End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +3238 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3.1
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3.1-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: oh_scale_x.25_compute_equal
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# oh_scale_x.25_compute_equal
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 2.3639
|
| 21 |
|
|
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3.1-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: oh_scale_x.25_compute_equal
|
|
|
|
| 16 |
|
| 17 |
# oh_scale_x.25_compute_equal
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the mlfoundations-dev/oh-dcft-v1.3_no-curation_gpt-4o-mini_scale_0.25x dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 2.3639
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 45.93475177304965,
|
| 3 |
+
"eval_loss": 2.3638699054718018,
|
| 4 |
+
"eval_runtime": 62.0556,
|
| 5 |
+
"eval_samples_per_second": 38.224,
|
| 6 |
+
"eval_steps_per_second": 0.612,
|
| 7 |
+
"total_flos": 6780341389885440.0,
|
| 8 |
+
"train_loss": 0.15424752476433726,
|
| 9 |
+
"train_runtime": 138136.8738,
|
| 10 |
+
"train_samples_per_second": 15.006,
|
| 11 |
+
"train_steps_per_second": 0.029
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 45.93475177304965,
|
| 3 |
+
"eval_loss": 2.3638699054718018,
|
| 4 |
+
"eval_runtime": 62.0556,
|
| 5 |
+
"eval_samples_per_second": 38.224,
|
| 6 |
+
"eval_steps_per_second": 0.612
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 45.93475177304965,
|
| 3 |
+
"total_flos": 6780341389885440.0,
|
| 4 |
+
"train_loss": 0.15424752476433726,
|
| 5 |
+
"train_runtime": 138136.8738,
|
| 6 |
+
"train_samples_per_second": 15.006,
|
| 7 |
+
"train_steps_per_second": 0.029
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3238 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 45.93475177304965,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4048,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11347517730496454,
|
| 13 |
+
"grad_norm": 2.3027738179328545,
|
| 14 |
+
"learning_rate": 5e-06,
|
| 15 |
+
"loss": 1.0453,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.22695035460992907,
|
| 20 |
+
"grad_norm": 2.6921655569071072,
|
| 21 |
+
"learning_rate": 5e-06,
|
| 22 |
+
"loss": 0.9428,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.3404255319148936,
|
| 27 |
+
"grad_norm": 3.106201627631174,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.9023,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.45390070921985815,
|
| 34 |
+
"grad_norm": 1.1134207204878144,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 0.8741,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5673758865248227,
|
| 41 |
+
"grad_norm": 1.5570541261244333,
|
| 42 |
+
"learning_rate": 5e-06,
|
| 43 |
+
"loss": 0.8561,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.6808510638297872,
|
| 48 |
+
"grad_norm": 1.2304331395600878,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.836,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.7943262411347518,
|
| 55 |
+
"grad_norm": 0.8463468873895378,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.8276,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.9078014184397163,
|
| 62 |
+
"grad_norm": 1.228944154284785,
|
| 63 |
+
"learning_rate": 5e-06,
|
| 64 |
+
"loss": 0.8194,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.9985815602836879,
|
| 69 |
+
"eval_loss": 0.8080979585647583,
|
| 70 |
+
"eval_runtime": 60.7189,
|
| 71 |
+
"eval_samples_per_second": 39.065,
|
| 72 |
+
"eval_steps_per_second": 0.626,
|
| 73 |
+
"step": 88
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.0212765957446808,
|
| 77 |
+
"grad_norm": 0.7120675758086178,
|
| 78 |
+
"learning_rate": 5e-06,
|
| 79 |
+
"loss": 0.8716,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.1347517730496455,
|
| 84 |
+
"grad_norm": 0.5665363188375315,
|
| 85 |
+
"learning_rate": 5e-06,
|
| 86 |
+
"loss": 0.7617,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.24822695035461,
|
| 91 |
+
"grad_norm": 0.7501407091895282,
|
| 92 |
+
"learning_rate": 5e-06,
|
| 93 |
+
"loss": 0.7499,
|
| 94 |
+
"step": 110
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.3617021276595744,
|
| 98 |
+
"grad_norm": 0.8887026186759622,
|
| 99 |
+
"learning_rate": 5e-06,
|
| 100 |
+
"loss": 0.7506,
|
| 101 |
+
"step": 120
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.475177304964539,
|
| 105 |
+
"grad_norm": 0.9778742655108387,
|
| 106 |
+
"learning_rate": 5e-06,
|
| 107 |
+
"loss": 0.7511,
|
| 108 |
+
"step": 130
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.5886524822695036,
|
| 112 |
+
"grad_norm": 0.9687851476126442,
|
| 113 |
+
"learning_rate": 5e-06,
|
| 114 |
+
"loss": 0.7457,
|
| 115 |
+
"step": 140
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.702127659574468,
|
| 119 |
+
"grad_norm": 0.9341422096680074,
|
| 120 |
+
"learning_rate": 5e-06,
|
| 121 |
+
"loss": 0.7436,
|
| 122 |
+
"step": 150
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.8156028368794326,
|
| 126 |
+
"grad_norm": 1.127646790775998,
|
| 127 |
+
"learning_rate": 5e-06,
|
| 128 |
+
"loss": 0.7376,
|
| 129 |
+
"step": 160
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.9290780141843973,
|
| 133 |
+
"grad_norm": 0.8756484845113162,
|
| 134 |
+
"learning_rate": 5e-06,
|
| 135 |
+
"loss": 0.7435,
|
| 136 |
+
"step": 170
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.9971631205673759,
|
| 140 |
+
"eval_loss": 0.7863057851791382,
|
| 141 |
+
"eval_runtime": 61.6081,
|
| 142 |
+
"eval_samples_per_second": 38.501,
|
| 143 |
+
"eval_steps_per_second": 0.617,
|
| 144 |
+
"step": 176
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.0425531914893615,
|
| 148 |
+
"grad_norm": 1.2181284988254628,
|
| 149 |
+
"learning_rate": 5e-06,
|
| 150 |
+
"loss": 0.7763,
|
| 151 |
+
"step": 180
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.1560283687943262,
|
| 155 |
+
"grad_norm": 1.1289889327658702,
|
| 156 |
+
"learning_rate": 5e-06,
|
| 157 |
+
"loss": 0.6823,
|
| 158 |
+
"step": 190
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.269503546099291,
|
| 162 |
+
"grad_norm": 0.8462647029938217,
|
| 163 |
+
"learning_rate": 5e-06,
|
| 164 |
+
"loss": 0.6753,
|
| 165 |
+
"step": 200
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.382978723404255,
|
| 169 |
+
"grad_norm": 0.7466282838810588,
|
| 170 |
+
"learning_rate": 5e-06,
|
| 171 |
+
"loss": 0.6782,
|
| 172 |
+
"step": 210
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.49645390070922,
|
| 176 |
+
"grad_norm": 1.5023120623230148,
|
| 177 |
+
"learning_rate": 5e-06,
|
| 178 |
+
"loss": 0.6788,
|
| 179 |
+
"step": 220
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.6099290780141846,
|
| 183 |
+
"grad_norm": 1.4790514704268571,
|
| 184 |
+
"learning_rate": 5e-06,
|
| 185 |
+
"loss": 0.6823,
|
| 186 |
+
"step": 230
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.723404255319149,
|
| 190 |
+
"grad_norm": 1.400821347106736,
|
| 191 |
+
"learning_rate": 5e-06,
|
| 192 |
+
"loss": 0.6851,
|
| 193 |
+
"step": 240
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.8368794326241136,
|
| 197 |
+
"grad_norm": 1.4979326003170097,
|
| 198 |
+
"learning_rate": 5e-06,
|
| 199 |
+
"loss": 0.6837,
|
| 200 |
+
"step": 250
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.950354609929078,
|
| 204 |
+
"grad_norm": 1.0055414499863942,
|
| 205 |
+
"learning_rate": 5e-06,
|
| 206 |
+
"loss": 0.6821,
|
| 207 |
+
"step": 260
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 2.9957446808510637,
|
| 211 |
+
"eval_loss": 0.7851720452308655,
|
| 212 |
+
"eval_runtime": 61.627,
|
| 213 |
+
"eval_samples_per_second": 38.49,
|
| 214 |
+
"eval_steps_per_second": 0.617,
|
| 215 |
+
"step": 264
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 3.0638297872340425,
|
| 219 |
+
"grad_norm": 1.6752123419187452,
|
| 220 |
+
"learning_rate": 5e-06,
|
| 221 |
+
"loss": 0.7145,
|
| 222 |
+
"step": 270
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 3.1773049645390072,
|
| 226 |
+
"grad_norm": 0.9962830747356258,
|
| 227 |
+
"learning_rate": 5e-06,
|
| 228 |
+
"loss": 0.6212,
|
| 229 |
+
"step": 280
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 3.2907801418439715,
|
| 233 |
+
"grad_norm": 0.972324563152424,
|
| 234 |
+
"learning_rate": 5e-06,
|
| 235 |
+
"loss": 0.6279,
|
| 236 |
+
"step": 290
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 3.404255319148936,
|
| 240 |
+
"grad_norm": 1.148916684368535,
|
| 241 |
+
"learning_rate": 5e-06,
|
| 242 |
+
"loss": 0.6234,
|
| 243 |
+
"step": 300
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 3.5177304964539005,
|
| 247 |
+
"grad_norm": 0.9255677977296973,
|
| 248 |
+
"learning_rate": 5e-06,
|
| 249 |
+
"loss": 0.6239,
|
| 250 |
+
"step": 310
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 3.631205673758865,
|
| 254 |
+
"grad_norm": 0.8748135331658988,
|
| 255 |
+
"learning_rate": 5e-06,
|
| 256 |
+
"loss": 0.6263,
|
| 257 |
+
"step": 320
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 3.74468085106383,
|
| 261 |
+
"grad_norm": 0.7126896324984585,
|
| 262 |
+
"learning_rate": 5e-06,
|
| 263 |
+
"loss": 0.6352,
|
| 264 |
+
"step": 330
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 3.8581560283687946,
|
| 268 |
+
"grad_norm": 0.7000351915001751,
|
| 269 |
+
"learning_rate": 5e-06,
|
| 270 |
+
"loss": 0.6357,
|
| 271 |
+
"step": 340
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 3.971631205673759,
|
| 275 |
+
"grad_norm": 0.9269318996308881,
|
| 276 |
+
"learning_rate": 5e-06,
|
| 277 |
+
"loss": 0.6325,
|
| 278 |
+
"step": 350
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 3.9943262411347518,
|
| 282 |
+
"eval_loss": 0.7985679507255554,
|
| 283 |
+
"eval_runtime": 61.9924,
|
| 284 |
+
"eval_samples_per_second": 38.263,
|
| 285 |
+
"eval_steps_per_second": 0.613,
|
| 286 |
+
"step": 352
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 4.085106382978723,
|
| 290 |
+
"grad_norm": 1.3604073736131337,
|
| 291 |
+
"learning_rate": 5e-06,
|
| 292 |
+
"loss": 0.6418,
|
| 293 |
+
"step": 360
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 4.198581560283688,
|
| 297 |
+
"grad_norm": 1.064592781888281,
|
| 298 |
+
"learning_rate": 5e-06,
|
| 299 |
+
"loss": 0.57,
|
| 300 |
+
"step": 370
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 4.3120567375886525,
|
| 304 |
+
"grad_norm": 1.0456672732780632,
|
| 305 |
+
"learning_rate": 5e-06,
|
| 306 |
+
"loss": 0.5737,
|
| 307 |
+
"step": 380
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 4.425531914893617,
|
| 311 |
+
"grad_norm": 0.8594065634354029,
|
| 312 |
+
"learning_rate": 5e-06,
|
| 313 |
+
"loss": 0.5751,
|
| 314 |
+
"step": 390
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 4.539007092198582,
|
| 318 |
+
"grad_norm": 0.8133394847298308,
|
| 319 |
+
"learning_rate": 5e-06,
|
| 320 |
+
"loss": 0.5814,
|
| 321 |
+
"step": 400
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 4.652482269503546,
|
| 325 |
+
"grad_norm": 0.8444234885021146,
|
| 326 |
+
"learning_rate": 5e-06,
|
| 327 |
+
"loss": 0.5767,
|
| 328 |
+
"step": 410
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 4.76595744680851,
|
| 332 |
+
"grad_norm": 0.9079077917670704,
|
| 333 |
+
"learning_rate": 5e-06,
|
| 334 |
+
"loss": 0.5791,
|
| 335 |
+
"step": 420
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 4.879432624113475,
|
| 339 |
+
"grad_norm": 0.9376422685057096,
|
| 340 |
+
"learning_rate": 5e-06,
|
| 341 |
+
"loss": 0.5868,
|
| 342 |
+
"step": 430
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 4.99290780141844,
|
| 346 |
+
"grad_norm": 0.973277316302177,
|
| 347 |
+
"learning_rate": 5e-06,
|
| 348 |
+
"loss": 0.5795,
|
| 349 |
+
"step": 440
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 4.99290780141844,
|
| 353 |
+
"eval_loss": 0.8202199339866638,
|
| 354 |
+
"eval_runtime": 61.7078,
|
| 355 |
+
"eval_samples_per_second": 38.439,
|
| 356 |
+
"eval_steps_per_second": 0.616,
|
| 357 |
+
"step": 440
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 5.1063829787234045,
|
| 361 |
+
"grad_norm": 1.3674880756976207,
|
| 362 |
+
"learning_rate": 5e-06,
|
| 363 |
+
"loss": 0.5655,
|
| 364 |
+
"step": 450
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 5.219858156028369,
|
| 368 |
+
"grad_norm": 1.7541639365749628,
|
| 369 |
+
"learning_rate": 5e-06,
|
| 370 |
+
"loss": 0.509,
|
| 371 |
+
"step": 460
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 5.333333333333333,
|
| 375 |
+
"grad_norm": 1.5509973437926652,
|
| 376 |
+
"learning_rate": 5e-06,
|
| 377 |
+
"loss": 0.5168,
|
| 378 |
+
"step": 470
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 5.446808510638298,
|
| 382 |
+
"grad_norm": 1.1574291804275065,
|
| 383 |
+
"learning_rate": 5e-06,
|
| 384 |
+
"loss": 0.5168,
|
| 385 |
+
"step": 480
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 5.560283687943262,
|
| 389 |
+
"grad_norm": 2.2427910706765473,
|
| 390 |
+
"learning_rate": 5e-06,
|
| 391 |
+
"loss": 0.5192,
|
| 392 |
+
"step": 490
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 5.673758865248227,
|
| 396 |
+
"grad_norm": 1.5662018655408698,
|
| 397 |
+
"learning_rate": 5e-06,
|
| 398 |
+
"loss": 0.5153,
|
| 399 |
+
"step": 500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 5.787234042553192,
|
| 403 |
+
"grad_norm": 1.1983365699034005,
|
| 404 |
+
"learning_rate": 5e-06,
|
| 405 |
+
"loss": 0.5235,
|
| 406 |
+
"step": 510
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 5.900709219858156,
|
| 410 |
+
"grad_norm": 0.9749148887193015,
|
| 411 |
+
"learning_rate": 5e-06,
|
| 412 |
+
"loss": 0.5193,
|
| 413 |
+
"step": 520
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 5.991489361702127,
|
| 417 |
+
"eval_loss": 0.8596100807189941,
|
| 418 |
+
"eval_runtime": 61.8556,
|
| 419 |
+
"eval_samples_per_second": 38.347,
|
| 420 |
+
"eval_steps_per_second": 0.614,
|
| 421 |
+
"step": 528
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 6.01418439716312,
|
| 425 |
+
"grad_norm": 1.9275086532414556,
|
| 426 |
+
"learning_rate": 5e-06,
|
| 427 |
+
"loss": 0.5558,
|
| 428 |
+
"step": 530
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 6.127659574468085,
|
| 432 |
+
"grad_norm": 1.6670293640581666,
|
| 433 |
+
"learning_rate": 5e-06,
|
| 434 |
+
"loss": 0.4668,
|
| 435 |
+
"step": 540
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 6.24113475177305,
|
| 439 |
+
"grad_norm": 1.4632224010725146,
|
| 440 |
+
"learning_rate": 5e-06,
|
| 441 |
+
"loss": 0.4606,
|
| 442 |
+
"step": 550
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 6.3546099290780145,
|
| 446 |
+
"grad_norm": 1.418941753583161,
|
| 447 |
+
"learning_rate": 5e-06,
|
| 448 |
+
"loss": 0.4659,
|
| 449 |
+
"step": 560
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 6.468085106382979,
|
| 453 |
+
"grad_norm": 1.2205311444987517,
|
| 454 |
+
"learning_rate": 5e-06,
|
| 455 |
+
"loss": 0.464,
|
| 456 |
+
"step": 570
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 6.581560283687943,
|
| 460 |
+
"grad_norm": 1.009077269752528,
|
| 461 |
+
"learning_rate": 5e-06,
|
| 462 |
+
"loss": 0.4692,
|
| 463 |
+
"step": 580
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 6.695035460992908,
|
| 467 |
+
"grad_norm": 0.9935626851765949,
|
| 468 |
+
"learning_rate": 5e-06,
|
| 469 |
+
"loss": 0.4726,
|
| 470 |
+
"step": 590
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 6.808510638297872,
|
| 474 |
+
"grad_norm": 1.6179148008775062,
|
| 475 |
+
"learning_rate": 5e-06,
|
| 476 |
+
"loss": 0.4747,
|
| 477 |
+
"step": 600
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 6.921985815602837,
|
| 481 |
+
"grad_norm": 1.321766747558007,
|
| 482 |
+
"learning_rate": 5e-06,
|
| 483 |
+
"loss": 0.4751,
|
| 484 |
+
"step": 610
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 6.990070921985816,
|
| 488 |
+
"eval_loss": 0.9138871431350708,
|
| 489 |
+
"eval_runtime": 61.7848,
|
| 490 |
+
"eval_samples_per_second": 38.391,
|
| 491 |
+
"eval_steps_per_second": 0.615,
|
| 492 |
+
"step": 616
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.035460992907802,
|
| 496 |
+
"grad_norm": 2.5479949839536995,
|
| 497 |
+
"learning_rate": 5e-06,
|
| 498 |
+
"loss": 0.4914,
|
| 499 |
+
"step": 620
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.148936170212766,
|
| 503 |
+
"grad_norm": 1.9532287720342911,
|
| 504 |
+
"learning_rate": 5e-06,
|
| 505 |
+
"loss": 0.4067,
|
| 506 |
+
"step": 630
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.26241134751773,
|
| 510 |
+
"grad_norm": 1.2471863458868895,
|
| 511 |
+
"learning_rate": 5e-06,
|
| 512 |
+
"loss": 0.407,
|
| 513 |
+
"step": 640
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.375886524822695,
|
| 517 |
+
"grad_norm": 1.380768007840979,
|
| 518 |
+
"learning_rate": 5e-06,
|
| 519 |
+
"loss": 0.4087,
|
| 520 |
+
"step": 650
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.48936170212766,
|
| 524 |
+
"grad_norm": 1.8120917410891308,
|
| 525 |
+
"learning_rate": 5e-06,
|
| 526 |
+
"loss": 0.4093,
|
| 527 |
+
"step": 660
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.602836879432624,
|
| 531 |
+
"grad_norm": 1.326806876928333,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.4097,
|
| 534 |
+
"step": 670
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.716312056737589,
|
| 538 |
+
"grad_norm": 1.6937687132719013,
|
| 539 |
+
"learning_rate": 5e-06,
|
| 540 |
+
"loss": 0.4158,
|
| 541 |
+
"step": 680
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.829787234042553,
|
| 545 |
+
"grad_norm": 1.3919385940617122,
|
| 546 |
+
"learning_rate": 5e-06,
|
| 547 |
+
"loss": 0.4179,
|
| 548 |
+
"step": 690
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.943262411347518,
|
| 552 |
+
"grad_norm": 1.7802559463891072,
|
| 553 |
+
"learning_rate": 5e-06,
|
| 554 |
+
"loss": 0.4221,
|
| 555 |
+
"step": 700
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 8.0,
|
| 559 |
+
"eval_loss": 1.000566840171814,
|
| 560 |
+
"eval_runtime": 61.7709,
|
| 561 |
+
"eval_samples_per_second": 38.4,
|
| 562 |
+
"eval_steps_per_second": 0.615,
|
| 563 |
+
"step": 705
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 8.056737588652481,
|
| 567 |
+
"grad_norm": 2.1738021359459454,
|
| 568 |
+
"learning_rate": 5e-06,
|
| 569 |
+
"loss": 0.424,
|
| 570 |
+
"step": 710
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 8.170212765957446,
|
| 574 |
+
"grad_norm": 2.0438827395655847,
|
| 575 |
+
"learning_rate": 5e-06,
|
| 576 |
+
"loss": 0.3553,
|
| 577 |
+
"step": 720
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 8.28368794326241,
|
| 581 |
+
"grad_norm": 1.8850254362666403,
|
| 582 |
+
"learning_rate": 5e-06,
|
| 583 |
+
"loss": 0.3587,
|
| 584 |
+
"step": 730
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 8.397163120567376,
|
| 588 |
+
"grad_norm": 2.5094106390967035,
|
| 589 |
+
"learning_rate": 5e-06,
|
| 590 |
+
"loss": 0.3574,
|
| 591 |
+
"step": 740
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 8.51063829787234,
|
| 595 |
+
"grad_norm": 2.2572139568178513,
|
| 596 |
+
"learning_rate": 5e-06,
|
| 597 |
+
"loss": 0.3619,
|
| 598 |
+
"step": 750
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 8.624113475177305,
|
| 602 |
+
"grad_norm": 1.2704163496310845,
|
| 603 |
+
"learning_rate": 5e-06,
|
| 604 |
+
"loss": 0.3583,
|
| 605 |
+
"step": 760
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 8.73758865248227,
|
| 609 |
+
"grad_norm": 1.8304249534362902,
|
| 610 |
+
"learning_rate": 5e-06,
|
| 611 |
+
"loss": 0.3604,
|
| 612 |
+
"step": 770
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 8.851063829787234,
|
| 616 |
+
"grad_norm": 1.6235522609505653,
|
| 617 |
+
"learning_rate": 5e-06,
|
| 618 |
+
"loss": 0.363,
|
| 619 |
+
"step": 780
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 8.964539007092199,
|
| 623 |
+
"grad_norm": 2.126756301230373,
|
| 624 |
+
"learning_rate": 5e-06,
|
| 625 |
+
"loss": 0.3649,
|
| 626 |
+
"step": 790
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 8.998581560283688,
|
| 630 |
+
"eval_loss": 1.0596354007720947,
|
| 631 |
+
"eval_runtime": 62.8012,
|
| 632 |
+
"eval_samples_per_second": 37.77,
|
| 633 |
+
"eval_steps_per_second": 0.605,
|
| 634 |
+
"step": 793
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 9.078014184397164,
|
| 638 |
+
"grad_norm": 2.421069501768367,
|
| 639 |
+
"learning_rate": 5e-06,
|
| 640 |
+
"loss": 0.3544,
|
| 641 |
+
"step": 800
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 9.191489361702128,
|
| 645 |
+
"grad_norm": 1.648060940441488,
|
| 646 |
+
"learning_rate": 5e-06,
|
| 647 |
+
"loss": 0.3044,
|
| 648 |
+
"step": 810
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 9.304964539007091,
|
| 652 |
+
"grad_norm": 1.6084806059069598,
|
| 653 |
+
"learning_rate": 5e-06,
|
| 654 |
+
"loss": 0.3059,
|
| 655 |
+
"step": 820
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 9.418439716312056,
|
| 659 |
+
"grad_norm": 1.5401041921688723,
|
| 660 |
+
"learning_rate": 5e-06,
|
| 661 |
+
"loss": 0.3058,
|
| 662 |
+
"step": 830
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 9.53191489361702,
|
| 666 |
+
"grad_norm": 2.0293807391338143,
|
| 667 |
+
"learning_rate": 5e-06,
|
| 668 |
+
"loss": 0.3115,
|
| 669 |
+
"step": 840
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 9.645390070921986,
|
| 673 |
+
"grad_norm": 2.314645079025493,
|
| 674 |
+
"learning_rate": 5e-06,
|
| 675 |
+
"loss": 0.3145,
|
| 676 |
+
"step": 850
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 9.75886524822695,
|
| 680 |
+
"grad_norm": 1.5570050085916418,
|
| 681 |
+
"learning_rate": 5e-06,
|
| 682 |
+
"loss": 0.3106,
|
| 683 |
+
"step": 860
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 9.872340425531915,
|
| 687 |
+
"grad_norm": 1.5311554302680037,
|
| 688 |
+
"learning_rate": 5e-06,
|
| 689 |
+
"loss": 0.3138,
|
| 690 |
+
"step": 870
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 9.98581560283688,
|
| 694 |
+
"grad_norm": 1.5287839268465846,
|
| 695 |
+
"learning_rate": 5e-06,
|
| 696 |
+
"loss": 0.3192,
|
| 697 |
+
"step": 880
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 9.997163120567375,
|
| 701 |
+
"eval_loss": 1.1392488479614258,
|
| 702 |
+
"eval_runtime": 62.4865,
|
| 703 |
+
"eval_samples_per_second": 37.96,
|
| 704 |
+
"eval_steps_per_second": 0.608,
|
| 705 |
+
"step": 881
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 10.099290780141844,
|
| 709 |
+
"grad_norm": 2.72751141563365,
|
| 710 |
+
"learning_rate": 5e-06,
|
| 711 |
+
"loss": 0.2913,
|
| 712 |
+
"step": 890
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 10.212765957446809,
|
| 716 |
+
"grad_norm": 1.695112625785905,
|
| 717 |
+
"learning_rate": 5e-06,
|
| 718 |
+
"loss": 0.2564,
|
| 719 |
+
"step": 900
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 10.326241134751774,
|
| 723 |
+
"grad_norm": 2.3429971574606965,
|
| 724 |
+
"learning_rate": 5e-06,
|
| 725 |
+
"loss": 0.2557,
|
| 726 |
+
"step": 910
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 10.439716312056738,
|
| 730 |
+
"grad_norm": 1.8027953487068278,
|
| 731 |
+
"learning_rate": 5e-06,
|
| 732 |
+
"loss": 0.2598,
|
| 733 |
+
"step": 920
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 10.553191489361701,
|
| 737 |
+
"grad_norm": 1.638203817216446,
|
| 738 |
+
"learning_rate": 5e-06,
|
| 739 |
+
"loss": 0.2629,
|
| 740 |
+
"step": 930
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 10.666666666666666,
|
| 744 |
+
"grad_norm": 1.7806426687502075,
|
| 745 |
+
"learning_rate": 5e-06,
|
| 746 |
+
"loss": 0.2634,
|
| 747 |
+
"step": 940
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 10.78014184397163,
|
| 751 |
+
"grad_norm": 1.5926893822981327,
|
| 752 |
+
"learning_rate": 5e-06,
|
| 753 |
+
"loss": 0.2638,
|
| 754 |
+
"step": 950
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 10.893617021276595,
|
| 758 |
+
"grad_norm": 1.7428329081390457,
|
| 759 |
+
"learning_rate": 5e-06,
|
| 760 |
+
"loss": 0.2658,
|
| 761 |
+
"step": 960
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 10.995744680851065,
|
| 765 |
+
"eval_loss": 1.251742959022522,
|
| 766 |
+
"eval_runtime": 61.8255,
|
| 767 |
+
"eval_samples_per_second": 38.366,
|
| 768 |
+
"eval_steps_per_second": 0.615,
|
| 769 |
+
"step": 969
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 11.00709219858156,
|
| 773 |
+
"grad_norm": 3.9263311549935636,
|
| 774 |
+
"learning_rate": 5e-06,
|
| 775 |
+
"loss": 0.2848,
|
| 776 |
+
"step": 970
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 11.120567375886525,
|
| 780 |
+
"grad_norm": 2.6049326617817044,
|
| 781 |
+
"learning_rate": 5e-06,
|
| 782 |
+
"loss": 0.2138,
|
| 783 |
+
"step": 980
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 11.23404255319149,
|
| 787 |
+
"grad_norm": 1.6222017763370542,
|
| 788 |
+
"learning_rate": 5e-06,
|
| 789 |
+
"loss": 0.213,
|
| 790 |
+
"step": 990
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 11.347517730496454,
|
| 794 |
+
"grad_norm": 2.5363164551454944,
|
| 795 |
+
"learning_rate": 5e-06,
|
| 796 |
+
"loss": 0.2145,
|
| 797 |
+
"step": 1000
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 11.460992907801419,
|
| 801 |
+
"grad_norm": 1.4653575698277383,
|
| 802 |
+
"learning_rate": 5e-06,
|
| 803 |
+
"loss": 0.2126,
|
| 804 |
+
"step": 1010
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 11.574468085106384,
|
| 808 |
+
"grad_norm": 1.661578038944481,
|
| 809 |
+
"learning_rate": 5e-06,
|
| 810 |
+
"loss": 0.2199,
|
| 811 |
+
"step": 1020
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 11.687943262411348,
|
| 815 |
+
"grad_norm": 1.6170075184374815,
|
| 816 |
+
"learning_rate": 5e-06,
|
| 817 |
+
"loss": 0.216,
|
| 818 |
+
"step": 1030
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 11.801418439716311,
|
| 822 |
+
"grad_norm": 2.84154531134498,
|
| 823 |
+
"learning_rate": 5e-06,
|
| 824 |
+
"loss": 0.2248,
|
| 825 |
+
"step": 1040
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 11.914893617021276,
|
| 829 |
+
"grad_norm": 2.304374165266374,
|
| 830 |
+
"learning_rate": 5e-06,
|
| 831 |
+
"loss": 0.2232,
|
| 832 |
+
"step": 1050
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 11.994326241134752,
|
| 836 |
+
"eval_loss": 1.3437931537628174,
|
| 837 |
+
"eval_runtime": 62.6563,
|
| 838 |
+
"eval_samples_per_second": 37.857,
|
| 839 |
+
"eval_steps_per_second": 0.606,
|
| 840 |
+
"step": 1057
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 12.02836879432624,
|
| 844 |
+
"grad_norm": 3.171954263474723,
|
| 845 |
+
"learning_rate": 5e-06,
|
| 846 |
+
"loss": 0.2283,
|
| 847 |
+
"step": 1060
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 12.141843971631205,
|
| 851 |
+
"grad_norm": 2.2055268906349044,
|
| 852 |
+
"learning_rate": 5e-06,
|
| 853 |
+
"loss": 0.1732,
|
| 854 |
+
"step": 1070
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 12.25531914893617,
|
| 858 |
+
"grad_norm": 1.966744769081506,
|
| 859 |
+
"learning_rate": 5e-06,
|
| 860 |
+
"loss": 0.1702,
|
| 861 |
+
"step": 1080
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 12.368794326241135,
|
| 865 |
+
"grad_norm": 1.4350271294458334,
|
| 866 |
+
"learning_rate": 5e-06,
|
| 867 |
+
"loss": 0.1708,
|
| 868 |
+
"step": 1090
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 12.4822695035461,
|
| 872 |
+
"grad_norm": 2.1908087303160437,
|
| 873 |
+
"learning_rate": 5e-06,
|
| 874 |
+
"loss": 0.1716,
|
| 875 |
+
"step": 1100
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 12.595744680851064,
|
| 879 |
+
"grad_norm": 2.1524139714858364,
|
| 880 |
+
"learning_rate": 5e-06,
|
| 881 |
+
"loss": 0.1769,
|
| 882 |
+
"step": 1110
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 12.709219858156029,
|
| 886 |
+
"grad_norm": 2.315412876413917,
|
| 887 |
+
"learning_rate": 5e-06,
|
| 888 |
+
"loss": 0.1774,
|
| 889 |
+
"step": 1120
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 12.822695035460994,
|
| 893 |
+
"grad_norm": 1.779835602481138,
|
| 894 |
+
"learning_rate": 5e-06,
|
| 895 |
+
"loss": 0.1818,
|
| 896 |
+
"step": 1130
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 12.936170212765958,
|
| 900 |
+
"grad_norm": 1.5338508372566295,
|
| 901 |
+
"learning_rate": 5e-06,
|
| 902 |
+
"loss": 0.1817,
|
| 903 |
+
"step": 1140
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 12.99290780141844,
|
| 907 |
+
"eval_loss": 1.4415781497955322,
|
| 908 |
+
"eval_runtime": 62.5739,
|
| 909 |
+
"eval_samples_per_second": 37.907,
|
| 910 |
+
"eval_steps_per_second": 0.607,
|
| 911 |
+
"step": 1145
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 13.049645390070921,
|
| 915 |
+
"grad_norm": 2.4012185669865347,
|
| 916 |
+
"learning_rate": 5e-06,
|
| 917 |
+
"loss": 0.1775,
|
| 918 |
+
"step": 1150
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 13.163120567375886,
|
| 922 |
+
"grad_norm": 2.6033021021980955,
|
| 923 |
+
"learning_rate": 5e-06,
|
| 924 |
+
"loss": 0.1378,
|
| 925 |
+
"step": 1160
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 13.27659574468085,
|
| 929 |
+
"grad_norm": 1.70623847553698,
|
| 930 |
+
"learning_rate": 5e-06,
|
| 931 |
+
"loss": 0.1358,
|
| 932 |
+
"step": 1170
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 13.390070921985815,
|
| 936 |
+
"grad_norm": 2.2305608008154665,
|
| 937 |
+
"learning_rate": 5e-06,
|
| 938 |
+
"loss": 0.1364,
|
| 939 |
+
"step": 1180
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 13.50354609929078,
|
| 943 |
+
"grad_norm": 2.7083160759097016,
|
| 944 |
+
"learning_rate": 5e-06,
|
| 945 |
+
"loss": 0.137,
|
| 946 |
+
"step": 1190
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 13.617021276595745,
|
| 950 |
+
"grad_norm": 2.69713739478201,
|
| 951 |
+
"learning_rate": 5e-06,
|
| 952 |
+
"loss": 0.1398,
|
| 953 |
+
"step": 1200
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 13.73049645390071,
|
| 957 |
+
"grad_norm": 2.3436045100684475,
|
| 958 |
+
"learning_rate": 5e-06,
|
| 959 |
+
"loss": 0.1399,
|
| 960 |
+
"step": 1210
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 13.843971631205674,
|
| 964 |
+
"grad_norm": 1.6050917746136801,
|
| 965 |
+
"learning_rate": 5e-06,
|
| 966 |
+
"loss": 0.1412,
|
| 967 |
+
"step": 1220
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 13.957446808510639,
|
| 971 |
+
"grad_norm": 2.5589880236873546,
|
| 972 |
+
"learning_rate": 5e-06,
|
| 973 |
+
"loss": 0.1418,
|
| 974 |
+
"step": 1230
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 13.991489361702127,
|
| 978 |
+
"eval_loss": 1.5400274991989136,
|
| 979 |
+
"eval_runtime": 62.1433,
|
| 980 |
+
"eval_samples_per_second": 38.17,
|
| 981 |
+
"eval_steps_per_second": 0.611,
|
| 982 |
+
"step": 1233
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 14.070921985815604,
|
| 986 |
+
"grad_norm": 2.6974199938795715,
|
| 987 |
+
"learning_rate": 5e-06,
|
| 988 |
+
"loss": 0.1335,
|
| 989 |
+
"step": 1240
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 14.184397163120567,
|
| 993 |
+
"grad_norm": 1.669747755191835,
|
| 994 |
+
"learning_rate": 5e-06,
|
| 995 |
+
"loss": 0.1045,
|
| 996 |
+
"step": 1250
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 14.297872340425531,
|
| 1000 |
+
"grad_norm": 1.6701536146823392,
|
| 1001 |
+
"learning_rate": 5e-06,
|
| 1002 |
+
"loss": 0.1047,
|
| 1003 |
+
"step": 1260
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 14.411347517730496,
|
| 1007 |
+
"grad_norm": 1.5692309327841514,
|
| 1008 |
+
"learning_rate": 5e-06,
|
| 1009 |
+
"loss": 0.1051,
|
| 1010 |
+
"step": 1270
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 14.52482269503546,
|
| 1014 |
+
"grad_norm": 1.5854537259008437,
|
| 1015 |
+
"learning_rate": 5e-06,
|
| 1016 |
+
"loss": 0.1092,
|
| 1017 |
+
"step": 1280
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 14.638297872340425,
|
| 1021 |
+
"grad_norm": 2.1977634811831757,
|
| 1022 |
+
"learning_rate": 5e-06,
|
| 1023 |
+
"loss": 0.1103,
|
| 1024 |
+
"step": 1290
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 14.75177304964539,
|
| 1028 |
+
"grad_norm": 1.4935460100019156,
|
| 1029 |
+
"learning_rate": 5e-06,
|
| 1030 |
+
"loss": 0.112,
|
| 1031 |
+
"step": 1300
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 14.865248226950355,
|
| 1035 |
+
"grad_norm": 2.2338875049474316,
|
| 1036 |
+
"learning_rate": 5e-06,
|
| 1037 |
+
"loss": 0.114,
|
| 1038 |
+
"step": 1310
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 14.97872340425532,
|
| 1042 |
+
"grad_norm": 1.6452083463902718,
|
| 1043 |
+
"learning_rate": 5e-06,
|
| 1044 |
+
"loss": 0.1144,
|
| 1045 |
+
"step": 1320
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 14.990070921985815,
|
| 1049 |
+
"eval_loss": 1.6749097108840942,
|
| 1050 |
+
"eval_runtime": 69.4669,
|
| 1051 |
+
"eval_samples_per_second": 34.146,
|
| 1052 |
+
"eval_steps_per_second": 0.547,
|
| 1053 |
+
"step": 1321
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 15.092198581560284,
|
| 1057 |
+
"grad_norm": 2.9909865909950932,
|
| 1058 |
+
"learning_rate": 5e-06,
|
| 1059 |
+
"loss": 0.0978,
|
| 1060 |
+
"step": 1330
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 15.205673758865249,
|
| 1064 |
+
"grad_norm": 1.753525186875363,
|
| 1065 |
+
"learning_rate": 5e-06,
|
| 1066 |
+
"loss": 0.0819,
|
| 1067 |
+
"step": 1340
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 15.319148936170214,
|
| 1071 |
+
"grad_norm": 1.49718571453873,
|
| 1072 |
+
"learning_rate": 5e-06,
|
| 1073 |
+
"loss": 0.0822,
|
| 1074 |
+
"step": 1350
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 15.432624113475176,
|
| 1078 |
+
"grad_norm": 1.5460276195217901,
|
| 1079 |
+
"learning_rate": 5e-06,
|
| 1080 |
+
"loss": 0.0842,
|
| 1081 |
+
"step": 1360
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 15.546099290780141,
|
| 1085 |
+
"grad_norm": 2.2302829005472993,
|
| 1086 |
+
"learning_rate": 5e-06,
|
| 1087 |
+
"loss": 0.0846,
|
| 1088 |
+
"step": 1370
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 15.659574468085106,
|
| 1092 |
+
"grad_norm": 1.8466734512848562,
|
| 1093 |
+
"learning_rate": 5e-06,
|
| 1094 |
+
"loss": 0.0854,
|
| 1095 |
+
"step": 1380
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 15.77304964539007,
|
| 1099 |
+
"grad_norm": 1.826155456821189,
|
| 1100 |
+
"learning_rate": 5e-06,
|
| 1101 |
+
"loss": 0.0859,
|
| 1102 |
+
"step": 1390
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 15.886524822695035,
|
| 1106 |
+
"grad_norm": 1.7480925191865653,
|
| 1107 |
+
"learning_rate": 5e-06,
|
| 1108 |
+
"loss": 0.0873,
|
| 1109 |
+
"step": 1400
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 16.0,
|
| 1113 |
+
"grad_norm": 4.703406049536698,
|
| 1114 |
+
"learning_rate": 5e-06,
|
| 1115 |
+
"loss": 0.0932,
|
| 1116 |
+
"step": 1410
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 16.0,
|
| 1120 |
+
"eval_loss": 1.773341178894043,
|
| 1121 |
+
"eval_runtime": 69.8689,
|
| 1122 |
+
"eval_samples_per_second": 33.949,
|
| 1123 |
+
"eval_steps_per_second": 0.544,
|
| 1124 |
+
"step": 1410
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 16.113475177304963,
|
| 1128 |
+
"grad_norm": 2.1931332653452653,
|
| 1129 |
+
"learning_rate": 5e-06,
|
| 1130 |
+
"loss": 0.0618,
|
| 1131 |
+
"step": 1420
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 16.22695035460993,
|
| 1135 |
+
"grad_norm": 1.561959511527612,
|
| 1136 |
+
"learning_rate": 5e-06,
|
| 1137 |
+
"loss": 0.0616,
|
| 1138 |
+
"step": 1430
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 16.340425531914892,
|
| 1142 |
+
"grad_norm": 2.08368980727909,
|
| 1143 |
+
"learning_rate": 5e-06,
|
| 1144 |
+
"loss": 0.0618,
|
| 1145 |
+
"step": 1440
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 16.45390070921986,
|
| 1149 |
+
"grad_norm": 2.633735186380944,
|
| 1150 |
+
"learning_rate": 5e-06,
|
| 1151 |
+
"loss": 0.0632,
|
| 1152 |
+
"step": 1450
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 16.56737588652482,
|
| 1156 |
+
"grad_norm": 1.6666433686469015,
|
| 1157 |
+
"learning_rate": 5e-06,
|
| 1158 |
+
"loss": 0.0628,
|
| 1159 |
+
"step": 1460
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 16.680851063829788,
|
| 1163 |
+
"grad_norm": 1.5125810239930932,
|
| 1164 |
+
"learning_rate": 5e-06,
|
| 1165 |
+
"loss": 0.063,
|
| 1166 |
+
"step": 1470
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 16.79432624113475,
|
| 1170 |
+
"grad_norm": 2.2573637438259464,
|
| 1171 |
+
"learning_rate": 5e-06,
|
| 1172 |
+
"loss": 0.0648,
|
| 1173 |
+
"step": 1480
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 16.907801418439718,
|
| 1177 |
+
"grad_norm": 1.766704303804737,
|
| 1178 |
+
"learning_rate": 5e-06,
|
| 1179 |
+
"loss": 0.0669,
|
| 1180 |
+
"step": 1490
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 16.99858156028369,
|
| 1184 |
+
"eval_loss": 1.9060131311416626,
|
| 1185 |
+
"eval_runtime": 69.1764,
|
| 1186 |
+
"eval_samples_per_second": 34.289,
|
| 1187 |
+
"eval_steps_per_second": 0.549,
|
| 1188 |
+
"step": 1498
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 17.02127659574468,
|
| 1192 |
+
"grad_norm": 2.8884446061839686,
|
| 1193 |
+
"learning_rate": 5e-06,
|
| 1194 |
+
"loss": 0.067,
|
| 1195 |
+
"step": 1500
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 17.134751773049647,
|
| 1199 |
+
"grad_norm": 1.4375657511058568,
|
| 1200 |
+
"learning_rate": 5e-06,
|
| 1201 |
+
"loss": 0.045,
|
| 1202 |
+
"step": 1510
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 17.24822695035461,
|
| 1206 |
+
"grad_norm": 1.4802152522135432,
|
| 1207 |
+
"learning_rate": 5e-06,
|
| 1208 |
+
"loss": 0.0447,
|
| 1209 |
+
"step": 1520
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 17.361702127659573,
|
| 1213 |
+
"grad_norm": 1.614881692739521,
|
| 1214 |
+
"learning_rate": 5e-06,
|
| 1215 |
+
"loss": 0.0473,
|
| 1216 |
+
"step": 1530
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 17.47517730496454,
|
| 1220 |
+
"grad_norm": 2.4617340461694472,
|
| 1221 |
+
"learning_rate": 5e-06,
|
| 1222 |
+
"loss": 0.0491,
|
| 1223 |
+
"step": 1540
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 17.588652482269502,
|
| 1227 |
+
"grad_norm": 1.4499102037181577,
|
| 1228 |
+
"learning_rate": 5e-06,
|
| 1229 |
+
"loss": 0.0483,
|
| 1230 |
+
"step": 1550
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 17.70212765957447,
|
| 1234 |
+
"grad_norm": 1.6979894468070986,
|
| 1235 |
+
"learning_rate": 5e-06,
|
| 1236 |
+
"loss": 0.0489,
|
| 1237 |
+
"step": 1560
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 17.81560283687943,
|
| 1241 |
+
"grad_norm": 2.081624941686834,
|
| 1242 |
+
"learning_rate": 5e-06,
|
| 1243 |
+
"loss": 0.0501,
|
| 1244 |
+
"step": 1570
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 17.929078014184398,
|
| 1248 |
+
"grad_norm": 1.9921542957081264,
|
| 1249 |
+
"learning_rate": 5e-06,
|
| 1250 |
+
"loss": 0.0506,
|
| 1251 |
+
"step": 1580
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 17.997163120567375,
|
| 1255 |
+
"eval_loss": 1.9450756311416626,
|
| 1256 |
+
"eval_runtime": 68.8214,
|
| 1257 |
+
"eval_samples_per_second": 34.466,
|
| 1258 |
+
"eval_steps_per_second": 0.552,
|
| 1259 |
+
"step": 1586
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 18.04255319148936,
|
| 1263 |
+
"grad_norm": 1.9224844379964918,
|
| 1264 |
+
"learning_rate": 5e-06,
|
| 1265 |
+
"loss": 0.0503,
|
| 1266 |
+
"step": 1590
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 18.156028368794328,
|
| 1270 |
+
"grad_norm": 1.714407414612998,
|
| 1271 |
+
"learning_rate": 5e-06,
|
| 1272 |
+
"loss": 0.0354,
|
| 1273 |
+
"step": 1600
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 18.26950354609929,
|
| 1277 |
+
"grad_norm": 1.380425999820701,
|
| 1278 |
+
"learning_rate": 5e-06,
|
| 1279 |
+
"loss": 0.036,
|
| 1280 |
+
"step": 1610
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 18.382978723404257,
|
| 1284 |
+
"grad_norm": 1.5291382013616857,
|
| 1285 |
+
"learning_rate": 5e-06,
|
| 1286 |
+
"loss": 0.0386,
|
| 1287 |
+
"step": 1620
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 18.49645390070922,
|
| 1291 |
+
"grad_norm": 2.2541583387536037,
|
| 1292 |
+
"learning_rate": 5e-06,
|
| 1293 |
+
"loss": 0.0399,
|
| 1294 |
+
"step": 1630
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 18.609929078014183,
|
| 1298 |
+
"grad_norm": 1.7584069820913808,
|
| 1299 |
+
"learning_rate": 5e-06,
|
| 1300 |
+
"loss": 0.0407,
|
| 1301 |
+
"step": 1640
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 18.72340425531915,
|
| 1305 |
+
"grad_norm": 1.7575163459677958,
|
| 1306 |
+
"learning_rate": 5e-06,
|
| 1307 |
+
"loss": 0.0408,
|
| 1308 |
+
"step": 1650
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 18.836879432624112,
|
| 1312 |
+
"grad_norm": 1.4369028755815154,
|
| 1313 |
+
"learning_rate": 5e-06,
|
| 1314 |
+
"loss": 0.04,
|
| 1315 |
+
"step": 1660
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 18.95035460992908,
|
| 1319 |
+
"grad_norm": 1.9850726261264064,
|
| 1320 |
+
"learning_rate": 5e-06,
|
| 1321 |
+
"loss": 0.0412,
|
| 1322 |
+
"step": 1670
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 18.995744680851065,
|
| 1326 |
+
"eval_loss": 2.018162727355957,
|
| 1327 |
+
"eval_runtime": 62.4156,
|
| 1328 |
+
"eval_samples_per_second": 38.003,
|
| 1329 |
+
"eval_steps_per_second": 0.609,
|
| 1330 |
+
"step": 1674
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 19.06382978723404,
|
| 1334 |
+
"grad_norm": 1.501965463597226,
|
| 1335 |
+
"learning_rate": 5e-06,
|
| 1336 |
+
"loss": 0.0382,
|
| 1337 |
+
"step": 1680
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 19.177304964539008,
|
| 1341 |
+
"grad_norm": 1.952580410332778,
|
| 1342 |
+
"learning_rate": 5e-06,
|
| 1343 |
+
"loss": 0.0306,
|
| 1344 |
+
"step": 1690
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 19.29078014184397,
|
| 1348 |
+
"grad_norm": 1.729911113753904,
|
| 1349 |
+
"learning_rate": 5e-06,
|
| 1350 |
+
"loss": 0.0306,
|
| 1351 |
+
"step": 1700
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 19.404255319148938,
|
| 1355 |
+
"grad_norm": 1.5141444330801705,
|
| 1356 |
+
"learning_rate": 5e-06,
|
| 1357 |
+
"loss": 0.0304,
|
| 1358 |
+
"step": 1710
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 19.5177304964539,
|
| 1362 |
+
"grad_norm": 1.3870058186319074,
|
| 1363 |
+
"learning_rate": 5e-06,
|
| 1364 |
+
"loss": 0.031,
|
| 1365 |
+
"step": 1720
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 19.631205673758867,
|
| 1369 |
+
"grad_norm": 1.8134623582061118,
|
| 1370 |
+
"learning_rate": 5e-06,
|
| 1371 |
+
"loss": 0.0321,
|
| 1372 |
+
"step": 1730
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 19.74468085106383,
|
| 1376 |
+
"grad_norm": 1.699089186796663,
|
| 1377 |
+
"learning_rate": 5e-06,
|
| 1378 |
+
"loss": 0.034,
|
| 1379 |
+
"step": 1740
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 19.858156028368793,
|
| 1383 |
+
"grad_norm": 1.3806789087183744,
|
| 1384 |
+
"learning_rate": 5e-06,
|
| 1385 |
+
"loss": 0.034,
|
| 1386 |
+
"step": 1750
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 19.97163120567376,
|
| 1390 |
+
"grad_norm": 1.4266442973936173,
|
| 1391 |
+
"learning_rate": 5e-06,
|
| 1392 |
+
"loss": 0.0336,
|
| 1393 |
+
"step": 1760
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 19.99432624113475,
|
| 1397 |
+
"eval_loss": 2.0948634147644043,
|
| 1398 |
+
"eval_runtime": 62.5678,
|
| 1399 |
+
"eval_samples_per_second": 37.911,
|
| 1400 |
+
"eval_steps_per_second": 0.607,
|
| 1401 |
+
"step": 1762
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 20.085106382978722,
|
| 1405 |
+
"grad_norm": 1.920078997696489,
|
| 1406 |
+
"learning_rate": 5e-06,
|
| 1407 |
+
"loss": 0.0294,
|
| 1408 |
+
"step": 1770
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 20.19858156028369,
|
| 1412 |
+
"grad_norm": 1.5002721536366188,
|
| 1413 |
+
"learning_rate": 5e-06,
|
| 1414 |
+
"loss": 0.0258,
|
| 1415 |
+
"step": 1780
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 20.31205673758865,
|
| 1419 |
+
"grad_norm": 1.5579738724358545,
|
| 1420 |
+
"learning_rate": 5e-06,
|
| 1421 |
+
"loss": 0.0267,
|
| 1422 |
+
"step": 1790
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 20.425531914893618,
|
| 1426 |
+
"grad_norm": 1.455169431793061,
|
| 1427 |
+
"learning_rate": 5e-06,
|
| 1428 |
+
"loss": 0.0266,
|
| 1429 |
+
"step": 1800
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 20.53900709219858,
|
| 1433 |
+
"grad_norm": 1.22886010057641,
|
| 1434 |
+
"learning_rate": 5e-06,
|
| 1435 |
+
"loss": 0.0269,
|
| 1436 |
+
"step": 1810
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 20.652482269503547,
|
| 1440 |
+
"grad_norm": 1.5116925302103017,
|
| 1441 |
+
"learning_rate": 5e-06,
|
| 1442 |
+
"loss": 0.0282,
|
| 1443 |
+
"step": 1820
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 20.76595744680851,
|
| 1447 |
+
"grad_norm": 1.5692932225328788,
|
| 1448 |
+
"learning_rate": 5e-06,
|
| 1449 |
+
"loss": 0.0288,
|
| 1450 |
+
"step": 1830
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 20.879432624113477,
|
| 1454 |
+
"grad_norm": 1.391741629484137,
|
| 1455 |
+
"learning_rate": 5e-06,
|
| 1456 |
+
"loss": 0.0298,
|
| 1457 |
+
"step": 1840
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 20.99290780141844,
|
| 1461 |
+
"grad_norm": 1.973700678085544,
|
| 1462 |
+
"learning_rate": 5e-06,
|
| 1463 |
+
"loss": 0.0299,
|
| 1464 |
+
"step": 1850
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 20.99290780141844,
|
| 1468 |
+
"eval_loss": 2.1437060832977295,
|
| 1469 |
+
"eval_runtime": 62.0314,
|
| 1470 |
+
"eval_samples_per_second": 38.239,
|
| 1471 |
+
"eval_steps_per_second": 0.613,
|
| 1472 |
+
"step": 1850
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 21.106382978723403,
|
| 1476 |
+
"grad_norm": 1.1791411901063809,
|
| 1477 |
+
"learning_rate": 5e-06,
|
| 1478 |
+
"loss": 0.0244,
|
| 1479 |
+
"step": 1860
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 21.21985815602837,
|
| 1483 |
+
"grad_norm": 1.267123446553761,
|
| 1484 |
+
"learning_rate": 5e-06,
|
| 1485 |
+
"loss": 0.0226,
|
| 1486 |
+
"step": 1870
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 21.333333333333332,
|
| 1490 |
+
"grad_norm": 1.3624423761722013,
|
| 1491 |
+
"learning_rate": 5e-06,
|
| 1492 |
+
"loss": 0.023,
|
| 1493 |
+
"step": 1880
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 21.4468085106383,
|
| 1497 |
+
"grad_norm": 1.426640337405648,
|
| 1498 |
+
"learning_rate": 5e-06,
|
| 1499 |
+
"loss": 0.0243,
|
| 1500 |
+
"step": 1890
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 21.56028368794326,
|
| 1504 |
+
"grad_norm": 1.4553540262096338,
|
| 1505 |
+
"learning_rate": 5e-06,
|
| 1506 |
+
"loss": 0.0242,
|
| 1507 |
+
"step": 1900
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 21.673758865248228,
|
| 1511 |
+
"grad_norm": 1.503724635353548,
|
| 1512 |
+
"learning_rate": 5e-06,
|
| 1513 |
+
"loss": 0.0247,
|
| 1514 |
+
"step": 1910
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 21.78723404255319,
|
| 1518 |
+
"grad_norm": 1.4126151297962473,
|
| 1519 |
+
"learning_rate": 5e-06,
|
| 1520 |
+
"loss": 0.0255,
|
| 1521 |
+
"step": 1920
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 21.900709219858157,
|
| 1525 |
+
"grad_norm": 1.2923917999452894,
|
| 1526 |
+
"learning_rate": 5e-06,
|
| 1527 |
+
"loss": 0.0255,
|
| 1528 |
+
"step": 1930
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 21.99148936170213,
|
| 1532 |
+
"eval_loss": 2.1743686199188232,
|
| 1533 |
+
"eval_runtime": 61.954,
|
| 1534 |
+
"eval_samples_per_second": 38.286,
|
| 1535 |
+
"eval_steps_per_second": 0.613,
|
| 1536 |
+
"step": 1938
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 22.01418439716312,
|
| 1540 |
+
"grad_norm": 1.2651742889807782,
|
| 1541 |
+
"learning_rate": 5e-06,
|
| 1542 |
+
"loss": 0.0257,
|
| 1543 |
+
"step": 1940
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 22.127659574468087,
|
| 1547 |
+
"grad_norm": 1.1234871648540001,
|
| 1548 |
+
"learning_rate": 5e-06,
|
| 1549 |
+
"loss": 0.0183,
|
| 1550 |
+
"step": 1950
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 22.24113475177305,
|
| 1554 |
+
"grad_norm": 1.1914682347381635,
|
| 1555 |
+
"learning_rate": 5e-06,
|
| 1556 |
+
"loss": 0.0197,
|
| 1557 |
+
"step": 1960
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 22.354609929078013,
|
| 1561 |
+
"grad_norm": 1.5236491827381375,
|
| 1562 |
+
"learning_rate": 5e-06,
|
| 1563 |
+
"loss": 0.0203,
|
| 1564 |
+
"step": 1970
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 22.46808510638298,
|
| 1568 |
+
"grad_norm": 1.152425475304377,
|
| 1569 |
+
"learning_rate": 5e-06,
|
| 1570 |
+
"loss": 0.0201,
|
| 1571 |
+
"step": 1980
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 22.581560283687942,
|
| 1575 |
+
"grad_norm": 1.5074453816966662,
|
| 1576 |
+
"learning_rate": 5e-06,
|
| 1577 |
+
"loss": 0.0203,
|
| 1578 |
+
"step": 1990
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 22.69503546099291,
|
| 1582 |
+
"grad_norm": 1.4232083123752879,
|
| 1583 |
+
"learning_rate": 5e-06,
|
| 1584 |
+
"loss": 0.02,
|
| 1585 |
+
"step": 2000
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 22.80851063829787,
|
| 1589 |
+
"grad_norm": 1.3274591200293286,
|
| 1590 |
+
"learning_rate": 5e-06,
|
| 1591 |
+
"loss": 0.0205,
|
| 1592 |
+
"step": 2010
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 22.921985815602838,
|
| 1596 |
+
"grad_norm": 1.2163483760719238,
|
| 1597 |
+
"learning_rate": 5e-06,
|
| 1598 |
+
"loss": 0.0214,
|
| 1599 |
+
"step": 2020
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 22.990070921985815,
|
| 1603 |
+
"eval_loss": 2.2530956268310547,
|
| 1604 |
+
"eval_runtime": 61.8444,
|
| 1605 |
+
"eval_samples_per_second": 38.354,
|
| 1606 |
+
"eval_steps_per_second": 0.614,
|
| 1607 |
+
"step": 2026
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 23.0354609929078,
|
| 1611 |
+
"grad_norm": 1.102674013054533,
|
| 1612 |
+
"learning_rate": 5e-06,
|
| 1613 |
+
"loss": 0.0211,
|
| 1614 |
+
"step": 2030
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 23.148936170212767,
|
| 1618 |
+
"grad_norm": 1.280352404689559,
|
| 1619 |
+
"learning_rate": 5e-06,
|
| 1620 |
+
"loss": 0.0166,
|
| 1621 |
+
"step": 2040
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 23.26241134751773,
|
| 1625 |
+
"grad_norm": 1.2676632223563031,
|
| 1626 |
+
"learning_rate": 5e-06,
|
| 1627 |
+
"loss": 0.0174,
|
| 1628 |
+
"step": 2050
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 23.375886524822697,
|
| 1632 |
+
"grad_norm": 1.338911086977905,
|
| 1633 |
+
"learning_rate": 5e-06,
|
| 1634 |
+
"loss": 0.0179,
|
| 1635 |
+
"step": 2060
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 23.48936170212766,
|
| 1639 |
+
"grad_norm": 1.3586671769170666,
|
| 1640 |
+
"learning_rate": 5e-06,
|
| 1641 |
+
"loss": 0.018,
|
| 1642 |
+
"step": 2070
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 23.602836879432623,
|
| 1646 |
+
"grad_norm": 1.6478284268488168,
|
| 1647 |
+
"learning_rate": 5e-06,
|
| 1648 |
+
"loss": 0.0182,
|
| 1649 |
+
"step": 2080
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 23.71631205673759,
|
| 1653 |
+
"grad_norm": 1.3656912406981414,
|
| 1654 |
+
"learning_rate": 5e-06,
|
| 1655 |
+
"loss": 0.0191,
|
| 1656 |
+
"step": 2090
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 23.829787234042552,
|
| 1660 |
+
"grad_norm": 1.2548808655348251,
|
| 1661 |
+
"learning_rate": 5e-06,
|
| 1662 |
+
"loss": 0.0182,
|
| 1663 |
+
"step": 2100
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 23.94326241134752,
|
| 1667 |
+
"grad_norm": 1.2709938993829644,
|
| 1668 |
+
"learning_rate": 5e-06,
|
| 1669 |
+
"loss": 0.0183,
|
| 1670 |
+
"step": 2110
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 24.0,
|
| 1674 |
+
"eval_loss": 2.2672340869903564,
|
| 1675 |
+
"eval_runtime": 61.7779,
|
| 1676 |
+
"eval_samples_per_second": 38.396,
|
| 1677 |
+
"eval_steps_per_second": 0.615,
|
| 1678 |
+
"step": 2115
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 24.05673758865248,
|
| 1682 |
+
"grad_norm": 1.1287652816162261,
|
| 1683 |
+
"learning_rate": 5e-06,
|
| 1684 |
+
"loss": 0.0176,
|
| 1685 |
+
"step": 2120
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 24.170212765957448,
|
| 1689 |
+
"grad_norm": 1.514856403878364,
|
| 1690 |
+
"learning_rate": 5e-06,
|
| 1691 |
+
"loss": 0.0149,
|
| 1692 |
+
"step": 2130
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 24.28368794326241,
|
| 1696 |
+
"grad_norm": 1.4283915780055492,
|
| 1697 |
+
"learning_rate": 5e-06,
|
| 1698 |
+
"loss": 0.0156,
|
| 1699 |
+
"step": 2140
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 24.397163120567377,
|
| 1703 |
+
"grad_norm": 1.059461534861265,
|
| 1704 |
+
"learning_rate": 5e-06,
|
| 1705 |
+
"loss": 0.0159,
|
| 1706 |
+
"step": 2150
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 24.51063829787234,
|
| 1710 |
+
"grad_norm": 1.1623165952765633,
|
| 1711 |
+
"learning_rate": 5e-06,
|
| 1712 |
+
"loss": 0.0161,
|
| 1713 |
+
"step": 2160
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 24.624113475177303,
|
| 1717 |
+
"grad_norm": 1.117151249059667,
|
| 1718 |
+
"learning_rate": 5e-06,
|
| 1719 |
+
"loss": 0.0167,
|
| 1720 |
+
"step": 2170
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 24.73758865248227,
|
| 1724 |
+
"grad_norm": 1.3585899962321324,
|
| 1725 |
+
"learning_rate": 5e-06,
|
| 1726 |
+
"loss": 0.0173,
|
| 1727 |
+
"step": 2180
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 24.851063829787233,
|
| 1731 |
+
"grad_norm": 1.2416797002653372,
|
| 1732 |
+
"learning_rate": 5e-06,
|
| 1733 |
+
"loss": 0.0176,
|
| 1734 |
+
"step": 2190
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 24.9645390070922,
|
| 1738 |
+
"grad_norm": 1.1197862040396234,
|
| 1739 |
+
"learning_rate": 5e-06,
|
| 1740 |
+
"loss": 0.0176,
|
| 1741 |
+
"step": 2200
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 24.99858156028369,
|
| 1745 |
+
"eval_loss": 2.2650251388549805,
|
| 1746 |
+
"eval_runtime": 61.7487,
|
| 1747 |
+
"eval_samples_per_second": 38.414,
|
| 1748 |
+
"eval_steps_per_second": 0.615,
|
| 1749 |
+
"step": 2203
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 25.078014184397162,
|
| 1753 |
+
"grad_norm": 0.9878141556093745,
|
| 1754 |
+
"learning_rate": 5e-06,
|
| 1755 |
+
"loss": 0.016,
|
| 1756 |
+
"step": 2210
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 25.19148936170213,
|
| 1760 |
+
"grad_norm": 1.1817739397428935,
|
| 1761 |
+
"learning_rate": 5e-06,
|
| 1762 |
+
"loss": 0.0145,
|
| 1763 |
+
"step": 2220
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 25.30496453900709,
|
| 1767 |
+
"grad_norm": 1.004904053519453,
|
| 1768 |
+
"learning_rate": 5e-06,
|
| 1769 |
+
"loss": 0.0152,
|
| 1770 |
+
"step": 2230
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 25.418439716312058,
|
| 1774 |
+
"grad_norm": 1.120987937715426,
|
| 1775 |
+
"learning_rate": 5e-06,
|
| 1776 |
+
"loss": 0.0153,
|
| 1777 |
+
"step": 2240
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 25.53191489361702,
|
| 1781 |
+
"grad_norm": 1.1581554681983095,
|
| 1782 |
+
"learning_rate": 5e-06,
|
| 1783 |
+
"loss": 0.0158,
|
| 1784 |
+
"step": 2250
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 25.645390070921987,
|
| 1788 |
+
"grad_norm": 0.9570504069445702,
|
| 1789 |
+
"learning_rate": 5e-06,
|
| 1790 |
+
"loss": 0.0159,
|
| 1791 |
+
"step": 2260
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 25.75886524822695,
|
| 1795 |
+
"grad_norm": 1.2223868766654897,
|
| 1796 |
+
"learning_rate": 5e-06,
|
| 1797 |
+
"loss": 0.0167,
|
| 1798 |
+
"step": 2270
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 25.872340425531917,
|
| 1802 |
+
"grad_norm": 1.235670723966881,
|
| 1803 |
+
"learning_rate": 5e-06,
|
| 1804 |
+
"loss": 0.0167,
|
| 1805 |
+
"step": 2280
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 25.98581560283688,
|
| 1809 |
+
"grad_norm": 1.0449436637521454,
|
| 1810 |
+
"learning_rate": 5e-06,
|
| 1811 |
+
"loss": 0.0165,
|
| 1812 |
+
"step": 2290
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 25.997163120567375,
|
| 1816 |
+
"eval_loss": 2.2784597873687744,
|
| 1817 |
+
"eval_runtime": 61.747,
|
| 1818 |
+
"eval_samples_per_second": 38.415,
|
| 1819 |
+
"eval_steps_per_second": 0.615,
|
| 1820 |
+
"step": 2291
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 26.099290780141843,
|
| 1824 |
+
"grad_norm": 0.937351445077704,
|
| 1825 |
+
"learning_rate": 5e-06,
|
| 1826 |
+
"loss": 0.0144,
|
| 1827 |
+
"step": 2300
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 26.21276595744681,
|
| 1831 |
+
"grad_norm": 1.4726294234141475,
|
| 1832 |
+
"learning_rate": 5e-06,
|
| 1833 |
+
"loss": 0.0138,
|
| 1834 |
+
"step": 2310
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 26.326241134751772,
|
| 1838 |
+
"grad_norm": 0.911974339928837,
|
| 1839 |
+
"learning_rate": 5e-06,
|
| 1840 |
+
"loss": 0.0141,
|
| 1841 |
+
"step": 2320
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 26.43971631205674,
|
| 1845 |
+
"grad_norm": 0.9153128396132686,
|
| 1846 |
+
"learning_rate": 5e-06,
|
| 1847 |
+
"loss": 0.0141,
|
| 1848 |
+
"step": 2330
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 26.5531914893617,
|
| 1852 |
+
"grad_norm": 1.0379804645600088,
|
| 1853 |
+
"learning_rate": 5e-06,
|
| 1854 |
+
"loss": 0.0146,
|
| 1855 |
+
"step": 2340
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 26.666666666666668,
|
| 1859 |
+
"grad_norm": 1.009567426911364,
|
| 1860 |
+
"learning_rate": 5e-06,
|
| 1861 |
+
"loss": 0.015,
|
| 1862 |
+
"step": 2350
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 26.78014184397163,
|
| 1866 |
+
"grad_norm": 0.9581577917417519,
|
| 1867 |
+
"learning_rate": 5e-06,
|
| 1868 |
+
"loss": 0.0155,
|
| 1869 |
+
"step": 2360
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 26.893617021276597,
|
| 1873 |
+
"grad_norm": 0.9973777175459498,
|
| 1874 |
+
"learning_rate": 5e-06,
|
| 1875 |
+
"loss": 0.0152,
|
| 1876 |
+
"step": 2370
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 26.995744680851065,
|
| 1880 |
+
"eval_loss": 2.272617816925049,
|
| 1881 |
+
"eval_runtime": 69.2573,
|
| 1882 |
+
"eval_samples_per_second": 34.249,
|
| 1883 |
+
"eval_steps_per_second": 0.549,
|
| 1884 |
+
"step": 2379
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 27.00709219858156,
|
| 1888 |
+
"grad_norm": 2.5151039373227637,
|
| 1889 |
+
"learning_rate": 5e-06,
|
| 1890 |
+
"loss": 0.0162,
|
| 1891 |
+
"step": 2380
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 27.120567375886523,
|
| 1895 |
+
"grad_norm": 0.8739287942149531,
|
| 1896 |
+
"learning_rate": 5e-06,
|
| 1897 |
+
"loss": 0.0119,
|
| 1898 |
+
"step": 2390
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 27.23404255319149,
|
| 1902 |
+
"grad_norm": 0.9923332279176865,
|
| 1903 |
+
"learning_rate": 5e-06,
|
| 1904 |
+
"loss": 0.0123,
|
| 1905 |
+
"step": 2400
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 27.347517730496453,
|
| 1909 |
+
"grad_norm": 0.8986384187935266,
|
| 1910 |
+
"learning_rate": 5e-06,
|
| 1911 |
+
"loss": 0.0123,
|
| 1912 |
+
"step": 2410
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 27.46099290780142,
|
| 1916 |
+
"grad_norm": 0.9527555500748222,
|
| 1917 |
+
"learning_rate": 5e-06,
|
| 1918 |
+
"loss": 0.0128,
|
| 1919 |
+
"step": 2420
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 27.574468085106382,
|
| 1923 |
+
"grad_norm": 0.9573410638147425,
|
| 1924 |
+
"learning_rate": 5e-06,
|
| 1925 |
+
"loss": 0.0131,
|
| 1926 |
+
"step": 2430
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 27.68794326241135,
|
| 1930 |
+
"grad_norm": 0.9206613579635727,
|
| 1931 |
+
"learning_rate": 5e-06,
|
| 1932 |
+
"loss": 0.0134,
|
| 1933 |
+
"step": 2440
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 27.80141843971631,
|
| 1937 |
+
"grad_norm": 1.0631805922440314,
|
| 1938 |
+
"learning_rate": 5e-06,
|
| 1939 |
+
"loss": 0.0136,
|
| 1940 |
+
"step": 2450
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 27.914893617021278,
|
| 1944 |
+
"grad_norm": 0.9873569774398828,
|
| 1945 |
+
"learning_rate": 5e-06,
|
| 1946 |
+
"loss": 0.0141,
|
| 1947 |
+
"step": 2460
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 27.99432624113475,
|
| 1951 |
+
"eval_loss": 2.3099989891052246,
|
| 1952 |
+
"eval_runtime": 61.8228,
|
| 1953 |
+
"eval_samples_per_second": 38.368,
|
| 1954 |
+
"eval_steps_per_second": 0.615,
|
| 1955 |
+
"step": 2467
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 28.02836879432624,
|
| 1959 |
+
"grad_norm": 0.8367467896199314,
|
| 1960 |
+
"learning_rate": 5e-06,
|
| 1961 |
+
"loss": 0.0142,
|
| 1962 |
+
"step": 2470
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 28.141843971631207,
|
| 1966 |
+
"grad_norm": 0.8750575092503129,
|
| 1967 |
+
"learning_rate": 5e-06,
|
| 1968 |
+
"loss": 0.011,
|
| 1969 |
+
"step": 2480
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 28.25531914893617,
|
| 1973 |
+
"grad_norm": 0.8433351422057557,
|
| 1974 |
+
"learning_rate": 5e-06,
|
| 1975 |
+
"loss": 0.0113,
|
| 1976 |
+
"step": 2490
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 28.368794326241133,
|
| 1980 |
+
"grad_norm": 0.9459082278215922,
|
| 1981 |
+
"learning_rate": 5e-06,
|
| 1982 |
+
"loss": 0.0114,
|
| 1983 |
+
"step": 2500
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 28.4822695035461,
|
| 1987 |
+
"grad_norm": 0.8476815269716437,
|
| 1988 |
+
"learning_rate": 5e-06,
|
| 1989 |
+
"loss": 0.0118,
|
| 1990 |
+
"step": 2510
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 28.595744680851062,
|
| 1994 |
+
"grad_norm": 0.9919619479253982,
|
| 1995 |
+
"learning_rate": 5e-06,
|
| 1996 |
+
"loss": 0.0119,
|
| 1997 |
+
"step": 2520
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 28.70921985815603,
|
| 2001 |
+
"grad_norm": 0.921713687988904,
|
| 2002 |
+
"learning_rate": 5e-06,
|
| 2003 |
+
"loss": 0.0119,
|
| 2004 |
+
"step": 2530
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 28.822695035460992,
|
| 2008 |
+
"grad_norm": 0.8904411861697781,
|
| 2009 |
+
"learning_rate": 5e-06,
|
| 2010 |
+
"loss": 0.012,
|
| 2011 |
+
"step": 2540
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 28.93617021276596,
|
| 2015 |
+
"grad_norm": 1.2704463003827144,
|
| 2016 |
+
"learning_rate": 5e-06,
|
| 2017 |
+
"loss": 0.0124,
|
| 2018 |
+
"step": 2550
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 28.99290780141844,
|
| 2022 |
+
"eval_loss": 2.332291841506958,
|
| 2023 |
+
"eval_runtime": 60.798,
|
| 2024 |
+
"eval_samples_per_second": 39.014,
|
| 2025 |
+
"eval_steps_per_second": 0.625,
|
| 2026 |
+
"step": 2555
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 29.04964539007092,
|
| 2030 |
+
"grad_norm": 0.8345079136380029,
|
| 2031 |
+
"learning_rate": 5e-06,
|
| 2032 |
+
"loss": 0.0116,
|
| 2033 |
+
"step": 2560
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 29.163120567375888,
|
| 2037 |
+
"grad_norm": 0.7727575462289974,
|
| 2038 |
+
"learning_rate": 5e-06,
|
| 2039 |
+
"loss": 0.0098,
|
| 2040 |
+
"step": 2570
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 29.27659574468085,
|
| 2044 |
+
"grad_norm": 0.7449623009916486,
|
| 2045 |
+
"learning_rate": 5e-06,
|
| 2046 |
+
"loss": 0.01,
|
| 2047 |
+
"step": 2580
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 29.390070921985817,
|
| 2051 |
+
"grad_norm": 0.823872924226129,
|
| 2052 |
+
"learning_rate": 5e-06,
|
| 2053 |
+
"loss": 0.01,
|
| 2054 |
+
"step": 2590
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 29.50354609929078,
|
| 2058 |
+
"grad_norm": 0.8397207934188576,
|
| 2059 |
+
"learning_rate": 5e-06,
|
| 2060 |
+
"loss": 0.0099,
|
| 2061 |
+
"step": 2600
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 29.617021276595743,
|
| 2065 |
+
"grad_norm": 0.8116256567720525,
|
| 2066 |
+
"learning_rate": 5e-06,
|
| 2067 |
+
"loss": 0.0102,
|
| 2068 |
+
"step": 2610
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 29.73049645390071,
|
| 2072 |
+
"grad_norm": 0.8294250388534566,
|
| 2073 |
+
"learning_rate": 5e-06,
|
| 2074 |
+
"loss": 0.0102,
|
| 2075 |
+
"step": 2620
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 29.843971631205672,
|
| 2079 |
+
"grad_norm": 0.8114849444813739,
|
| 2080 |
+
"learning_rate": 5e-06,
|
| 2081 |
+
"loss": 0.0104,
|
| 2082 |
+
"step": 2630
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 29.95744680851064,
|
| 2086 |
+
"grad_norm": 0.8852232975531666,
|
| 2087 |
+
"learning_rate": 5e-06,
|
| 2088 |
+
"loss": 0.0106,
|
| 2089 |
+
"step": 2640
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 29.99148936170213,
|
| 2093 |
+
"eval_loss": 2.357081651687622,
|
| 2094 |
+
"eval_runtime": 62.2636,
|
| 2095 |
+
"eval_samples_per_second": 38.096,
|
| 2096 |
+
"eval_steps_per_second": 0.61,
|
| 2097 |
+
"step": 2643
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 30.070921985815602,
|
| 2101 |
+
"grad_norm": 0.7031992810809015,
|
| 2102 |
+
"learning_rate": 5e-06,
|
| 2103 |
+
"loss": 0.0096,
|
| 2104 |
+
"step": 2650
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 30.18439716312057,
|
| 2108 |
+
"grad_norm": 0.7627672024963855,
|
| 2109 |
+
"learning_rate": 5e-06,
|
| 2110 |
+
"loss": 0.0084,
|
| 2111 |
+
"step": 2660
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 30.29787234042553,
|
| 2115 |
+
"grad_norm": 0.7110603755577746,
|
| 2116 |
+
"learning_rate": 5e-06,
|
| 2117 |
+
"loss": 0.0084,
|
| 2118 |
+
"step": 2670
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 30.411347517730498,
|
| 2122 |
+
"grad_norm": 0.7420329356605581,
|
| 2123 |
+
"learning_rate": 5e-06,
|
| 2124 |
+
"loss": 0.0085,
|
| 2125 |
+
"step": 2680
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 30.52482269503546,
|
| 2129 |
+
"grad_norm": 0.7808701527194697,
|
| 2130 |
+
"learning_rate": 5e-06,
|
| 2131 |
+
"loss": 0.0085,
|
| 2132 |
+
"step": 2690
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 30.638297872340427,
|
| 2136 |
+
"grad_norm": 0.7874064236769093,
|
| 2137 |
+
"learning_rate": 5e-06,
|
| 2138 |
+
"loss": 0.0085,
|
| 2139 |
+
"step": 2700
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 30.75177304964539,
|
| 2143 |
+
"grad_norm": 0.7661676156662441,
|
| 2144 |
+
"learning_rate": 5e-06,
|
| 2145 |
+
"loss": 0.0087,
|
| 2146 |
+
"step": 2710
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 30.865248226950353,
|
| 2150 |
+
"grad_norm": 0.8535116914325056,
|
| 2151 |
+
"learning_rate": 5e-06,
|
| 2152 |
+
"loss": 0.0088,
|
| 2153 |
+
"step": 2720
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 30.97872340425532,
|
| 2157 |
+
"grad_norm": 0.7958972876528424,
|
| 2158 |
+
"learning_rate": 5e-06,
|
| 2159 |
+
"loss": 0.0091,
|
| 2160 |
+
"step": 2730
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 30.990070921985815,
|
| 2164 |
+
"eval_loss": 2.4116153717041016,
|
| 2165 |
+
"eval_runtime": 62.2832,
|
| 2166 |
+
"eval_samples_per_second": 38.084,
|
| 2167 |
+
"eval_steps_per_second": 0.61,
|
| 2168 |
+
"step": 2731
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 31.092198581560282,
|
| 2172 |
+
"grad_norm": 0.6485885567051449,
|
| 2173 |
+
"learning_rate": 5e-06,
|
| 2174 |
+
"loss": 0.0078,
|
| 2175 |
+
"step": 2740
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 31.20567375886525,
|
| 2179 |
+
"grad_norm": 0.6541844022555007,
|
| 2180 |
+
"learning_rate": 5e-06,
|
| 2181 |
+
"loss": 0.0072,
|
| 2182 |
+
"step": 2750
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 31.319148936170212,
|
| 2186 |
+
"grad_norm": 0.776841660846075,
|
| 2187 |
+
"learning_rate": 5e-06,
|
| 2188 |
+
"loss": 0.0073,
|
| 2189 |
+
"step": 2760
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 31.43262411347518,
|
| 2193 |
+
"grad_norm": 0.7710605909492212,
|
| 2194 |
+
"learning_rate": 5e-06,
|
| 2195 |
+
"loss": 0.0072,
|
| 2196 |
+
"step": 2770
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 31.54609929078014,
|
| 2200 |
+
"grad_norm": 0.6816480202732877,
|
| 2201 |
+
"learning_rate": 5e-06,
|
| 2202 |
+
"loss": 0.0072,
|
| 2203 |
+
"step": 2780
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 31.659574468085108,
|
| 2207 |
+
"grad_norm": 0.8400344753440614,
|
| 2208 |
+
"learning_rate": 5e-06,
|
| 2209 |
+
"loss": 0.0073,
|
| 2210 |
+
"step": 2790
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 31.77304964539007,
|
| 2214 |
+
"grad_norm": 0.7577115728922557,
|
| 2215 |
+
"learning_rate": 5e-06,
|
| 2216 |
+
"loss": 0.0076,
|
| 2217 |
+
"step": 2800
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 31.886524822695037,
|
| 2221 |
+
"grad_norm": 0.8366427457618262,
|
| 2222 |
+
"learning_rate": 5e-06,
|
| 2223 |
+
"loss": 0.0077,
|
| 2224 |
+
"step": 2810
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 32.0,
|
| 2228 |
+
"grad_norm": 1.7417184938662238,
|
| 2229 |
+
"learning_rate": 5e-06,
|
| 2230 |
+
"loss": 0.0083,
|
| 2231 |
+
"step": 2820
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 32.0,
|
| 2235 |
+
"eval_loss": 2.5119211673736572,
|
| 2236 |
+
"eval_runtime": 62.2182,
|
| 2237 |
+
"eval_samples_per_second": 38.124,
|
| 2238 |
+
"eval_steps_per_second": 0.611,
|
| 2239 |
+
"step": 2820
|
| 2240 |
+
},
|
| 2241 |
+
{
|
| 2242 |
+
"epoch": 32.11347517730496,
|
| 2243 |
+
"grad_norm": 0.6760402035549387,
|
| 2244 |
+
"learning_rate": 5e-06,
|
| 2245 |
+
"loss": 0.006,
|
| 2246 |
+
"step": 2830
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 32.226950354609926,
|
| 2250 |
+
"grad_norm": 0.6247844370602216,
|
| 2251 |
+
"learning_rate": 5e-06,
|
| 2252 |
+
"loss": 0.0064,
|
| 2253 |
+
"step": 2840
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 32.340425531914896,
|
| 2257 |
+
"grad_norm": 0.8298072821546006,
|
| 2258 |
+
"learning_rate": 5e-06,
|
| 2259 |
+
"loss": 0.0065,
|
| 2260 |
+
"step": 2850
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 32.45390070921986,
|
| 2264 |
+
"grad_norm": 0.6225777220105869,
|
| 2265 |
+
"learning_rate": 5e-06,
|
| 2266 |
+
"loss": 0.0064,
|
| 2267 |
+
"step": 2860
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 32.56737588652482,
|
| 2271 |
+
"grad_norm": 0.6622379639401021,
|
| 2272 |
+
"learning_rate": 5e-06,
|
| 2273 |
+
"loss": 0.0064,
|
| 2274 |
+
"step": 2870
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 32.680851063829785,
|
| 2278 |
+
"grad_norm": 0.6622926501947951,
|
| 2279 |
+
"learning_rate": 5e-06,
|
| 2280 |
+
"loss": 0.0065,
|
| 2281 |
+
"step": 2880
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 32.794326241134755,
|
| 2285 |
+
"grad_norm": 0.6348291026878606,
|
| 2286 |
+
"learning_rate": 5e-06,
|
| 2287 |
+
"loss": 0.0072,
|
| 2288 |
+
"step": 2890
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 32.90780141843972,
|
| 2292 |
+
"grad_norm": 2.0690919155019616,
|
| 2293 |
+
"learning_rate": 5e-06,
|
| 2294 |
+
"loss": 0.0071,
|
| 2295 |
+
"step": 2900
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 32.99858156028369,
|
| 2299 |
+
"eval_loss": 2.4598617553710938,
|
| 2300 |
+
"eval_runtime": 61.935,
|
| 2301 |
+
"eval_samples_per_second": 38.298,
|
| 2302 |
+
"eval_steps_per_second": 0.614,
|
| 2303 |
+
"step": 2908
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 33.02127659574468,
|
| 2307 |
+
"grad_norm": 0.536551007465182,
|
| 2308 |
+
"learning_rate": 5e-06,
|
| 2309 |
+
"loss": 0.0072,
|
| 2310 |
+
"step": 2910
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 33.13475177304964,
|
| 2314 |
+
"grad_norm": 0.6265202665061125,
|
| 2315 |
+
"learning_rate": 5e-06,
|
| 2316 |
+
"loss": 0.0055,
|
| 2317 |
+
"step": 2920
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 33.248226950354606,
|
| 2321 |
+
"grad_norm": 0.6066247337398882,
|
| 2322 |
+
"learning_rate": 5e-06,
|
| 2323 |
+
"loss": 0.0061,
|
| 2324 |
+
"step": 2930
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 33.361702127659576,
|
| 2328 |
+
"grad_norm": 0.6329886038923753,
|
| 2329 |
+
"learning_rate": 5e-06,
|
| 2330 |
+
"loss": 0.0062,
|
| 2331 |
+
"step": 2940
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 33.47517730496454,
|
| 2335 |
+
"grad_norm": 0.6434289034891333,
|
| 2336 |
+
"learning_rate": 5e-06,
|
| 2337 |
+
"loss": 0.006,
|
| 2338 |
+
"step": 2950
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 33.5886524822695,
|
| 2342 |
+
"grad_norm": 0.6060277242963108,
|
| 2343 |
+
"learning_rate": 5e-06,
|
| 2344 |
+
"loss": 0.0062,
|
| 2345 |
+
"step": 2960
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 33.702127659574465,
|
| 2349 |
+
"grad_norm": 0.6231645801668283,
|
| 2350 |
+
"learning_rate": 5e-06,
|
| 2351 |
+
"loss": 0.0065,
|
| 2352 |
+
"step": 2970
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 33.815602836879435,
|
| 2356 |
+
"grad_norm": 0.6726293815659068,
|
| 2357 |
+
"learning_rate": 5e-06,
|
| 2358 |
+
"loss": 0.0062,
|
| 2359 |
+
"step": 2980
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 33.9290780141844,
|
| 2363 |
+
"grad_norm": 0.7144217421088135,
|
| 2364 |
+
"learning_rate": 5e-06,
|
| 2365 |
+
"loss": 0.0066,
|
| 2366 |
+
"step": 2990
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 33.99716312056738,
|
| 2370 |
+
"eval_loss": 2.4769325256347656,
|
| 2371 |
+
"eval_runtime": 60.7622,
|
| 2372 |
+
"eval_samples_per_second": 39.037,
|
| 2373 |
+
"eval_steps_per_second": 0.625,
|
| 2374 |
+
"step": 2996
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 34.04255319148936,
|
| 2378 |
+
"grad_norm": 0.640917218147658,
|
| 2379 |
+
"learning_rate": 5e-06,
|
| 2380 |
+
"loss": 0.0066,
|
| 2381 |
+
"step": 3000
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 34.156028368794324,
|
| 2385 |
+
"grad_norm": 0.5848948305959304,
|
| 2386 |
+
"learning_rate": 5e-06,
|
| 2387 |
+
"loss": 0.0059,
|
| 2388 |
+
"step": 3010
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 34.269503546099294,
|
| 2392 |
+
"grad_norm": 0.6093130739371995,
|
| 2393 |
+
"learning_rate": 5e-06,
|
| 2394 |
+
"loss": 0.0056,
|
| 2395 |
+
"step": 3020
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 34.38297872340426,
|
| 2399 |
+
"grad_norm": 0.6028806624019981,
|
| 2400 |
+
"learning_rate": 5e-06,
|
| 2401 |
+
"loss": 0.0058,
|
| 2402 |
+
"step": 3030
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 34.49645390070922,
|
| 2406 |
+
"grad_norm": 0.5872374541059426,
|
| 2407 |
+
"learning_rate": 5e-06,
|
| 2408 |
+
"loss": 0.0057,
|
| 2409 |
+
"step": 3040
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 34.60992907801418,
|
| 2413 |
+
"grad_norm": 0.6073657131269332,
|
| 2414 |
+
"learning_rate": 5e-06,
|
| 2415 |
+
"loss": 0.0057,
|
| 2416 |
+
"step": 3050
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 34.723404255319146,
|
| 2420 |
+
"grad_norm": 0.586833117804181,
|
| 2421 |
+
"learning_rate": 5e-06,
|
| 2422 |
+
"loss": 0.0058,
|
| 2423 |
+
"step": 3060
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 34.836879432624116,
|
| 2427 |
+
"grad_norm": 0.6606622175333071,
|
| 2428 |
+
"learning_rate": 5e-06,
|
| 2429 |
+
"loss": 0.0064,
|
| 2430 |
+
"step": 3070
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 34.95035460992908,
|
| 2434 |
+
"grad_norm": 0.6279271205274822,
|
| 2435 |
+
"learning_rate": 5e-06,
|
| 2436 |
+
"loss": 0.0061,
|
| 2437 |
+
"step": 3080
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 34.99574468085106,
|
| 2441 |
+
"eval_loss": 2.4636948108673096,
|
| 2442 |
+
"eval_runtime": 62.1741,
|
| 2443 |
+
"eval_samples_per_second": 38.151,
|
| 2444 |
+
"eval_steps_per_second": 0.611,
|
| 2445 |
+
"step": 3084
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 35.06382978723404,
|
| 2449 |
+
"grad_norm": 0.5598672191700598,
|
| 2450 |
+
"learning_rate": 5e-06,
|
| 2451 |
+
"loss": 0.0059,
|
| 2452 |
+
"step": 3090
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 35.177304964539005,
|
| 2456 |
+
"grad_norm": 0.5838009342325976,
|
| 2457 |
+
"learning_rate": 5e-06,
|
| 2458 |
+
"loss": 0.0052,
|
| 2459 |
+
"step": 3100
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 35.290780141843975,
|
| 2463 |
+
"grad_norm": 0.5591904839383254,
|
| 2464 |
+
"learning_rate": 5e-06,
|
| 2465 |
+
"loss": 0.0055,
|
| 2466 |
+
"step": 3110
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 35.40425531914894,
|
| 2470 |
+
"grad_norm": 0.52651340630151,
|
| 2471 |
+
"learning_rate": 5e-06,
|
| 2472 |
+
"loss": 0.0052,
|
| 2473 |
+
"step": 3120
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 35.5177304964539,
|
| 2477 |
+
"grad_norm": 0.5557733635066537,
|
| 2478 |
+
"learning_rate": 5e-06,
|
| 2479 |
+
"loss": 0.0054,
|
| 2480 |
+
"step": 3130
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 35.63120567375886,
|
| 2484 |
+
"grad_norm": 0.5555621169894426,
|
| 2485 |
+
"learning_rate": 5e-06,
|
| 2486 |
+
"loss": 0.0056,
|
| 2487 |
+
"step": 3140
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 35.744680851063826,
|
| 2491 |
+
"grad_norm": 0.6174053410534251,
|
| 2492 |
+
"learning_rate": 5e-06,
|
| 2493 |
+
"loss": 0.006,
|
| 2494 |
+
"step": 3150
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 35.858156028368796,
|
| 2498 |
+
"grad_norm": 0.6219808794998958,
|
| 2499 |
+
"learning_rate": 5e-06,
|
| 2500 |
+
"loss": 0.0058,
|
| 2501 |
+
"step": 3160
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 35.97163120567376,
|
| 2505 |
+
"grad_norm": 0.6453593970939066,
|
| 2506 |
+
"learning_rate": 5e-06,
|
| 2507 |
+
"loss": 0.0059,
|
| 2508 |
+
"step": 3170
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 35.99432624113475,
|
| 2512 |
+
"eval_loss": 2.446837902069092,
|
| 2513 |
+
"eval_runtime": 70.4261,
|
| 2514 |
+
"eval_samples_per_second": 33.681,
|
| 2515 |
+
"eval_steps_per_second": 0.54,
|
| 2516 |
+
"step": 3172
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 36.08510638297872,
|
| 2520 |
+
"grad_norm": 0.5743613529876505,
|
| 2521 |
+
"learning_rate": 5e-06,
|
| 2522 |
+
"loss": 0.0054,
|
| 2523 |
+
"step": 3180
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 36.198581560283685,
|
| 2527 |
+
"grad_norm": 0.5822243660764591,
|
| 2528 |
+
"learning_rate": 5e-06,
|
| 2529 |
+
"loss": 0.0053,
|
| 2530 |
+
"step": 3190
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 36.312056737588655,
|
| 2534 |
+
"grad_norm": 0.6021189745310633,
|
| 2535 |
+
"learning_rate": 5e-06,
|
| 2536 |
+
"loss": 0.0052,
|
| 2537 |
+
"step": 3200
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 36.42553191489362,
|
| 2541 |
+
"grad_norm": 0.5853179594306633,
|
| 2542 |
+
"learning_rate": 5e-06,
|
| 2543 |
+
"loss": 0.0053,
|
| 2544 |
+
"step": 3210
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 36.53900709219858,
|
| 2548 |
+
"grad_norm": 0.5676622286224795,
|
| 2549 |
+
"learning_rate": 5e-06,
|
| 2550 |
+
"loss": 0.0056,
|
| 2551 |
+
"step": 3220
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 36.652482269503544,
|
| 2555 |
+
"grad_norm": 0.6378528653719511,
|
| 2556 |
+
"learning_rate": 5e-06,
|
| 2557 |
+
"loss": 0.0055,
|
| 2558 |
+
"step": 3230
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 36.765957446808514,
|
| 2562 |
+
"grad_norm": 0.5795559134784821,
|
| 2563 |
+
"learning_rate": 5e-06,
|
| 2564 |
+
"loss": 0.0054,
|
| 2565 |
+
"step": 3240
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 36.87943262411348,
|
| 2569 |
+
"grad_norm": 0.5739478607744279,
|
| 2570 |
+
"learning_rate": 5e-06,
|
| 2571 |
+
"loss": 0.0056,
|
| 2572 |
+
"step": 3250
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 36.99290780141844,
|
| 2576 |
+
"grad_norm": 0.6761276352174411,
|
| 2577 |
+
"learning_rate": 5e-06,
|
| 2578 |
+
"loss": 0.0058,
|
| 2579 |
+
"step": 3260
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 36.99290780141844,
|
| 2583 |
+
"eval_loss": 2.4386816024780273,
|
| 2584 |
+
"eval_runtime": 64.5541,
|
| 2585 |
+
"eval_samples_per_second": 36.744,
|
| 2586 |
+
"eval_steps_per_second": 0.589,
|
| 2587 |
+
"step": 3260
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 37.1063829787234,
|
| 2591 |
+
"grad_norm": 0.5639138084508524,
|
| 2592 |
+
"learning_rate": 5e-06,
|
| 2593 |
+
"loss": 0.0052,
|
| 2594 |
+
"step": 3270
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 37.219858156028366,
|
| 2598 |
+
"grad_norm": 0.5552500627772322,
|
| 2599 |
+
"learning_rate": 5e-06,
|
| 2600 |
+
"loss": 0.0051,
|
| 2601 |
+
"step": 3280
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 37.333333333333336,
|
| 2605 |
+
"grad_norm": 0.5505999304380521,
|
| 2606 |
+
"learning_rate": 5e-06,
|
| 2607 |
+
"loss": 0.0052,
|
| 2608 |
+
"step": 3290
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 37.4468085106383,
|
| 2612 |
+
"grad_norm": 0.6290595858485541,
|
| 2613 |
+
"learning_rate": 5e-06,
|
| 2614 |
+
"loss": 0.0053,
|
| 2615 |
+
"step": 3300
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 37.56028368794326,
|
| 2619 |
+
"grad_norm": 0.6689930337019874,
|
| 2620 |
+
"learning_rate": 5e-06,
|
| 2621 |
+
"loss": 0.0052,
|
| 2622 |
+
"step": 3310
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 37.673758865248224,
|
| 2626 |
+
"grad_norm": 0.6361115119336014,
|
| 2627 |
+
"learning_rate": 5e-06,
|
| 2628 |
+
"loss": 0.0053,
|
| 2629 |
+
"step": 3320
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 37.787234042553195,
|
| 2633 |
+
"grad_norm": 0.5758613903708901,
|
| 2634 |
+
"learning_rate": 5e-06,
|
| 2635 |
+
"loss": 0.0057,
|
| 2636 |
+
"step": 3330
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 37.90070921985816,
|
| 2640 |
+
"grad_norm": 0.5488463771978056,
|
| 2641 |
+
"learning_rate": 5e-06,
|
| 2642 |
+
"loss": 0.0056,
|
| 2643 |
+
"step": 3340
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 37.99148936170213,
|
| 2647 |
+
"eval_loss": 2.409132480621338,
|
| 2648 |
+
"eval_runtime": 61.4398,
|
| 2649 |
+
"eval_samples_per_second": 38.607,
|
| 2650 |
+
"eval_steps_per_second": 0.618,
|
| 2651 |
+
"step": 3348
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 38.01418439716312,
|
| 2655 |
+
"grad_norm": 0.502740899332166,
|
| 2656 |
+
"learning_rate": 5e-06,
|
| 2657 |
+
"loss": 0.006,
|
| 2658 |
+
"step": 3350
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 38.12765957446808,
|
| 2662 |
+
"grad_norm": 0.5325341716894662,
|
| 2663 |
+
"learning_rate": 5e-06,
|
| 2664 |
+
"loss": 0.0051,
|
| 2665 |
+
"step": 3360
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 38.241134751773046,
|
| 2669 |
+
"grad_norm": 0.5469915840136621,
|
| 2670 |
+
"learning_rate": 5e-06,
|
| 2671 |
+
"loss": 0.0052,
|
| 2672 |
+
"step": 3370
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 38.354609929078016,
|
| 2676 |
+
"grad_norm": 0.5898577174992149,
|
| 2677 |
+
"learning_rate": 5e-06,
|
| 2678 |
+
"loss": 0.0053,
|
| 2679 |
+
"step": 3380
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 38.46808510638298,
|
| 2683 |
+
"grad_norm": 0.591194838597715,
|
| 2684 |
+
"learning_rate": 5e-06,
|
| 2685 |
+
"loss": 0.0053,
|
| 2686 |
+
"step": 3390
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 38.58156028368794,
|
| 2690 |
+
"grad_norm": 0.5781105524402917,
|
| 2691 |
+
"learning_rate": 5e-06,
|
| 2692 |
+
"loss": 0.0054,
|
| 2693 |
+
"step": 3400
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 38.695035460992905,
|
| 2697 |
+
"grad_norm": 0.5818767708267519,
|
| 2698 |
+
"learning_rate": 5e-06,
|
| 2699 |
+
"loss": 0.0055,
|
| 2700 |
+
"step": 3410
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 38.808510638297875,
|
| 2704 |
+
"grad_norm": 0.6196108940872473,
|
| 2705 |
+
"learning_rate": 5e-06,
|
| 2706 |
+
"loss": 0.0056,
|
| 2707 |
+
"step": 3420
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 38.92198581560284,
|
| 2711 |
+
"grad_norm": 0.6091311582013242,
|
| 2712 |
+
"learning_rate": 5e-06,
|
| 2713 |
+
"loss": 0.0056,
|
| 2714 |
+
"step": 3430
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 38.99007092198582,
|
| 2718 |
+
"eval_loss": 2.4180121421813965,
|
| 2719 |
+
"eval_runtime": 67.9293,
|
| 2720 |
+
"eval_samples_per_second": 34.919,
|
| 2721 |
+
"eval_steps_per_second": 0.559,
|
| 2722 |
+
"step": 3436
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 39.0354609929078,
|
| 2726 |
+
"grad_norm": 0.5203631693062609,
|
| 2727 |
+
"learning_rate": 5e-06,
|
| 2728 |
+
"loss": 0.0059,
|
| 2729 |
+
"step": 3440
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 39.148936170212764,
|
| 2733 |
+
"grad_norm": 0.5220175235403596,
|
| 2734 |
+
"learning_rate": 5e-06,
|
| 2735 |
+
"loss": 0.0054,
|
| 2736 |
+
"step": 3450
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 39.262411347517734,
|
| 2740 |
+
"grad_norm": 0.5306053779378094,
|
| 2741 |
+
"learning_rate": 5e-06,
|
| 2742 |
+
"loss": 0.0051,
|
| 2743 |
+
"step": 3460
|
| 2744 |
+
},
|
| 2745 |
+
{
|
| 2746 |
+
"epoch": 39.3758865248227,
|
| 2747 |
+
"grad_norm": 0.5733275129436737,
|
| 2748 |
+
"learning_rate": 5e-06,
|
| 2749 |
+
"loss": 0.005,
|
| 2750 |
+
"step": 3470
|
| 2751 |
+
},
|
| 2752 |
+
{
|
| 2753 |
+
"epoch": 39.48936170212766,
|
| 2754 |
+
"grad_norm": 0.6322314662873444,
|
| 2755 |
+
"learning_rate": 5e-06,
|
| 2756 |
+
"loss": 0.0053,
|
| 2757 |
+
"step": 3480
|
| 2758 |
+
},
|
| 2759 |
+
{
|
| 2760 |
+
"epoch": 39.60283687943262,
|
| 2761 |
+
"grad_norm": 0.5610008695326577,
|
| 2762 |
+
"learning_rate": 5e-06,
|
| 2763 |
+
"loss": 0.0054,
|
| 2764 |
+
"step": 3490
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"epoch": 39.716312056737586,
|
| 2768 |
+
"grad_norm": 0.6512700300720512,
|
| 2769 |
+
"learning_rate": 5e-06,
|
| 2770 |
+
"loss": 0.0054,
|
| 2771 |
+
"step": 3500
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 39.829787234042556,
|
| 2775 |
+
"grad_norm": 0.6190687455862945,
|
| 2776 |
+
"learning_rate": 5e-06,
|
| 2777 |
+
"loss": 0.0056,
|
| 2778 |
+
"step": 3510
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 39.94326241134752,
|
| 2782 |
+
"grad_norm": 0.6109666894463458,
|
| 2783 |
+
"learning_rate": 5e-06,
|
| 2784 |
+
"loss": 0.0057,
|
| 2785 |
+
"step": 3520
|
| 2786 |
+
},
|
| 2787 |
+
{
|
| 2788 |
+
"epoch": 40.0,
|
| 2789 |
+
"eval_loss": 2.419253349304199,
|
| 2790 |
+
"eval_runtime": 62.0127,
|
| 2791 |
+
"eval_samples_per_second": 38.25,
|
| 2792 |
+
"eval_steps_per_second": 0.613,
|
| 2793 |
+
"step": 3525
|
| 2794 |
+
},
|
| 2795 |
+
{
|
| 2796 |
+
"epoch": 40.05673758865248,
|
| 2797 |
+
"grad_norm": 0.5065903135496583,
|
| 2798 |
+
"learning_rate": 5e-06,
|
| 2799 |
+
"loss": 0.0057,
|
| 2800 |
+
"step": 3530
|
| 2801 |
+
},
|
| 2802 |
+
{
|
| 2803 |
+
"epoch": 40.170212765957444,
|
| 2804 |
+
"grad_norm": 0.5264705355565352,
|
| 2805 |
+
"learning_rate": 5e-06,
|
| 2806 |
+
"loss": 0.0052,
|
| 2807 |
+
"step": 3540
|
| 2808 |
+
},
|
| 2809 |
+
{
|
| 2810 |
+
"epoch": 40.283687943262414,
|
| 2811 |
+
"grad_norm": 0.5771888267651291,
|
| 2812 |
+
"learning_rate": 5e-06,
|
| 2813 |
+
"loss": 0.0052,
|
| 2814 |
+
"step": 3550
|
| 2815 |
+
},
|
| 2816 |
+
{
|
| 2817 |
+
"epoch": 40.39716312056738,
|
| 2818 |
+
"grad_norm": 0.5990984567590389,
|
| 2819 |
+
"learning_rate": 5e-06,
|
| 2820 |
+
"loss": 0.0053,
|
| 2821 |
+
"step": 3560
|
| 2822 |
+
},
|
| 2823 |
+
{
|
| 2824 |
+
"epoch": 40.51063829787234,
|
| 2825 |
+
"grad_norm": 0.5735226625803138,
|
| 2826 |
+
"learning_rate": 5e-06,
|
| 2827 |
+
"loss": 0.0053,
|
| 2828 |
+
"step": 3570
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 40.6241134751773,
|
| 2832 |
+
"grad_norm": 0.6419191932394923,
|
| 2833 |
+
"learning_rate": 5e-06,
|
| 2834 |
+
"loss": 0.0054,
|
| 2835 |
+
"step": 3580
|
| 2836 |
+
},
|
| 2837 |
+
{
|
| 2838 |
+
"epoch": 40.737588652482266,
|
| 2839 |
+
"grad_norm": 0.6128946066765568,
|
| 2840 |
+
"learning_rate": 5e-06,
|
| 2841 |
+
"loss": 0.0054,
|
| 2842 |
+
"step": 3590
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 40.851063829787236,
|
| 2846 |
+
"grad_norm": 0.6089024429918992,
|
| 2847 |
+
"learning_rate": 5e-06,
|
| 2848 |
+
"loss": 0.0058,
|
| 2849 |
+
"step": 3600
|
| 2850 |
+
},
|
| 2851 |
+
{
|
| 2852 |
+
"epoch": 40.9645390070922,
|
| 2853 |
+
"grad_norm": 0.5901415629350402,
|
| 2854 |
+
"learning_rate": 5e-06,
|
| 2855 |
+
"loss": 0.0057,
|
| 2856 |
+
"step": 3610
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"epoch": 40.99858156028369,
|
| 2860 |
+
"eval_loss": 2.4677248001098633,
|
| 2861 |
+
"eval_runtime": 61.1468,
|
| 2862 |
+
"eval_samples_per_second": 38.792,
|
| 2863 |
+
"eval_steps_per_second": 0.621,
|
| 2864 |
+
"step": 3613
|
| 2865 |
+
},
|
| 2866 |
+
{
|
| 2867 |
+
"epoch": 41.07801418439716,
|
| 2868 |
+
"grad_norm": 0.5208524139132839,
|
| 2869 |
+
"learning_rate": 5e-06,
|
| 2870 |
+
"loss": 0.0053,
|
| 2871 |
+
"step": 3620
|
| 2872 |
+
},
|
| 2873 |
+
{
|
| 2874 |
+
"epoch": 41.191489361702125,
|
| 2875 |
+
"grad_norm": 0.5081832637563751,
|
| 2876 |
+
"learning_rate": 5e-06,
|
| 2877 |
+
"loss": 0.0051,
|
| 2878 |
+
"step": 3630
|
| 2879 |
+
},
|
| 2880 |
+
{
|
| 2881 |
+
"epoch": 41.304964539007095,
|
| 2882 |
+
"grad_norm": 0.5335717303413468,
|
| 2883 |
+
"learning_rate": 5e-06,
|
| 2884 |
+
"loss": 0.0053,
|
| 2885 |
+
"step": 3640
|
| 2886 |
+
},
|
| 2887 |
+
{
|
| 2888 |
+
"epoch": 41.41843971631206,
|
| 2889 |
+
"grad_norm": 0.5613593262396614,
|
| 2890 |
+
"learning_rate": 5e-06,
|
| 2891 |
+
"loss": 0.0054,
|
| 2892 |
+
"step": 3650
|
| 2893 |
+
},
|
| 2894 |
+
{
|
| 2895 |
+
"epoch": 41.53191489361702,
|
| 2896 |
+
"grad_norm": 0.6327268082709085,
|
| 2897 |
+
"learning_rate": 5e-06,
|
| 2898 |
+
"loss": 0.0054,
|
| 2899 |
+
"step": 3660
|
| 2900 |
+
},
|
| 2901 |
+
{
|
| 2902 |
+
"epoch": 41.645390070921984,
|
| 2903 |
+
"grad_norm": 0.5361250969249659,
|
| 2904 |
+
"learning_rate": 5e-06,
|
| 2905 |
+
"loss": 0.0054,
|
| 2906 |
+
"step": 3670
|
| 2907 |
+
},
|
| 2908 |
+
{
|
| 2909 |
+
"epoch": 41.758865248226954,
|
| 2910 |
+
"grad_norm": 0.5516534040602991,
|
| 2911 |
+
"learning_rate": 5e-06,
|
| 2912 |
+
"loss": 0.0057,
|
| 2913 |
+
"step": 3680
|
| 2914 |
+
},
|
| 2915 |
+
{
|
| 2916 |
+
"epoch": 41.87234042553192,
|
| 2917 |
+
"grad_norm": 0.6408365010304582,
|
| 2918 |
+
"learning_rate": 5e-06,
|
| 2919 |
+
"loss": 0.0057,
|
| 2920 |
+
"step": 3690
|
| 2921 |
+
},
|
| 2922 |
+
{
|
| 2923 |
+
"epoch": 41.98581560283688,
|
| 2924 |
+
"grad_norm": 0.6048119305492722,
|
| 2925 |
+
"learning_rate": 5e-06,
|
| 2926 |
+
"loss": 0.0058,
|
| 2927 |
+
"step": 3700
|
| 2928 |
+
},
|
| 2929 |
+
{
|
| 2930 |
+
"epoch": 41.99716312056738,
|
| 2931 |
+
"eval_loss": 2.364220142364502,
|
| 2932 |
+
"eval_runtime": 61.7215,
|
| 2933 |
+
"eval_samples_per_second": 38.431,
|
| 2934 |
+
"eval_steps_per_second": 0.616,
|
| 2935 |
+
"step": 3701
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 42.09929078014184,
|
| 2939 |
+
"grad_norm": 0.5693627921008746,
|
| 2940 |
+
"learning_rate": 5e-06,
|
| 2941 |
+
"loss": 0.0054,
|
| 2942 |
+
"step": 3710
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 42.212765957446805,
|
| 2946 |
+
"grad_norm": 0.566082918703353,
|
| 2947 |
+
"learning_rate": 5e-06,
|
| 2948 |
+
"loss": 0.0052,
|
| 2949 |
+
"step": 3720
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 42.326241134751776,
|
| 2953 |
+
"grad_norm": 0.5511507866097272,
|
| 2954 |
+
"learning_rate": 5e-06,
|
| 2955 |
+
"loss": 0.0053,
|
| 2956 |
+
"step": 3730
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 42.43971631205674,
|
| 2960 |
+
"grad_norm": 0.559826817930835,
|
| 2961 |
+
"learning_rate": 5e-06,
|
| 2962 |
+
"loss": 0.0054,
|
| 2963 |
+
"step": 3740
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 42.5531914893617,
|
| 2967 |
+
"grad_norm": 0.58340912778088,
|
| 2968 |
+
"learning_rate": 5e-06,
|
| 2969 |
+
"loss": 0.0054,
|
| 2970 |
+
"step": 3750
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 42.666666666666664,
|
| 2974 |
+
"grad_norm": 0.6002758035231259,
|
| 2975 |
+
"learning_rate": 5e-06,
|
| 2976 |
+
"loss": 0.0055,
|
| 2977 |
+
"step": 3760
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 42.780141843971634,
|
| 2981 |
+
"grad_norm": 0.5989358562823475,
|
| 2982 |
+
"learning_rate": 5e-06,
|
| 2983 |
+
"loss": 0.0056,
|
| 2984 |
+
"step": 3770
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 42.8936170212766,
|
| 2988 |
+
"grad_norm": 0.626190447474167,
|
| 2989 |
+
"learning_rate": 5e-06,
|
| 2990 |
+
"loss": 0.0058,
|
| 2991 |
+
"step": 3780
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 42.99574468085106,
|
| 2995 |
+
"eval_loss": 2.4230844974517822,
|
| 2996 |
+
"eval_runtime": 69.8551,
|
| 2997 |
+
"eval_samples_per_second": 33.956,
|
| 2998 |
+
"eval_steps_per_second": 0.544,
|
| 2999 |
+
"step": 3789
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 43.00709219858156,
|
| 3003 |
+
"grad_norm": 1.2904921052987568,
|
| 3004 |
+
"learning_rate": 5e-06,
|
| 3005 |
+
"loss": 0.006,
|
| 3006 |
+
"step": 3790
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 43.12056737588652,
|
| 3010 |
+
"grad_norm": 0.5946681024606572,
|
| 3011 |
+
"learning_rate": 5e-06,
|
| 3012 |
+
"loss": 0.0051,
|
| 3013 |
+
"step": 3800
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 43.234042553191486,
|
| 3017 |
+
"grad_norm": 0.6247714127627697,
|
| 3018 |
+
"learning_rate": 5e-06,
|
| 3019 |
+
"loss": 0.0053,
|
| 3020 |
+
"step": 3810
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 43.347517730496456,
|
| 3024 |
+
"grad_norm": 0.5677050922672181,
|
| 3025 |
+
"learning_rate": 5e-06,
|
| 3026 |
+
"loss": 0.0054,
|
| 3027 |
+
"step": 3820
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 43.46099290780142,
|
| 3031 |
+
"grad_norm": 0.583683610881685,
|
| 3032 |
+
"learning_rate": 5e-06,
|
| 3033 |
+
"loss": 0.0055,
|
| 3034 |
+
"step": 3830
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 43.57446808510638,
|
| 3038 |
+
"grad_norm": 0.5921120744442437,
|
| 3039 |
+
"learning_rate": 5e-06,
|
| 3040 |
+
"loss": 0.0054,
|
| 3041 |
+
"step": 3840
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 43.687943262411345,
|
| 3045 |
+
"grad_norm": 0.6264915981813468,
|
| 3046 |
+
"learning_rate": 5e-06,
|
| 3047 |
+
"loss": 0.0056,
|
| 3048 |
+
"step": 3850
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 43.801418439716315,
|
| 3052 |
+
"grad_norm": 0.6015317942981112,
|
| 3053 |
+
"learning_rate": 5e-06,
|
| 3054 |
+
"loss": 0.0058,
|
| 3055 |
+
"step": 3860
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 43.91489361702128,
|
| 3059 |
+
"grad_norm": 0.6365257752447925,
|
| 3060 |
+
"learning_rate": 5e-06,
|
| 3061 |
+
"loss": 0.0059,
|
| 3062 |
+
"step": 3870
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 43.99432624113475,
|
| 3066 |
+
"eval_loss": 2.413682222366333,
|
| 3067 |
+
"eval_runtime": 62.0169,
|
| 3068 |
+
"eval_samples_per_second": 38.248,
|
| 3069 |
+
"eval_steps_per_second": 0.613,
|
| 3070 |
+
"step": 3877
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 44.02836879432624,
|
| 3074 |
+
"grad_norm": 0.5790636141901765,
|
| 3075 |
+
"learning_rate": 5e-06,
|
| 3076 |
+
"loss": 0.0062,
|
| 3077 |
+
"step": 3880
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 44.141843971631204,
|
| 3081 |
+
"grad_norm": 0.5977139442568224,
|
| 3082 |
+
"learning_rate": 5e-06,
|
| 3083 |
+
"loss": 0.0052,
|
| 3084 |
+
"step": 3890
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 44.255319148936174,
|
| 3088 |
+
"grad_norm": 0.561775775158954,
|
| 3089 |
+
"learning_rate": 5e-06,
|
| 3090 |
+
"loss": 0.0053,
|
| 3091 |
+
"step": 3900
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 44.36879432624114,
|
| 3095 |
+
"grad_norm": 0.6449694049046988,
|
| 3096 |
+
"learning_rate": 5e-06,
|
| 3097 |
+
"loss": 0.0056,
|
| 3098 |
+
"step": 3910
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 44.4822695035461,
|
| 3102 |
+
"grad_norm": 0.5502986632322298,
|
| 3103 |
+
"learning_rate": 5e-06,
|
| 3104 |
+
"loss": 0.0054,
|
| 3105 |
+
"step": 3920
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 44.59574468085106,
|
| 3109 |
+
"grad_norm": 0.6004460230999634,
|
| 3110 |
+
"learning_rate": 5e-06,
|
| 3111 |
+
"loss": 0.0056,
|
| 3112 |
+
"step": 3930
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 44.709219858156025,
|
| 3116 |
+
"grad_norm": 0.5848689800167252,
|
| 3117 |
+
"learning_rate": 5e-06,
|
| 3118 |
+
"loss": 0.0055,
|
| 3119 |
+
"step": 3940
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 44.822695035460995,
|
| 3123 |
+
"grad_norm": 0.5660229423422497,
|
| 3124 |
+
"learning_rate": 5e-06,
|
| 3125 |
+
"loss": 0.0055,
|
| 3126 |
+
"step": 3950
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 44.93617021276596,
|
| 3130 |
+
"grad_norm": 0.6235698314913066,
|
| 3131 |
+
"learning_rate": 5e-06,
|
| 3132 |
+
"loss": 0.0057,
|
| 3133 |
+
"step": 3960
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 44.99290780141844,
|
| 3137 |
+
"eval_loss": 2.4166228771209717,
|
| 3138 |
+
"eval_runtime": 69.0142,
|
| 3139 |
+
"eval_samples_per_second": 34.37,
|
| 3140 |
+
"eval_steps_per_second": 0.551,
|
| 3141 |
+
"step": 3965
|
| 3142 |
+
},
|
| 3143 |
+
{
|
| 3144 |
+
"epoch": 45.04964539007092,
|
| 3145 |
+
"grad_norm": 0.602636820068635,
|
| 3146 |
+
"learning_rate": 5e-06,
|
| 3147 |
+
"loss": 0.0057,
|
| 3148 |
+
"step": 3970
|
| 3149 |
+
},
|
| 3150 |
+
{
|
| 3151 |
+
"epoch": 45.163120567375884,
|
| 3152 |
+
"grad_norm": 0.5512354587640222,
|
| 3153 |
+
"learning_rate": 5e-06,
|
| 3154 |
+
"loss": 0.0052,
|
| 3155 |
+
"step": 3980
|
| 3156 |
+
},
|
| 3157 |
+
{
|
| 3158 |
+
"epoch": 45.276595744680854,
|
| 3159 |
+
"grad_norm": 0.5360556245484226,
|
| 3160 |
+
"learning_rate": 5e-06,
|
| 3161 |
+
"loss": 0.0053,
|
| 3162 |
+
"step": 3990
|
| 3163 |
+
},
|
| 3164 |
+
{
|
| 3165 |
+
"epoch": 45.39007092198582,
|
| 3166 |
+
"grad_norm": 0.584753693926125,
|
| 3167 |
+
"learning_rate": 5e-06,
|
| 3168 |
+
"loss": 0.0052,
|
| 3169 |
+
"step": 4000
|
| 3170 |
+
},
|
| 3171 |
+
{
|
| 3172 |
+
"epoch": 45.50354609929078,
|
| 3173 |
+
"grad_norm": 0.5475446229877688,
|
| 3174 |
+
"learning_rate": 5e-06,
|
| 3175 |
+
"loss": 0.0053,
|
| 3176 |
+
"step": 4010
|
| 3177 |
+
},
|
| 3178 |
+
{
|
| 3179 |
+
"epoch": 45.61702127659574,
|
| 3180 |
+
"grad_norm": 0.5833576450707664,
|
| 3181 |
+
"learning_rate": 5e-06,
|
| 3182 |
+
"loss": 0.0055,
|
| 3183 |
+
"step": 4020
|
| 3184 |
+
},
|
| 3185 |
+
{
|
| 3186 |
+
"epoch": 45.730496453900706,
|
| 3187 |
+
"grad_norm": 0.514889308216256,
|
| 3188 |
+
"learning_rate": 5e-06,
|
| 3189 |
+
"loss": 0.0055,
|
| 3190 |
+
"step": 4030
|
| 3191 |
+
},
|
| 3192 |
+
{
|
| 3193 |
+
"epoch": 45.843971631205676,
|
| 3194 |
+
"grad_norm": 0.6093305891992143,
|
| 3195 |
+
"learning_rate": 5e-06,
|
| 3196 |
+
"loss": 0.0056,
|
| 3197 |
+
"step": 4040
|
| 3198 |
+
},
|
| 3199 |
+
{
|
| 3200 |
+
"epoch": 45.93475177304965,
|
| 3201 |
+
"eval_loss": 2.3638699054718018,
|
| 3202 |
+
"eval_runtime": 59.5586,
|
| 3203 |
+
"eval_samples_per_second": 39.826,
|
| 3204 |
+
"eval_steps_per_second": 0.638,
|
| 3205 |
+
"step": 4048
|
| 3206 |
+
},
|
| 3207 |
+
{
|
| 3208 |
+
"epoch": 45.93475177304965,
|
| 3209 |
+
"step": 4048,
|
| 3210 |
+
"total_flos": 6780341389885440.0,
|
| 3211 |
+
"train_loss": 0.15424752476433726,
|
| 3212 |
+
"train_runtime": 138136.8738,
|
| 3213 |
+
"train_samples_per_second": 15.006,
|
| 3214 |
+
"train_steps_per_second": 0.029
|
| 3215 |
+
}
|
| 3216 |
+
],
|
| 3217 |
+
"logging_steps": 10,
|
| 3218 |
+
"max_steps": 4048,
|
| 3219 |
+
"num_input_tokens_seen": 0,
|
| 3220 |
+
"num_train_epochs": 46,
|
| 3221 |
+
"save_steps": 500,
|
| 3222 |
+
"stateful_callbacks": {
|
| 3223 |
+
"TrainerControl": {
|
| 3224 |
+
"args": {
|
| 3225 |
+
"should_epoch_stop": false,
|
| 3226 |
+
"should_evaluate": false,
|
| 3227 |
+
"should_log": false,
|
| 3228 |
+
"should_save": true,
|
| 3229 |
+
"should_training_stop": true
|
| 3230 |
+
},
|
| 3231 |
+
"attributes": {}
|
| 3232 |
+
}
|
| 3233 |
+
},
|
| 3234 |
+
"total_flos": 6780341389885440.0,
|
| 3235 |
+
"train_batch_size": 8,
|
| 3236 |
+
"trial_name": null,
|
| 3237 |
+
"trial_params": null
|
| 3238 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|