grad_norm": 0.13140219449996948,
|
| 2642 |
+
"learning_rate": 0.001,
|
| 2643 |
+
"loss": 0.0496,
|
| 2644 |
+
"step": 698
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 24.35,
|
| 2648 |
+
"grad_norm": 0.13666340708732605,
|
| 2649 |
+
"learning_rate": 0.001,
|
| 2650 |
+
"loss": 0.0453,
|
| 2651 |
+
"step": 700
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 24.42,
|
| 2655 |
+
"grad_norm": 0.14786505699157715,
|
| 2656 |
+
"learning_rate": 0.001,
|
| 2657 |
+
"loss": 0.0458,
|
| 2658 |
+
"step": 702
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 24.49,
|
| 2662 |
+
"grad_norm": 0.14177779853343964,
|
| 2663 |
+
"learning_rate": 0.001,
|
| 2664 |
+
"loss": 0.0454,
|
| 2665 |
+
"step": 704
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 24.56,
|
| 2669 |
+
"grad_norm": 0.15088017284870148,
|
| 2670 |
+
"learning_rate": 0.001,
|
| 2671 |
+
"loss": 0.0486,
|
| 2672 |
+
"step": 706
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 24.63,
|
| 2676 |
+
"grad_norm": 0.1343182623386383,
|
| 2677 |
+
"learning_rate": 0.001,
|
| 2678 |
+
"loss": 0.0466,
|
| 2679 |
+
"step": 708
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 24.7,
|
| 2683 |
+
"grad_norm": 0.14765577018260956,
|
| 2684 |
+
"learning_rate": 0.001,
|
| 2685 |
+
"loss": 0.051,
|
| 2686 |
+
"step": 710
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 24.77,
|
| 2690 |
+
"grad_norm": 0.1491304486989975,
|
| 2691 |
+
"learning_rate": 0.001,
|
| 2692 |
+
"loss": 0.0497,
|
| 2693 |
+
"step": 712
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 24.83,
|
| 2697 |
+
"grad_norm": 0.1482744961977005,
|
| 2698 |
+
"learning_rate": 0.001,
|
| 2699 |
+
"loss": 0.0525,
|
| 2700 |
+
"step": 714
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 24.9,
|
| 2704 |
+
"grad_norm": 0.15700866281986237,
|
| 2705 |
+
"learning_rate": 0.001,
|
| 2706 |
+
"loss": 0.0497,
|
| 2707 |
+
"step": 716
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 24.97,
|
| 2711 |
+
"grad_norm": 0.15650177001953125,
|
| 2712 |
+
"learning_rate": 0.001,
|
| 2713 |
+
"loss": 0.0525,
|
| 2714 |
+
"step": 718
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 24.97,
|
| 2718 |
+
"eval_loss": 4.934982776641846,
|
| 2719 |
+
"eval_runtime": 14.2072,
|
| 2720 |
+
"eval_samples_per_second": 23.791,
|
| 2721 |
+
"eval_steps_per_second": 3.027,
|
| 2722 |
+
"step": 718
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 25.04,
|
| 2726 |
+
"grad_norm": 0.15254813432693481,
|
| 2727 |
+
"learning_rate": 0.001,
|
| 2728 |
+
"loss": 0.0501,
|
| 2729 |
+
"step": 720
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 25.11,
|
| 2733 |
+
"grad_norm": 0.1353287696838379,
|
| 2734 |
+
"learning_rate": 0.001,
|
| 2735 |
+
"loss": 0.0471,
|
| 2736 |
+
"step": 722
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 25.18,
|
| 2740 |
+
"grad_norm": 0.155950665473938,
|
| 2741 |
+
"learning_rate": 0.001,
|
| 2742 |
+
"loss": 0.0478,
|
| 2743 |
+
"step": 724
|
| 2744 |
+
},
|
| 2745 |
+
{
|
| 2746 |
+
"epoch": 25.25,
|
| 2747 |
+
"grad_norm": 0.14119938015937805,
|
| 2748 |
+
"learning_rate": 0.001,
|
| 2749 |
+
"loss": 0.0484,
|
| 2750 |
+
"step": 726
|
| 2751 |
+
},
|
| 2752 |
+
{
|
| 2753 |
+
"epoch": 25.32,
|
| 2754 |
+
"grad_norm": 0.14679110050201416,
|
| 2755 |
+
"learning_rate": 0.001,
|
| 2756 |
+
"loss": 0.0432,
|
| 2757 |
+
"step": 728
|
| 2758 |
+
},
|
| 2759 |
+
{
|
| 2760 |
+
"epoch": 25.39,
|
| 2761 |
+
"grad_norm": 0.13392671942710876,
|
| 2762 |
+
"learning_rate": 0.001,
|
| 2763 |
+
"loss": 0.0486,
|
| 2764 |
+
"step": 730
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"epoch": 25.46,
|
| 2768 |
+
"grad_norm": 0.13733159005641937,
|
| 2769 |
+
"learning_rate": 0.001,
|
| 2770 |
+
"loss": 0.0451,
|
| 2771 |
+
"step": 732
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 25.53,
|
| 2775 |
+
"grad_norm": 0.13778570294380188,
|
| 2776 |
+
"learning_rate": 0.001,
|
| 2777 |
+
"loss": 0.0452,
|
| 2778 |
+
"step": 734
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 25.6,
|
| 2782 |
+
"grad_norm": 0.15224093198776245,
|
| 2783 |
+
"learning_rate": 0.001,
|
| 2784 |
+
"loss": 0.0469,
|
| 2785 |
+
"step": 736
|
| 2786 |
+
},
|
| 2787 |
+
{
|
| 2788 |
+
"epoch": 25.67,
|
| 2789 |
+
"grad_norm": 0.15857195854187012,
|
| 2790 |
+
"learning_rate": 0.001,
|
| 2791 |
+
"loss": 0.0489,
|
| 2792 |
+
"step": 738
|
| 2793 |
+
},
|
| 2794 |
+
{
|
| 2795 |
+
"epoch": 25.74,
|
| 2796 |
+
"grad_norm": 0.15234865248203278,
|
| 2797 |
+
"learning_rate": 0.001,
|
| 2798 |
+
"loss": 0.0536,
|
| 2799 |
+
"step": 740
|
| 2800 |
+
},
|
| 2801 |
+
{
|
| 2802 |
+
"epoch": 25.81,
|
| 2803 |
+
"grad_norm": 0.17310313880443573,
|
| 2804 |
+
"learning_rate": 0.001,
|
| 2805 |
+
"loss": 0.0518,
|
| 2806 |
+
"step": 742
|
| 2807 |
+
},
|
| 2808 |
+
{
|
| 2809 |
+
"epoch": 25.88,
|
| 2810 |
+
"grad_norm": 0.23332546651363373,
|
| 2811 |
+
"learning_rate": 0.001,
|
| 2812 |
+
"loss": 0.0601,
|
| 2813 |
+
"step": 744
|
| 2814 |
+
},
|
| 2815 |
+
{
|
| 2816 |
+
"epoch": 25.95,
|
| 2817 |
+
"grad_norm": 0.22406938672065735,
|
| 2818 |
+
"learning_rate": 0.001,
|
| 2819 |
+
"loss": 0.081,
|
| 2820 |
+
"step": 746
|
| 2821 |
+
},
|
| 2822 |
+
{
|
| 2823 |
+
"epoch": 25.98,
|
| 2824 |
+
"eval_loss": 4.690272331237793,
|
| 2825 |
+
"eval_runtime": 14.2141,
|
| 2826 |
+
"eval_samples_per_second": 23.779,
|
| 2827 |
+
"eval_steps_per_second": 3.025,
|
| 2828 |
+
"step": 747
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 26.02,
|
| 2832 |
+
"grad_norm": 0.14596611261367798,
|
| 2833 |
+
"learning_rate": 0.001,
|
| 2834 |
+
"loss": 0.0551,
|
| 2835 |
+
"step": 748
|
| 2836 |
+
},
|
| 2837 |
+
{
|
| 2838 |
+
"epoch": 26.09,
|
| 2839 |
+
"grad_norm": 0.11857102066278458,
|
| 2840 |
+
"learning_rate": 0.001,
|
| 2841 |
+
"loss": 0.0398,
|
| 2842 |
+
"step": 750
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 26.16,
|
| 2846 |
+
"grad_norm": 0.13823005557060242,
|
| 2847 |
+
"learning_rate": 0.001,
|
| 2848 |
+
"loss": 0.0424,
|
| 2849 |
+
"step": 752
|
| 2850 |
+
},
|
| 2851 |
+
{
|
| 2852 |
+
"epoch": 26.23,
|
| 2853 |
+
"grad_norm": 0.12586718797683716,
|
| 2854 |
+
"learning_rate": 0.001,
|
| 2855 |
+
"loss": 0.0422,
|
| 2856 |
+
"step": 754
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"epoch": 26.3,
|
| 2860 |
+
"grad_norm": 0.12402280420064926,
|
| 2861 |
+
"learning_rate": 0.001,
|
| 2862 |
+
"loss": 0.04,
|
| 2863 |
+
"step": 756
|
| 2864 |
+
},
|
| 2865 |
+
{
|
| 2866 |
+
"epoch": 26.37,
|
| 2867 |
+
"grad_norm": 0.1546175628900528,
|
| 2868 |
+
"learning_rate": 0.001,
|
| 2869 |
+
"loss": 0.0434,
|
| 2870 |
+
"step": 758
|
| 2871 |
+
},
|
| 2872 |
+
{
|
| 2873 |
+
"epoch": 26.43,
|
| 2874 |
+
"grad_norm": 0.15831632912158966,
|
| 2875 |
+
"learning_rate": 0.001,
|
| 2876 |
+
"loss": 0.041,
|
| 2877 |
+
"step": 760
|
| 2878 |
+
},
|
| 2879 |
+
{
|
| 2880 |
+
"epoch": 26.5,
|
| 2881 |
+
"grad_norm": 0.1380414068698883,
|
| 2882 |
+
"learning_rate": 0.001,
|
| 2883 |
+
"loss": 0.0411,
|
| 2884 |
+
"step": 762
|
| 2885 |
+
},
|
| 2886 |
+
{
|
| 2887 |
+
"epoch": 26.57,
|
| 2888 |
+
"grad_norm": 0.13040709495544434,
|
| 2889 |
+
"learning_rate": 0.001,
|
| 2890 |
+
"loss": 0.0412,
|
| 2891 |
+
"step": 764
|
| 2892 |
+
},
|
| 2893 |
+
{
|
| 2894 |
+
"epoch": 26.64,
|
| 2895 |
+
"grad_norm": 0.14285586774349213,
|
| 2896 |
+
"learning_rate": 0.001,
|
| 2897 |
+
"loss": 0.0426,
|
| 2898 |
+
"step": 766
|
| 2899 |
+
},
|
| 2900 |
+
{
|
| 2901 |
+
"epoch": 26.71,
|
| 2902 |
+
"grad_norm": 0.14807988703250885,
|
| 2903 |
+
"learning_rate": 0.001,
|
| 2904 |
+
"loss": 0.045,
|
| 2905 |
+
"step": 768
|
| 2906 |
+
},
|
| 2907 |
+
{
|
| 2908 |
+
"epoch": 26.78,
|
| 2909 |
+
"grad_norm": 0.14204862713813782,
|
| 2910 |
+
"learning_rate": 0.001,
|
| 2911 |
+
"loss": 0.0517,
|
| 2912 |
+
"step": 770
|
| 2913 |
+
},
|
| 2914 |
+
{
|
| 2915 |
+
"epoch": 26.85,
|
| 2916 |
+
"grad_norm": 0.13837137818336487,
|
| 2917 |
+
"learning_rate": 0.001,
|
| 2918 |
+
"loss": 0.0452,
|
| 2919 |
+
"step": 772
|
| 2920 |
+
},
|
| 2921 |
+
{
|
| 2922 |
+
"epoch": 26.92,
|
| 2923 |
+
"grad_norm": 0.14084497094154358,
|
| 2924 |
+
"learning_rate": 0.001,
|
| 2925 |
+
"loss": 0.0475,
|
| 2926 |
+
"step": 774
|
| 2927 |
+
},
|
| 2928 |
+
{
|
| 2929 |
+
"epoch": 26.99,
|
| 2930 |
+
"grad_norm": 0.13428126275539398,
|
| 2931 |
+
"learning_rate": 0.001,
|
| 2932 |
+
"loss": 0.0505,
|
| 2933 |
+
"step": 776
|
| 2934 |
+
},
|
| 2935 |
+
{
|
| 2936 |
+
"epoch": 26.99,
|
| 2937 |
+
"eval_loss": 5.000490665435791,
|
| 2938 |
+
"eval_runtime": 14.2197,
|
| 2939 |
+
"eval_samples_per_second": 23.77,
|
| 2940 |
+
"eval_steps_per_second": 3.024,
|
| 2941 |
+
"step": 776
|
| 2942 |
+
},
|
| 2943 |
+
{
|
| 2944 |
+
"epoch": 27.06,
|
| 2945 |
+
"grad_norm": 0.1252642720937729,
|
| 2946 |
+
"learning_rate": 0.001,
|
| 2947 |
+
"loss": 0.0372,
|
| 2948 |
+
"step": 778
|
| 2949 |
+
},
|
| 2950 |
+
{
|
| 2951 |
+
"epoch": 27.13,
|
| 2952 |
+
"grad_norm": 0.1609179526567459,
|
| 2953 |
+
"learning_rate": 0.001,
|
| 2954 |
+
"loss": 0.0422,
|
| 2955 |
+
"step": 780
|
| 2956 |
+
},
|
| 2957 |
+
{
|
| 2958 |
+
"epoch": 27.2,
|
| 2959 |
+
"grad_norm": 0.15185533463954926,
|
| 2960 |
+
"learning_rate": 0.001,
|
| 2961 |
+
"loss": 0.0427,
|
| 2962 |
+
"step": 782
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"epoch": 27.27,
|
| 2966 |
+
"grad_norm": 0.1398567408323288,
|
| 2967 |
+
"learning_rate": 0.001,
|
| 2968 |
+
"loss": 0.0452,
|
| 2969 |
+
"step": 784
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 27.34,
|
| 2973 |
+
"grad_norm": 0.13516364991664886,
|
| 2974 |
+
"learning_rate": 0.001,
|
| 2975 |
+
"loss": 0.0431,
|
| 2976 |
+
"step": 786
|
| 2977 |
+
},
|
| 2978 |
+
{
|
| 2979 |
+
"epoch": 27.41,
|
| 2980 |
+
"grad_norm": 0.1388152688741684,
|
| 2981 |
+
"learning_rate": 0.001,
|
| 2982 |
+
"loss": 0.0424,
|
| 2983 |
+
"step": 788
|
| 2984 |
+
},
|
| 2985 |
+
{
|
| 2986 |
+
"epoch": 27.48,
|
| 2987 |
+
"grad_norm": 0.14043325185775757,
|
| 2988 |
+
"learning_rate": 0.001,
|
| 2989 |
+
"loss": 0.0428,
|
| 2990 |
+
"step": 790
|
| 2991 |
+
},
|
| 2992 |
+
{
|
| 2993 |
+
"epoch": 27.55,
|
| 2994 |
+
"grad_norm": 0.14878694713115692,
|
| 2995 |
+
"learning_rate": 0.001,
|
| 2996 |
+
"loss": 0.0473,
|
| 2997 |
+
"step": 792
|
| 2998 |
+
},
|
| 2999 |
+
{
|
| 3000 |
+
"epoch": 27.62,
|
| 3001 |
+
"grad_norm": 0.15647174417972565,
|
| 3002 |
+
"learning_rate": 0.001,
|
| 3003 |
+
"loss": 0.0448,
|
| 3004 |
+
"step": 794
|
| 3005 |
+
},
|
| 3006 |
+
{
|
| 3007 |
+
"epoch": 27.69,
|
| 3008 |
+
"grad_norm": 0.1453557014465332,
|
| 3009 |
+
"learning_rate": 0.001,
|
| 3010 |
+
"loss": 0.0441,
|
| 3011 |
+
"step": 796
|
| 3012 |
+
},
|
| 3013 |
+
{
|
| 3014 |
+
"epoch": 27.76,
|
| 3015 |
+
"grad_norm": 0.13821879029273987,
|
| 3016 |
+
"learning_rate": 0.001,
|
| 3017 |
+
"loss": 0.0473,
|
| 3018 |
+
"step": 798
|
| 3019 |
+
},
|
| 3020 |
+
{
|
| 3021 |
+
"epoch": 27.83,
|
| 3022 |
+
"grad_norm": 0.1780141144990921,
|
| 3023 |
+
"learning_rate": 0.001,
|
| 3024 |
+
"loss": 0.048,
|
| 3025 |
+
"step": 800
|
| 3026 |
+
},
|
| 3027 |
+
{
|
| 3028 |
+
"epoch": 27.9,
|
| 3029 |
+
"grad_norm": 0.16689875721931458,
|
| 3030 |
+
"learning_rate": 0.001,
|
| 3031 |
+
"loss": 0.0568,
|
| 3032 |
+
"step": 802
|
| 3033 |
+
},
|
| 3034 |
+
{
|
| 3035 |
+
"epoch": 27.97,
|
| 3036 |
+
"grad_norm": 0.16479206085205078,
|
| 3037 |
+
"learning_rate": 0.001,
|
| 3038 |
+
"loss": 0.0576,
|
| 3039 |
+
"step": 804
|
| 3040 |
+
},
|
| 3041 |
+
{
|
| 3042 |
+
"epoch": 28.0,
|
| 3043 |
+
"eval_loss": 5.019590377807617,
|
| 3044 |
+
"eval_runtime": 14.2181,
|
| 3045 |
+
"eval_samples_per_second": 23.772,
|
| 3046 |
+
"eval_steps_per_second": 3.024,
|
| 3047 |
+
"step": 805
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 28.03,
|
| 3051 |
+
"grad_norm": 0.17976313829421997,
|
| 3052 |
+
"learning_rate": 0.001,
|
| 3053 |
+
"loss": 0.0586,
|
| 3054 |
+
"step": 806
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 28.1,
|
| 3058 |
+
"grad_norm": 0.15373440086841583,
|
| 3059 |
+
"learning_rate": 0.001,
|
| 3060 |
+
"loss": 0.0479,
|
| 3061 |
+
"step": 808
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 28.17,
|
| 3065 |
+
"grad_norm": 0.13314908742904663,
|
| 3066 |
+
"learning_rate": 0.001,
|
| 3067 |
+
"loss": 0.044,
|
| 3068 |
+
"step": 810
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 28.24,
|
| 3072 |
+
"grad_norm": 0.12571971118450165,
|
| 3073 |
+
"learning_rate": 0.001,
|
| 3074 |
+
"loss": 0.0382,
|
| 3075 |
+
"step": 812
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 28.31,
|
| 3079 |
+
"grad_norm": 0.1362116038799286,
|
| 3080 |
+
"learning_rate": 0.001,
|
| 3081 |
+
"loss": 0.0412,
|
| 3082 |
+
"step": 814
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 28.38,
|
| 3086 |
+
"grad_norm": 0.1244189664721489,
|
| 3087 |
+
"learning_rate": 0.001,
|
| 3088 |
+
"loss": 0.0422,
|
| 3089 |
+
"step": 816
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 28.45,
|
| 3093 |
+
"grad_norm": 0.1321047842502594,
|
| 3094 |
+
"learning_rate": 0.001,
|
| 3095 |
+
"loss": 0.0419,
|
| 3096 |
+
"step": 818
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 28.52,
|
| 3100 |
+
"grad_norm": 0.14995931088924408,
|
| 3101 |
+
"learning_rate": 0.001,
|
| 3102 |
+
"loss": 0.0453,
|
| 3103 |
+
"step": 820
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 28.59,
|
| 3107 |
+
"grad_norm": 0.1319851577281952,
|
| 3108 |
+
"learning_rate": 0.001,
|
| 3109 |
+
"loss": 0.0428,
|
| 3110 |
+
"step": 822
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 28.66,
|
| 3114 |
+
"grad_norm": 0.14392071962356567,
|
| 3115 |
+
"learning_rate": 0.001,
|
| 3116 |
+
"loss": 0.0417,
|
| 3117 |
+
"step": 824
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 28.73,
|
| 3121 |
+
"grad_norm": 0.13848307728767395,
|
| 3122 |
+
"learning_rate": 0.001,
|
| 3123 |
+
"loss": 0.0445,
|
| 3124 |
+
"step": 826
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 28.8,
|
| 3128 |
+
"grad_norm": 0.1495485007762909,
|
| 3129 |
+
"learning_rate": 0.001,
|
| 3130 |
+
"loss": 0.0451,
|
| 3131 |
+
"step": 828
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 28.87,
|
| 3135 |
+
"grad_norm": 0.13798055052757263,
|
| 3136 |
+
"learning_rate": 0.001,
|
| 3137 |
+
"loss": 0.0456,
|
| 3138 |
+
"step": 830
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 28.94,
|
| 3142 |
+
"grad_norm": 0.1452612578868866,
|
| 3143 |
+
"learning_rate": 0.001,
|
| 3144 |
+
"loss": 0.0448,
|
| 3145 |
+
"step": 832
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 28.97,
|
| 3149 |
+
"eval_loss": 5.109987735748291,
|
| 3150 |
+
"eval_runtime": 14.2104,
|
| 3151 |
+
"eval_samples_per_second": 23.785,
|
| 3152 |
+
"eval_steps_per_second": 3.026,
|
| 3153 |
+
"step": 833
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 29.01,
|
| 3157 |
+
"grad_norm": 0.17332157492637634,
|
| 3158 |
+
"learning_rate": 0.001,
|
| 3159 |
+
"loss": 0.0438,
|
| 3160 |
+
"step": 834
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 29.08,
|
| 3164 |
+
"grad_norm": 0.12759405374526978,
|
| 3165 |
+
"learning_rate": 0.001,
|
| 3166 |
+
"loss": 0.0391,
|
| 3167 |
+
"step": 836
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 29.15,
|
| 3171 |
+
"grad_norm": 0.13717012107372284,
|
| 3172 |
+
"learning_rate": 0.001,
|
| 3173 |
+
"loss": 0.0392,
|
| 3174 |
+
"step": 838
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 29.22,
|
| 3178 |
+
"grad_norm": 0.12544453144073486,
|
| 3179 |
+
"learning_rate": 0.001,
|
| 3180 |
+
"loss": 0.0373,
|
| 3181 |
+
"step": 840
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 29.29,
|
| 3185 |
+
"grad_norm": 0.12997865676879883,
|
| 3186 |
+
"learning_rate": 0.001,
|
| 3187 |
+
"loss": 0.0403,
|
| 3188 |
+
"step": 842
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 29.36,
|
| 3192 |
+
"grad_norm": 0.1252465397119522,
|
| 3193 |
+
"learning_rate": 0.001,
|
| 3194 |
+
"loss": 0.0355,
|
| 3195 |
+
"step": 844
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 29.43,
|
| 3199 |
+
"grad_norm": 0.13186001777648926,
|
| 3200 |
+
"learning_rate": 0.001,
|
| 3201 |
+
"loss": 0.046,
|
| 3202 |
+
"step": 846
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 29.5,
|
| 3206 |
+
"grad_norm": 0.1309279203414917,
|
| 3207 |
+
"learning_rate": 0.001,
|
| 3208 |
+
"loss": 0.0423,
|
| 3209 |
+
"step": 848
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 29.57,
|
| 3213 |
+
"grad_norm": 0.13001705706119537,
|
| 3214 |
+
"learning_rate": 0.001,
|
| 3215 |
+
"loss": 0.0401,
|
| 3216 |
+
"step": 850
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 29.63,
|
| 3220 |
+
"grad_norm": 0.13039255142211914,
|
| 3221 |
+
"learning_rate": 0.001,
|
| 3222 |
+
"loss": 0.0396,
|
| 3223 |
+
"step": 852
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 29.7,
|
| 3227 |
+
"grad_norm": 0.13635265827178955,
|
| 3228 |
+
"learning_rate": 0.001,
|
| 3229 |
+
"loss": 0.0426,
|
| 3230 |
+
"step": 854
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 29.77,
|
| 3234 |
+
"grad_norm": 0.16096022725105286,
|
| 3235 |
+
"learning_rate": 0.001,
|
| 3236 |
+
"loss": 0.0444,
|
| 3237 |
+
"step": 856
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 29.84,
|
| 3241 |
+
"grad_norm": 0.12575644254684448,
|
| 3242 |
+
"learning_rate": 0.001,
|
| 3243 |
+
"loss": 0.0427,
|
| 3244 |
+
"step": 858
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 29.91,
|
| 3248 |
+
"grad_norm": 0.13646484911441803,
|
| 3249 |
+
"learning_rate": 0.001,
|
| 3250 |
+
"loss": 0.0442,
|
| 3251 |
+
"step": 860
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 29.98,
|
| 3255 |
+
"grad_norm": 0.13415445387363434,
|
| 3256 |
+
"learning_rate": 0.001,
|
| 3257 |
+
"loss": 0.0457,
|
| 3258 |
+
"step": 862
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 29.98,
|
| 3262 |
+
"eval_loss": 5.0008440017700195,
|
| 3263 |
+
"eval_runtime": 14.2056,
|
| 3264 |
+
"eval_samples_per_second": 23.793,
|
| 3265 |
+
"eval_steps_per_second": 3.027,
|
| 3266 |
+
"step": 862
|
| 3267 |
+
},
|
| 3268 |
+
{
|
| 3269 |
+
"epoch": 30.05,
|
| 3270 |
+
"grad_norm": 0.12284772843122482,
|
| 3271 |
+
"learning_rate": 0.001,
|
| 3272 |
+
"loss": 0.0331,
|
| 3273 |
+
"step": 864
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 30.12,
|
| 3277 |
+
"grad_norm": 0.12438986450433731,
|
| 3278 |
+
"learning_rate": 0.001,
|
| 3279 |
+
"loss": 0.0372,
|
| 3280 |
+
"step": 866
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"epoch": 30.19,
|
| 3284 |
+
"grad_norm": 0.12020161747932434,
|
| 3285 |
+
"learning_rate": 0.001,
|
| 3286 |
+
"loss": 0.0366,
|
| 3287 |
+
"step": 868
|
| 3288 |
+
},
|
| 3289 |
+
{
|
| 3290 |
+
"epoch": 30.26,
|
| 3291 |
+
"grad_norm": 0.12868288159370422,
|
| 3292 |
+
"learning_rate": 0.001,
|
| 3293 |
+
"loss": 0.0384,
|
| 3294 |
+
"step": 870
|
| 3295 |
+
},
|
| 3296 |
+
{
|
| 3297 |
+
"epoch": 30.33,
|
| 3298 |
+
"grad_norm": 0.1276613026857376,
|
| 3299 |
+
"learning_rate": 0.001,
|
| 3300 |
+
"loss": 0.0415,
|
| 3301 |
+
"step": 872
|
| 3302 |
+
},
|
| 3303 |
+
{
|
| 3304 |
+
"epoch": 30.4,
|
| 3305 |
+
"grad_norm": 0.13801202178001404,
|
| 3306 |
+
"learning_rate": 0.001,
|
| 3307 |
+
"loss": 0.0403,
|
| 3308 |
+
"step": 874
|
| 3309 |
+
},
|
| 3310 |
+
{
|
| 3311 |
+
"epoch": 30.47,
|
| 3312 |
+
"grad_norm": 0.1174420416355133,
|
| 3313 |
+
"learning_rate": 0.001,
|
| 3314 |
+
"loss": 0.0389,
|
| 3315 |
+
"step": 876
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 30.54,
|
| 3319 |
+
"grad_norm": 0.11008256673812866,
|
| 3320 |
+
"learning_rate": 0.001,
|
| 3321 |
+
"loss": 0.0342,
|
| 3322 |
+
"step": 878
|
| 3323 |
+
},
|
| 3324 |
+
{
|
| 3325 |
+
"epoch": 30.61,
|
| 3326 |
+
"grad_norm": 0.12188894301652908,
|
| 3327 |
+
"learning_rate": 0.001,
|
| 3328 |
+
"loss": 0.0382,
|
| 3329 |
+
"step": 880
|
| 3330 |
+
},
|
| 3331 |
+
{
|
| 3332 |
+
"epoch": 30.68,
|
| 3333 |
+
"grad_norm": 0.12297231703996658,
|
| 3334 |
+
"learning_rate": 0.001,
|
| 3335 |
+
"loss": 0.0382,
|
| 3336 |
+
"step": 882
|
| 3337 |
+
},
|
| 3338 |
+
{
|
| 3339 |
+
"epoch": 30.75,
|
| 3340 |
+
"grad_norm": 0.12277786433696747,
|
| 3341 |
+
"learning_rate": 0.001,
|
| 3342 |
+
"loss": 0.0381,
|
| 3343 |
+
"step": 884
|
| 3344 |
+
},
|
| 3345 |
+
{
|
| 3346 |
+
"epoch": 30.82,
|
| 3347 |
+
"grad_norm": 0.13909707963466644,
|
| 3348 |
+
"learning_rate": 0.001,
|
| 3349 |
+
"loss": 0.0385,
|
| 3350 |
+
"step": 886
|
| 3351 |
+
},
|
| 3352 |
+
{
|
| 3353 |
+
"epoch": 30.89,
|
| 3354 |
+
"grad_norm": 0.13425001502037048,
|
| 3355 |
+
"learning_rate": 0.001,
|
| 3356 |
+
"loss": 0.0424,
|
| 3357 |
+
"step": 888
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 30.96,
|
| 3361 |
+
"grad_norm": 0.14270208775997162,
|
| 3362 |
+
"learning_rate": 0.001,
|
| 3363 |
+
"loss": 0.0442,
|
| 3364 |
+
"step": 890
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 30.99,
|
| 3368 |
+
"eval_loss": 5.50925874710083,
|
| 3369 |
+
"eval_runtime": 14.2033,
|
| 3370 |
+
"eval_samples_per_second": 23.797,
|
| 3371 |
+
"eval_steps_per_second": 3.027,
|
| 3372 |
+
"step": 891
|
| 3373 |
+
},
|
| 3374 |
+
{
|
| 3375 |
+
"epoch": 31.03,
|
| 3376 |
+
"grad_norm": 0.14988122880458832,
|
| 3377 |
+
"learning_rate": 0.001,
|
| 3378 |
+
"loss": 0.0391,
|
| 3379 |
+
"step": 892
|
| 3380 |
+
},
|
| 3381 |
+
{
|
| 3382 |
+
"epoch": 31.1,
|
| 3383 |
+
"grad_norm": 0.12136206030845642,
|
| 3384 |
+
"learning_rate": 0.001,
|
| 3385 |
+
"loss": 0.0368,
|
| 3386 |
+
"step": 894
|
| 3387 |
+
},
|
| 3388 |
+
{
|
| 3389 |
+
"epoch": 31.17,
|
| 3390 |
+
"grad_norm": 0.1474607139825821,
|
| 3391 |
+
"learning_rate": 0.001,
|
| 3392 |
+
"loss": 0.0366,
|
| 3393 |
+
"step": 896
|
| 3394 |
+
},
|
| 3395 |
+
{
|
| 3396 |
+
"epoch": 31.23,
|
| 3397 |
+
"grad_norm": 0.1664348542690277,
|
| 3398 |
+
"learning_rate": 0.001,
|
| 3399 |
+
"loss": 0.0364,
|
| 3400 |
+
"step": 898
|
| 3401 |
+
},
|
| 3402 |
+
{
|
| 3403 |
+
"epoch": 31.3,
|
| 3404 |
+
"grad_norm": 0.12352883070707321,
|
| 3405 |
+
"learning_rate": 0.001,
|
| 3406 |
+
"loss": 0.0368,
|
| 3407 |
+
"step": 900
|
| 3408 |
+
},
|
| 3409 |
+
{
|
| 3410 |
+
"epoch": 31.37,
|
| 3411 |
+
"grad_norm": 0.1317415088415146,
|
| 3412 |
+
"learning_rate": 0.001,
|
| 3413 |
+
"loss": 0.0344,
|
| 3414 |
+
"step": 902
|
| 3415 |
+
},
|
| 3416 |
+
{
|
| 3417 |
+
"epoch": 31.44,
|
| 3418 |
+
"grad_norm": 0.12790362536907196,
|
| 3419 |
+
"learning_rate": 0.001,
|
| 3420 |
+
"loss": 0.034,
|
| 3421 |
+
"step": 904
|
| 3422 |
+
},
|
| 3423 |
+
{
|
| 3424 |
+
"epoch": 31.51,
|
| 3425 |
+
"grad_norm": 0.13639821112155914,
|
| 3426 |
+
"learning_rate": 0.001,
|
| 3427 |
+
"loss": 0.0385,
|
| 3428 |
+
"step": 906
|
| 3429 |
+
},
|
| 3430 |
+
{
|
| 3431 |
+
"epoch": 31.58,
|
| 3432 |
+
"grad_norm": 0.14075782895088196,
|
| 3433 |
+
"learning_rate": 0.001,
|
| 3434 |
+
"loss": 0.0414,
|
| 3435 |
+
"step": 908
|
| 3436 |
+
},
|
| 3437 |
+
{
|
| 3438 |
+
"epoch": 31.65,
|
| 3439 |
+
"grad_norm": 0.1388426125049591,
|
| 3440 |
+
"learning_rate": 0.001,
|
| 3441 |
+
"loss": 0.0423,
|
| 3442 |
+
"step": 910
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 31.72,
|
| 3446 |
+
"grad_norm": 0.1304369866847992,
|
| 3447 |
+
"learning_rate": 0.001,
|
| 3448 |
+
"loss": 0.0418,
|
| 3449 |
+
"step": 912
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 31.79,
|
| 3453 |
+
"grad_norm": 0.13760504126548767,
|
| 3454 |
+
"learning_rate": 0.001,
|
| 3455 |
+
"loss": 0.0438,
|
| 3456 |
+
"step": 914
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 31.86,
|
| 3460 |
+
"grad_norm": 0.13388106226921082,
|
| 3461 |
+
"learning_rate": 0.001,
|
| 3462 |
+
"loss": 0.0442,
|
| 3463 |
+
"step": 916
|
| 3464 |
+
},
|
| 3465 |
+
{
|
| 3466 |
+
"epoch": 31.93,
|
| 3467 |
+
"grad_norm": 0.13234779238700867,
|
| 3468 |
+
"learning_rate": 0.001,
|
| 3469 |
+
"loss": 0.0403,
|
| 3470 |
+
"step": 918
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 32.0,
|
| 3474 |
+
"grad_norm": 0.12893211841583252,
|
| 3475 |
+
"learning_rate": 0.001,
|
| 3476 |
+
"loss": 0.0391,
|
| 3477 |
+
"step": 920
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 32.0,
|
| 3481 |
+
"eval_loss": 5.429612159729004,
|
| 3482 |
+
"eval_runtime": 14.2128,
|
| 3483 |
+
"eval_samples_per_second": 23.781,
|
| 3484 |
+
"eval_steps_per_second": 3.025,
|
| 3485 |
+
"step": 920
|
| 3486 |
+
},
|
| 3487 |
+
{
|
| 3488 |
+
"epoch": 32.07,
|
| 3489 |
+
"grad_norm": 0.12455160915851593,
|
| 3490 |
+
"learning_rate": 0.001,
|
| 3491 |
+
"loss": 0.0343,
|
| 3492 |
+
"step": 922
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 32.14,
|
| 3496 |
+
"grad_norm": 0.13458769023418427,
|
| 3497 |
+
"learning_rate": 0.001,
|
| 3498 |
+
"loss": 0.0358,
|
| 3499 |
+
"step": 924
|
| 3500 |
+
},
|
| 3501 |
+
{
|
| 3502 |
+
"epoch": 32.21,
|
| 3503 |
+
"grad_norm": 0.11662837117910385,
|
| 3504 |
+
"learning_rate": 0.001,
|
| 3505 |
+
"loss": 0.0359,
|
| 3506 |
+
"step": 926
|
| 3507 |
+
},
|
| 3508 |
+
{
|
| 3509 |
+
"epoch": 32.28,
|
| 3510 |
+
"grad_norm": 0.1250811070203781,
|
| 3511 |
+
"learning_rate": 0.001,
|
| 3512 |
+
"loss": 0.0366,
|
| 3513 |
+
"step": 928
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 32.35,
|
| 3517 |
+
"grad_norm": 0.1264219582080841,
|
| 3518 |
+
"learning_rate": 0.001,
|
| 3519 |
+
"loss": 0.0374,
|
| 3520 |
+
"step": 930
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 32.42,
|
| 3524 |
+
"grad_norm": 0.12272246927022934,
|
| 3525 |
+
"learning_rate": 0.001,
|
| 3526 |
+
"loss": 0.0378,
|
| 3527 |
+
"step": 932
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 32.49,
|
| 3531 |
+
"grad_norm": 0.12346846610307693,
|
| 3532 |
+
"learning_rate": 0.001,
|
| 3533 |
+
"loss": 0.0387,
|
| 3534 |
+
"step": 934
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 32.56,
|
| 3538 |
+
"grad_norm": 0.12550891935825348,
|
| 3539 |
+
"learning_rate": 0.001,
|
| 3540 |
+
"loss": 0.0355,
|
| 3541 |
+
"step": 936
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 32.63,
|
| 3545 |
+
"grad_norm": 0.13145896792411804,
|
| 3546 |
+
"learning_rate": 0.001,
|
| 3547 |
+
"loss": 0.0371,
|
| 3548 |
+
"step": 938
|
| 3549 |
+
},
|
| 3550 |
+
{
|
| 3551 |
+
"epoch": 32.7,
|
| 3552 |
+
"grad_norm": 0.12549975514411926,
|
| 3553 |
+
"learning_rate": 0.001,
|
| 3554 |
+
"loss": 0.038,
|
| 3555 |
+
"step": 940
|
| 3556 |
+
},
|
| 3557 |
+
{
|
| 3558 |
+
"epoch": 32.77,
|
| 3559 |
+
"grad_norm": 0.13180097937583923,
|
| 3560 |
+
"learning_rate": 0.001,
|
| 3561 |
+
"loss": 0.0361,
|
| 3562 |
+
"step": 942
|
| 3563 |
+
},
|
| 3564 |
+
{
|
| 3565 |
+
"epoch": 32.83,
|
| 3566 |
+
"grad_norm": 0.1240687221288681,
|
| 3567 |
+
"learning_rate": 0.001,
|
| 3568 |
+
"loss": 0.0364,
|
| 3569 |
+
"step": 944
|
| 3570 |
+
},
|
| 3571 |
+
{
|
| 3572 |
+
"epoch": 32.9,
|
| 3573 |
+
"grad_norm": 0.14110788702964783,
|
| 3574 |
+
"learning_rate": 0.001,
|
| 3575 |
+
"loss": 0.0421,
|
| 3576 |
+
"step": 946
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 32.97,
|
| 3580 |
+
"grad_norm": 0.14410488307476044,
|
| 3581 |
+
"learning_rate": 0.001,
|
| 3582 |
+
"loss": 0.0392,
|
| 3583 |
+
"step": 948
|
| 3584 |
+
},
|
| 3585 |
+
{
|
| 3586 |
+
"epoch": 32.97,
|
| 3587 |
+
"eval_loss": 5.23565673828125,
|
| 3588 |
+
"eval_runtime": 14.2111,
|
| 3589 |
+
"eval_samples_per_second": 23.784,
|
| 3590 |
+
"eval_steps_per_second": 3.026,
|
| 3591 |
+
"step": 948
|
| 3592 |
+
},
|
| 3593 |
+
{
|
| 3594 |
+
"epoch": 33.04,
|
| 3595 |
+
"grad_norm": 0.12806980311870575,
|
| 3596 |
+
"learning_rate": 0.001,
|
| 3597 |
+
"loss": 0.041,
|
| 3598 |
+
"step": 950
|
| 3599 |
+
},
|
| 3600 |
+
{
|
| 3601 |
+
"epoch": 33.11,
|
| 3602 |
+
"grad_norm": 0.12071588635444641,
|
| 3603 |
+
"learning_rate": 0.001,
|
| 3604 |
+
"loss": 0.0401,
|
| 3605 |
+
"step": 952
|
| 3606 |
+
},
|
| 3607 |
+
{
|
| 3608 |
+
"epoch": 33.18,
|
| 3609 |
+
"grad_norm": 0.11720431596040726,
|
| 3610 |
+
"learning_rate": 0.001,
|
| 3611 |
+
"loss": 0.035,
|
| 3612 |
+
"step": 954
|
| 3613 |
+
},
|
| 3614 |
+
{
|
| 3615 |
+
"epoch": 33.25,
|
| 3616 |
+
"grad_norm": 0.12089748680591583,
|
| 3617 |
+
"learning_rate": 0.001,
|
| 3618 |
+
"loss": 0.0389,
|
| 3619 |
+
"step": 956
|
| 3620 |
+
},
|
| 3621 |
+
{
|
| 3622 |
+
"epoch": 33.32,
|
| 3623 |
+
"grad_norm": 0.1227785050868988,
|
| 3624 |
+
"learning_rate": 0.001,
|
| 3625 |
+
"loss": 0.0389,
|
| 3626 |
+
"step": 958
|
| 3627 |
+
},
|
| 3628 |
+
{
|
| 3629 |
+
"epoch": 33.39,
|
| 3630 |
+
"grad_norm": 0.12859608232975006,
|
| 3631 |
+
"learning_rate": 0.001,
|
| 3632 |
+
"loss": 0.0342,
|
| 3633 |
+
"step": 960
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 33.46,
|
| 3637 |
+
"grad_norm": 0.2180158495903015,
|
| 3638 |
+
"learning_rate": 0.001,
|
| 3639 |
+
"loss": 0.041,
|
| 3640 |
+
"step": 962
|
| 3641 |
+
},
|
| 3642 |
+
{
|
| 3643 |
+
"epoch": 33.53,
|
| 3644 |
+
"grad_norm": 0.2960141897201538,
|
| 3645 |
+
"learning_rate": 0.001,
|
| 3646 |
+
"loss": 0.0989,
|
| 3647 |
+
"step": 964
|
| 3648 |
+
},
|
| 3649 |
+
{
|
| 3650 |
+
"epoch": 33.6,
|
| 3651 |
+
"grad_norm": 0.15140074491500854,
|
| 3652 |
+
"learning_rate": 0.001,
|
| 3653 |
+
"loss": 0.0503,
|
| 3654 |
+
"step": 966
|
| 3655 |
+
},
|
| 3656 |
+
{
|
| 3657 |
+
"epoch": 33.67,
|
| 3658 |
+
"grad_norm": 0.13876180350780487,
|
| 3659 |
+
"learning_rate": 0.001,
|
| 3660 |
+
"loss": 0.0418,
|
| 3661 |
+
"step": 968
|
| 3662 |
+
},
|
| 3663 |
+
{
|
| 3664 |
+
"epoch": 33.74,
|
| 3665 |
+
"grad_norm": 0.1290924996137619,
|
| 3666 |
+
"learning_rate": 0.001,
|
| 3667 |
+
"loss": 0.0413,
|
| 3668 |
+
"step": 970
|
| 3669 |
+
},
|
| 3670 |
+
{
|
| 3671 |
+
"epoch": 33.81,
|
| 3672 |
+
"grad_norm": 0.13152842223644257,
|
| 3673 |
+
"learning_rate": 0.001,
|
| 3674 |
+
"loss": 0.0398,
|
| 3675 |
+
"step": 972
|
| 3676 |
+
},
|
| 3677 |
+
{
|
| 3678 |
+
"epoch": 33.88,
|
| 3679 |
+
"grad_norm": 0.1308111995458603,
|
| 3680 |
+
"learning_rate": 0.001,
|
| 3681 |
+
"loss": 0.0384,
|
| 3682 |
+
"step": 974
|
| 3683 |
+
},
|
| 3684 |
+
{
|
| 3685 |
+
"epoch": 33.95,
|
| 3686 |
+
"grad_norm": 0.12206979840993881,
|
| 3687 |
+
"learning_rate": 0.001,
|
| 3688 |
+
"loss": 0.0376,
|
| 3689 |
+
"step": 976
|
| 3690 |
+
},
|
| 3691 |
+
{
|
| 3692 |
+
"epoch": 33.98,
|
| 3693 |
+
"eval_loss": 5.226629734039307,
|
| 3694 |
+
"eval_runtime": 14.2113,
|
| 3695 |
+
"eval_samples_per_second": 23.784,
|
| 3696 |
+
"eval_steps_per_second": 3.026,
|
| 3697 |
+
"step": 977
|
| 3698 |
+
},
|
| 3699 |
+
{
|
| 3700 |
+
"epoch": 34.02,
|
| 3701 |
+
"grad_norm": 0.15338914096355438,
|
| 3702 |
+
"learning_rate": 0.001,
|
| 3703 |
+
"loss": 0.036,
|
| 3704 |
+
"step": 978
|
| 3705 |
+
},
|
| 3706 |
+
{
|
| 3707 |
+
"epoch": 34.09,
|
| 3708 |
+
"grad_norm": 0.12300385534763336,
|
| 3709 |
+
"learning_rate": 0.001,
|
| 3710 |
+
"loss": 0.0357,
|
| 3711 |
+
"step": 980
|
| 3712 |
+
},
|
| 3713 |
+
{
|
| 3714 |
+
"epoch": 34.16,
|
| 3715 |
+
"grad_norm": 0.10930658131837845,
|
| 3716 |
+
"learning_rate": 0.001,
|
| 3717 |
+
"loss": 0.03,
|
| 3718 |
+
"step": 982
|
| 3719 |
+
},
|
| 3720 |
+
{
|
| 3721 |
+
"epoch": 34.23,
|
| 3722 |
+
"grad_norm": 0.12387796491384506,
|
| 3723 |
+
"learning_rate": 0.001,
|
| 3724 |
+
"loss": 0.0393,
|
| 3725 |
+
"step": 984
|
| 3726 |
+
},
|
| 3727 |
+
{
|
| 3728 |
+
"epoch": 34.3,
|
| 3729 |
+
"grad_norm": 0.1442791372537613,
|
| 3730 |
+
"learning_rate": 0.001,
|
| 3731 |
+
"loss": 0.0327,
|
| 3732 |
+
"step": 986
|
| 3733 |
+
},
|
| 3734 |
+
{
|
| 3735 |
+
"epoch": 34.37,
|
| 3736 |
+
"grad_norm": 0.12000349164009094,
|
| 3737 |
+
"learning_rate": 0.001,
|
| 3738 |
+
"loss": 0.0356,
|
| 3739 |
+
"step": 988
|
| 3740 |
+
},
|
| 3741 |
+
{
|
| 3742 |
+
"epoch": 34.43,
|
| 3743 |
+
"grad_norm": 0.12804967164993286,
|
| 3744 |
+
"learning_rate": 0.001,
|
| 3745 |
+
"loss": 0.0415,
|
| 3746 |
+
"step": 990
|
| 3747 |
+
},
|
| 3748 |
+
{
|
| 3749 |
+
"epoch": 34.5,
|
| 3750 |
+
"grad_norm": 0.12344588339328766,
|
| 3751 |
+
"learning_rate": 0.001,
|
| 3752 |
+
"loss": 0.0348,
|
| 3753 |
+
"step": 992
|
| 3754 |
+
},
|
| 3755 |
+
{
|
| 3756 |
+
"epoch": 34.57,
|
| 3757 |
+
"grad_norm": 0.12841206789016724,
|
| 3758 |
+
"learning_rate": 0.001,
|
| 3759 |
+
"loss": 0.036,
|
| 3760 |
+
"step": 994
|
| 3761 |
+
},
|
| 3762 |
+
{
|
| 3763 |
+
"epoch": 34.64,
|
| 3764 |
+
"grad_norm": 0.11577708274126053,
|
| 3765 |
+
"learning_rate": 0.001,
|
| 3766 |
+
"loss": 0.0349,
|
| 3767 |
+
"step": 996
|
| 3768 |
+
},
|
| 3769 |
+
{
|
| 3770 |
+
"epoch": 34.71,
|
| 3771 |
+
"grad_norm": 0.12154724448919296,
|
| 3772 |
+
"learning_rate": 0.001,
|
| 3773 |
+
"loss": 0.0384,
|
| 3774 |
+
"step": 998
|
| 3775 |
+
},
|
| 3776 |
+
{
|
| 3777 |
+
"epoch": 34.78,
|
| 3778 |
+
"grad_norm": 0.12926706671714783,
|
| 3779 |
+
"learning_rate": 0.001,
|
| 3780 |
+
"loss": 0.0376,
|
| 3781 |
+
"step": 1000
|
| 3782 |
+
},
|
| 3783 |
+
{
|
| 3784 |
+
"epoch": 34.85,
|
| 3785 |
+
"grad_norm": 0.11031629890203476,
|
| 3786 |
+
"learning_rate": 0.001,
|
| 3787 |
+
"loss": 0.0341,
|
| 3788 |
+
"step": 1002
|
| 3789 |
+
},
|
| 3790 |
+
{
|
| 3791 |
+
"epoch": 34.92,
|
| 3792 |
+
"grad_norm": 0.13376009464263916,
|
| 3793 |
+
"learning_rate": 0.001,
|
| 3794 |
+
"loss": 0.0352,
|
| 3795 |
+
"step": 1004
|
| 3796 |
+
},
|
| 3797 |
+
{
|
| 3798 |
+
"epoch": 34.99,
|
| 3799 |
+
"grad_norm": 0.1240614578127861,
|
| 3800 |
+
"learning_rate": 0.001,
|
| 3801 |
+
"loss": 0.0381,
|
| 3802 |
+
"step": 1006
|
| 3803 |
+
},
|
| 3804 |
+
{
|
| 3805 |
+
"epoch": 34.99,
|
| 3806 |
+
"eval_loss": 5.263000965118408,
|
| 3807 |
+
"eval_runtime": 14.2061,
|
| 3808 |
+
"eval_samples_per_second": 23.793,
|
| 3809 |
+
"eval_steps_per_second": 3.027,
|
| 3810 |
+
"step": 1006
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 35.06,
|
| 3814 |
+
"grad_norm": 0.12797795236110687,
|
| 3815 |
+
"learning_rate": 0.001,
|
| 3816 |
+
"loss": 0.0339,
|
| 3817 |
+
"step": 1008
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 35.13,
|
| 3821 |
+
"grad_norm": 0.13305290043354034,
|
| 3822 |
+
"learning_rate": 0.001,
|
| 3823 |
+
"loss": 0.0333,
|
| 3824 |
+
"step": 1010
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 35.2,
|
| 3828 |
+
"grad_norm": 0.12507498264312744,
|
| 3829 |
+
"learning_rate": 0.001,
|
| 3830 |
+
"loss": 0.0316,
|
| 3831 |
+
"step": 1012
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 35.27,
|
| 3835 |
+
"grad_norm": 0.1163344755768776,
|
| 3836 |
+
"learning_rate": 0.001,
|
| 3837 |
+
"loss": 0.0323,
|
| 3838 |
+
"step": 1014
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 35.34,
|
| 3842 |
+
"grad_norm": 0.11232735961675644,
|
| 3843 |
+
"learning_rate": 0.001,
|
| 3844 |
+
"loss": 0.035,
|
| 3845 |
+
"step": 1016
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 35.41,
|
| 3849 |
+
"grad_norm": 0.11937879770994186,
|
| 3850 |
+
"learning_rate": 0.001,
|
| 3851 |
+
"loss": 0.0343,
|
| 3852 |
+
"step": 1018
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 35.48,
|
| 3856 |
+
"grad_norm": 0.1227354183793068,
|
| 3857 |
+
"learning_rate": 0.001,
|
| 3858 |
+
"loss": 0.0316,
|
| 3859 |
+
"step": 1020
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 35.55,
|
| 3863 |
+
"grad_norm": 0.1201673150062561,
|
| 3864 |
+
"learning_rate": 0.001,
|
| 3865 |
+
"loss": 0.0345,
|
| 3866 |
+
"step": 1022
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 35.62,
|
| 3870 |
+
"grad_norm": 0.11625051498413086,
|
| 3871 |
+
"learning_rate": 0.001,
|
| 3872 |
+
"loss": 0.0305,
|
| 3873 |
+
"step": 1024
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 35.69,
|
| 3877 |
+
"grad_norm": 0.1144721657037735,
|
| 3878 |
+
"learning_rate": 0.001,
|
| 3879 |
+
"loss": 0.03,
|
| 3880 |
+
"step": 1026
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 35.76,
|
| 3884 |
+
"grad_norm": 0.12341687083244324,
|
| 3885 |
+
"learning_rate": 0.001,
|
| 3886 |
+
"loss": 0.0374,
|
| 3887 |
+
"step": 1028
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 35.83,
|
| 3891 |
+
"grad_norm": 0.12577305734157562,
|
| 3892 |
+
"learning_rate": 0.001,
|
| 3893 |
+
"loss": 0.0347,
|
| 3894 |
+
"step": 1030
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 35.9,
|
| 3898 |
+
"grad_norm": 0.11321111768484116,
|
| 3899 |
+
"learning_rate": 0.001,
|
| 3900 |
+
"loss": 0.0336,
|
| 3901 |
+
"step": 1032
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 35.97,
|
| 3905 |
+
"grad_norm": 0.11875444650650024,
|
| 3906 |
+
"learning_rate": 0.001,
|
| 3907 |
+
"loss": 0.0339,
|
| 3908 |
+
"step": 1034
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 36.0,
|
| 3912 |
+
"eval_loss": 5.353184700012207,
|
| 3913 |
+
"eval_runtime": 14.1981,
|
| 3914 |
+
"eval_samples_per_second": 23.806,
|
| 3915 |
+
"eval_steps_per_second": 3.029,
|
| 3916 |
+
"step": 1035
|
| 3917 |
+
},
|
| 3918 |
+
{
|
| 3919 |
+
"epoch": 36.03,
|
| 3920 |
+
"grad_norm": 0.122185118496418,
|
| 3921 |
+
"learning_rate": 0.001,
|
| 3922 |
+
"loss": 0.0285,
|
| 3923 |
+
"step": 1036
|
| 3924 |
+
},
|
| 3925 |
+
{
|
| 3926 |
+
"epoch": 36.1,
|
| 3927 |
+
"grad_norm": 0.1075131967663765,
|
| 3928 |
+
"learning_rate": 0.001,
|
| 3929 |
+
"loss": 0.0312,
|
| 3930 |
+
"step": 1038
|
| 3931 |
+
},
|
| 3932 |
+
{
|
| 3933 |
+
"epoch": 36.17,
|
| 3934 |
+
"grad_norm": 0.12375488132238388,
|
| 3935 |
+
"learning_rate": 0.001,
|
| 3936 |
+
"loss": 0.0291,
|
| 3937 |
+
"step": 1040
|
| 3938 |
+
},
|
| 3939 |
+
{
|
| 3940 |
+
"epoch": 36.24,
|
| 3941 |
+
"grad_norm": 0.12728258967399597,
|
| 3942 |
+
"learning_rate": 0.001,
|
| 3943 |
+
"loss": 0.0333,
|
| 3944 |
+
"step": 1042
|
| 3945 |
+
},
|
| 3946 |
+
{
|
| 3947 |
+
"epoch": 36.31,
|
| 3948 |
+
"grad_norm": 0.1169193834066391,
|
| 3949 |
+
"learning_rate": 0.001,
|
| 3950 |
+
"loss": 0.0309,
|
| 3951 |
+
"step": 1044
|
| 3952 |
+
},
|
| 3953 |
+
{
|
| 3954 |
+
"epoch": 36.38,
|
| 3955 |
+
"grad_norm": 0.11048964411020279,
|
| 3956 |
+
"learning_rate": 0.001,
|
| 3957 |
+
"loss": 0.0299,
|
| 3958 |
+
"step": 1046
|
| 3959 |
+
},
|
| 3960 |
+
{
|
| 3961 |
+
"epoch": 36.45,
|
| 3962 |
+
"grad_norm": 0.11923568695783615,
|
| 3963 |
+
"learning_rate": 0.001,
|
| 3964 |
+
"loss": 0.0301,
|
| 3965 |
+
"step": 1048
|
| 3966 |
+
},
|
| 3967 |
+
{
|
| 3968 |
+
"epoch": 36.52,
|
| 3969 |
+
"grad_norm": 0.1160750761628151,
|
| 3970 |
+
"learning_rate": 0.001,
|
| 3971 |
+
"loss": 0.0287,
|
| 3972 |
+
"step": 1050
|
| 3973 |
+
},
|
| 3974 |
+
{
|
| 3975 |
+
"epoch": 36.59,
|
| 3976 |
+
"grad_norm": 0.12535539269447327,
|
| 3977 |
+
"learning_rate": 0.001,
|
| 3978 |
+
"loss": 0.0335,
|
| 3979 |
+
"step": 1052
|
| 3980 |
+
},
|
| 3981 |
+
{
|
| 3982 |
+
"epoch": 36.66,
|
| 3983 |
+
"grad_norm": 0.13712206482887268,
|
| 3984 |
+
"learning_rate": 0.001,
|
| 3985 |
+
"loss": 0.0358,
|
| 3986 |
+
"step": 1054
|
| 3987 |
+
},
|
| 3988 |
+
{
|
| 3989 |
+
"epoch": 36.73,
|
| 3990 |
+
"grad_norm": 0.12085587531328201,
|
| 3991 |
+
"learning_rate": 0.001,
|
| 3992 |
+
"loss": 0.0362,
|
| 3993 |
+
"step": 1056
|
| 3994 |
+
},
|
| 3995 |
+
{
|
| 3996 |
+
"epoch": 36.8,
|
| 3997 |
+
"grad_norm": 0.1010272428393364,
|
| 3998 |
+
"learning_rate": 0.001,
|
| 3999 |
+
"loss": 0.0324,
|
| 4000 |
+
"step": 1058
|
| 4001 |
+
},
|
| 4002 |
+
{
|
| 4003 |
+
"epoch": 36.87,
|
| 4004 |
+
"grad_norm": 0.11820247769355774,
|
| 4005 |
+
"learning_rate": 0.001,
|
| 4006 |
+
"loss": 0.0315,
|
| 4007 |
+
"step": 1060
|
| 4008 |
+
},
|
| 4009 |
+
{
|
| 4010 |
+
"epoch": 36.94,
|
| 4011 |
+
"grad_norm": 0.1259121298789978,
|
| 4012 |
+
"learning_rate": 0.001,
|
| 4013 |
+
"loss": 0.0377,
|
| 4014 |
+
"step": 1062
|
| 4015 |
+
},
|
| 4016 |
+
{
|
| 4017 |
+
"epoch": 36.97,
|
| 4018 |
+
"eval_loss": 5.44431734085083,
|
| 4019 |
+
"eval_runtime": 14.2065,
|
| 4020 |
+
"eval_samples_per_second": 23.792,
|
| 4021 |
+
"eval_steps_per_second": 3.027,
|
| 4022 |
+
"step": 1063
|
| 4023 |
+
},
|
| 4024 |
+
{
|
| 4025 |
+
"epoch": 37.01,
|
| 4026 |
+
"grad_norm": 0.12725719809532166,
|
| 4027 |
+
"learning_rate": 0.001,
|
| 4028 |
+
"loss": 0.0277,
|
| 4029 |
+
"step": 1064
|
| 4030 |
+
},
|
| 4031 |
+
{
|
| 4032 |
+
"epoch": 37.08,
|
| 4033 |
+
"grad_norm": 0.11410412192344666,
|
| 4034 |
+
"learning_rate": 0.001,
|
| 4035 |
+
"loss": 0.0282,
|
| 4036 |
+
"step": 1066
|
| 4037 |
+
},
|
| 4038 |
+
{
|
| 4039 |
+
"epoch": 37.15,
|
| 4040 |
+
"grad_norm": 0.11205872148275375,
|
| 4041 |
+
"learning_rate": 0.001,
|
| 4042 |
+
"loss": 0.0302,
|
| 4043 |
+
"step": 1068
|
| 4044 |
+
},
|
| 4045 |
+
{
|
| 4046 |
+
"epoch": 37.22,
|
| 4047 |
+
"grad_norm": 0.09593572467565536,
|
| 4048 |
+
"learning_rate": 0.001,
|
| 4049 |
+
"loss": 0.0263,
|
| 4050 |
+
"step": 1070
|
| 4051 |
+
},
|
| 4052 |
+
{
|
| 4053 |
+
"epoch": 37.29,
|
| 4054 |
+
"grad_norm": 0.12462112307548523,
|
| 4055 |
+
"learning_rate": 0.001,
|
| 4056 |
+
"loss": 0.0328,
|
| 4057 |
+
"step": 1072
|
| 4058 |
+
},
|
| 4059 |
+
{
|
| 4060 |
+
"epoch": 37.36,
|
| 4061 |
+
"grad_norm": 0.11873652786016464,
|
| 4062 |
+
"learning_rate": 0.001,
|
| 4063 |
+
"loss": 0.0328,
|
| 4064 |
+
"step": 1074
|
| 4065 |
+
},
|
| 4066 |
+
{
|
| 4067 |
+
"epoch": 37.43,
|
| 4068 |
+
"grad_norm": 0.1105048656463623,
|
| 4069 |
+
"learning_rate": 0.001,
|
| 4070 |
+
"loss": 0.0298,
|
| 4071 |
+
"step": 1076
|
| 4072 |
+
},
|
| 4073 |
+
{
|
| 4074 |
+
"epoch": 37.5,
|
| 4075 |
+
"grad_norm": 0.10692736506462097,
|
| 4076 |
+
"learning_rate": 0.001,
|
| 4077 |
+
"loss": 0.0328,
|
| 4078 |
+
"step": 1078
|
| 4079 |
+
},
|
| 4080 |
+
{
|
| 4081 |
+
"epoch": 37.57,
|
| 4082 |
+
"grad_norm": 0.12755931913852692,
|
| 4083 |
+
"learning_rate": 0.001,
|
| 4084 |
+
"loss": 0.0321,
|
| 4085 |
+
"step": 1080
|
| 4086 |
+
},
|
| 4087 |
+
{
|
| 4088 |
+
"epoch": 37.63,
|
| 4089 |
+
"grad_norm": 0.13307899236679077,
|
| 4090 |
+
"learning_rate": 0.001,
|
| 4091 |
+
"loss": 0.0329,
|
| 4092 |
+
"step": 1082
|
| 4093 |
+
},
|
| 4094 |
+
{
|
| 4095 |
+
"epoch": 37.7,
|
| 4096 |
+
"grad_norm": 0.12125247716903687,
|
| 4097 |
+
"learning_rate": 0.001,
|
| 4098 |
+
"loss": 0.034,
|
| 4099 |
+
"step": 1084
|
| 4100 |
+
},
|
| 4101 |
+
{
|
| 4102 |
+
"epoch": 37.77,
|
| 4103 |
+
"grad_norm": 0.12050389498472214,
|
| 4104 |
+
"learning_rate": 0.001,
|
| 4105 |
+
"loss": 0.031,
|
| 4106 |
+
"step": 1086
|
| 4107 |
+
},
|
| 4108 |
+
{
|
| 4109 |
+
"epoch": 37.84,
|
| 4110 |
+
"grad_norm": 0.11437363177537918,
|
| 4111 |
+
"learning_rate": 0.001,
|
| 4112 |
+
"loss": 0.0308,
|
| 4113 |
+
"step": 1088
|
| 4114 |
+
},
|
| 4115 |
+
{
|
| 4116 |
+
"epoch": 37.91,
|
| 4117 |
+
"grad_norm": 0.13162565231323242,
|
| 4118 |
+
"learning_rate": 0.001,
|
| 4119 |
+
"loss": 0.037,
|
| 4120 |
+
"step": 1090
|
| 4121 |
+
},
|
| 4122 |
+
{
|
| 4123 |
+
"epoch": 37.98,
|
| 4124 |
+
"grad_norm": 0.13537472486495972,
|
| 4125 |
+
"learning_rate": 0.001,
|
| 4126 |
+
"loss": 0.0336,
|
| 4127 |
+
"step": 1092
|
| 4128 |
+
},
|
| 4129 |
+
{
|
| 4130 |
+
"epoch": 37.98,
|
| 4131 |
+
"eval_loss": 5.080876350402832,
|
| 4132 |
+
"eval_runtime": 14.2123,
|
| 4133 |
+
"eval_samples_per_second": 23.782,
|
| 4134 |
+
"eval_steps_per_second": 3.026,
|
| 4135 |
+
"step": 1092
|
| 4136 |
+
},
|
| 4137 |
+
{
|
| 4138 |
+
"epoch": 38.05,
|
| 4139 |
+
"grad_norm": 0.10324922204017639,
|
| 4140 |
+
"learning_rate": 0.001,
|
| 4141 |
+
"loss": 0.0311,
|
| 4142 |
+
"step": 1094
|
| 4143 |
+
},
|
| 4144 |
+
{
|
| 4145 |
+
"epoch": 38.12,
|
| 4146 |
+
"grad_norm": 0.10196162760257721,
|
| 4147 |
+
"learning_rate": 0.001,
|
| 4148 |
+
"loss": 0.0274,
|
| 4149 |
+
"step": 1096
|
| 4150 |
+
},
|
| 4151 |
+
{
|
| 4152 |
+
"epoch": 38.19,
|
| 4153 |
+
"grad_norm": 0.10867214947938919,
|
| 4154 |
+
"learning_rate": 0.001,
|
| 4155 |
+
"loss": 0.0281,
|
| 4156 |
+
"step": 1098
|
| 4157 |
+
},
|
| 4158 |
+
{
|
| 4159 |
+
"epoch": 38.26,
|
| 4160 |
+
"grad_norm": 0.10762563347816467,
|
| 4161 |
+
"learning_rate": 0.001,
|
| 4162 |
+
"loss": 0.0301,
|
| 4163 |
+
"step": 1100
|
| 4164 |
+
},
|
| 4165 |
+
{
|
| 4166 |
+
"epoch": 38.33,
|
| 4167 |
+
"grad_norm": 0.10636897385120392,
|
| 4168 |
+
"learning_rate": 0.001,
|
| 4169 |
+
"loss": 0.0305,
|
| 4170 |
+
"step": 1102
|
| 4171 |
+
},
|
| 4172 |
+
{
|
| 4173 |
+
"epoch": 38.4,
|
| 4174 |
+
"grad_norm": 0.10994244366884232,
|
| 4175 |
+
"learning_rate": 0.001,
|
| 4176 |
+
"loss": 0.0328,
|
| 4177 |
+
"step": 1104
|
| 4178 |
+
},
|
| 4179 |
+
{
|
| 4180 |
+
"epoch": 38.47,
|
| 4181 |
+
"grad_norm": 0.11419042944908142,
|
| 4182 |
+
"learning_rate": 0.001,
|
| 4183 |
+
"loss": 0.0328,
|
| 4184 |
+
"step": 1106
|
| 4185 |
+
},
|
| 4186 |
+
{
|
| 4187 |
+
"epoch": 38.54,
|
| 4188 |
+
"grad_norm": 0.10719581693410873,
|
| 4189 |
+
"learning_rate": 0.001,
|
| 4190 |
+
"loss": 0.0359,
|
| 4191 |
+
"step": 1108
|
| 4192 |
+
},
|
| 4193 |
+
{
|
| 4194 |
+
"epoch": 38.61,
|
| 4195 |
+
"grad_norm": 0.13429002463817596,
|
| 4196 |
+
"learning_rate": 0.001,
|
| 4197 |
+
"loss": 0.0302,
|
| 4198 |
+
"step": 1110
|
| 4199 |
+
},
|
| 4200 |
+
{
|
| 4201 |
+
"epoch": 38.68,
|
| 4202 |
+
"grad_norm": 0.11462274938821793,
|
| 4203 |
+
"learning_rate": 0.001,
|
| 4204 |
+
"loss": 0.0311,
|
| 4205 |
+
"step": 1112
|
| 4206 |
+
},
|
| 4207 |
+
{
|
| 4208 |
+
"epoch": 38.75,
|
| 4209 |
+
"grad_norm": 0.12587036192417145,
|
| 4210 |
+
"learning_rate": 0.001,
|
| 4211 |
+
"loss": 0.0303,
|
| 4212 |
+
"step": 1114
|
| 4213 |
+
},
|
| 4214 |
+
{
|
| 4215 |
+
"epoch": 38.82,
|
| 4216 |
+
"grad_norm": 0.11628645658493042,
|
| 4217 |
+
"learning_rate": 0.001,
|
| 4218 |
+
"loss": 0.0294,
|
| 4219 |
+
"step": 1116
|
| 4220 |
+
},
|
| 4221 |
+
{
|
| 4222 |
+
"epoch": 38.89,
|
| 4223 |
+
"grad_norm": 0.1132420152425766,
|
| 4224 |
+
"learning_rate": 0.001,
|
| 4225 |
+
"loss": 0.0344,
|
| 4226 |
+
"step": 1118
|
| 4227 |
+
},
|
| 4228 |
+
{
|
| 4229 |
+
"epoch": 38.96,
|
| 4230 |
+
"grad_norm": 0.12212924659252167,
|
| 4231 |
+
"learning_rate": 0.001,
|
| 4232 |
+
"loss": 0.0316,
|
| 4233 |
+
"step": 1120
|
| 4234 |
+
},
|
| 4235 |
+
{
|
| 4236 |
+
"epoch": 38.96,
|
| 4237 |
+
"eval_loss": 5.267875671386719,
|
| 4238 |
+
"eval_runtime": 14.2092,
|
| 4239 |
+
"eval_samples_per_second": 23.787,
|
| 4240 |
+
"eval_steps_per_second": 3.026,
|
| 4241 |
+
"step": 1120
|
| 4242 |
+
},
|
| 4243 |
+
{
|
| 4244 |
+
"epoch": 38.96,
|
| 4245 |
+
"step": 1120,
|
| 4246 |
+
"total_flos": 1.0970359673435505e+18,
|
| 4247 |
+
"train_loss": 0.32971865670822026,
|
| 4248 |
+
"train_runtime": 25027.7089,
|
| 4249 |
+
"train_samples_per_second": 5.87,
|
| 4250 |
+
"train_steps_per_second": 0.045
|
| 4251 |
+
}
|
| 4252 |
+
],
|
| 4253 |
+
"logging_steps": 2,
|
| 4254 |
+
"max_steps": 1120,
|
| 4255 |
+
"num_input_tokens_seen": 0,
|
| 4256 |
+
"num_train_epochs": 40,
|
| 4257 |
+
"save_steps": 500,
|
| 4258 |
+
"total_flos": 1.0970359673435505e+18,
|
| 4259 |
+
"train_batch_size": 8,
|
| 4260 |
+
"trial_name": null,
|
| 4261 |
+
"trial_params": null
|
| 4262 |
+
}
|