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 +3126 -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.5_compute_equal
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# oh_scale_x.5_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: 2.4058
|
| 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.5_compute_equal
|
|
|
|
| 16 |
|
| 17 |
# oh_scale_x.5_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_0.5x dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 2.4058
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 24.868123587038433,
|
| 3 |
+
"eval_loss": 2.4057514667510986,
|
| 4 |
+
"eval_runtime": 113.9035,
|
| 5 |
+
"eval_samples_per_second": 39.235,
|
| 6 |
+
"eval_steps_per_second": 0.615,
|
| 7 |
+
"total_flos": 6909319257784320.0,
|
| 8 |
+
"train_loss": 0.2661740447047985,
|
| 9 |
+
"train_runtime": 139979.5045,
|
| 10 |
+
"train_samples_per_second": 15.162,
|
| 11 |
+
"train_steps_per_second": 0.029
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 24.868123587038433,
|
| 3 |
+
"eval_loss": 2.4057514667510986,
|
| 4 |
+
"eval_runtime": 113.9035,
|
| 5 |
+
"eval_samples_per_second": 39.235,
|
| 6 |
+
"eval_steps_per_second": 0.615
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 24.868123587038433,
|
| 3 |
+
"total_flos": 6909319257784320.0,
|
| 4 |
+
"train_loss": 0.2661740447047985,
|
| 5 |
+
"train_runtime": 139979.5045,
|
| 6 |
+
"train_samples_per_second": 15.162,
|
| 7 |
+
"train_steps_per_second": 0.029
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3126 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 24.868123587038433,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4125,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06028636021100226,
|
| 13 |
+
"grad_norm": 14.353901908542692,
|
| 14 |
+
"learning_rate": 5e-06,
|
| 15 |
+
"loss": 1.0595,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.12057272042200452,
|
| 20 |
+
"grad_norm": 1.3180865308999647,
|
| 21 |
+
"learning_rate": 5e-06,
|
| 22 |
+
"loss": 0.9403,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1808590806330068,
|
| 27 |
+
"grad_norm": 0.9782137323160548,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.89,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.24114544084400905,
|
| 34 |
+
"grad_norm": 1.3518217541449848,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 0.8618,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.30143180105501133,
|
| 41 |
+
"grad_norm": 0.8934828143592951,
|
| 42 |
+
"learning_rate": 5e-06,
|
| 43 |
+
"loss": 0.8447,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.3617181612660136,
|
| 48 |
+
"grad_norm": 0.9919016971761021,
|
| 49 |
+
"learning_rate": 5e-06,
|
| 50 |
+
"loss": 0.8318,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.42200452147701584,
|
| 55 |
+
"grad_norm": 0.9227497685306041,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.8177,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.4822908816880181,
|
| 62 |
+
"grad_norm": 0.9363087183332998,
|
| 63 |
+
"learning_rate": 5e-06,
|
| 64 |
+
"loss": 0.8106,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.5425772418990203,
|
| 69 |
+
"grad_norm": 1.0736624468965914,
|
| 70 |
+
"learning_rate": 5e-06,
|
| 71 |
+
"loss": 0.8052,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.6028636021100227,
|
| 76 |
+
"grad_norm": 1.353564519927803,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.7907,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.6631499623210249,
|
| 83 |
+
"grad_norm": 1.0455405570196998,
|
| 84 |
+
"learning_rate": 5e-06,
|
| 85 |
+
"loss": 0.7913,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.7234363225320272,
|
| 90 |
+
"grad_norm": 0.9569192147667731,
|
| 91 |
+
"learning_rate": 5e-06,
|
| 92 |
+
"loss": 0.7828,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.7837226827430294,
|
| 97 |
+
"grad_norm": 0.8298672277424939,
|
| 98 |
+
"learning_rate": 5e-06,
|
| 99 |
+
"loss": 0.7841,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.8440090429540317,
|
| 104 |
+
"grad_norm": 1.2437896161299962,
|
| 105 |
+
"learning_rate": 5e-06,
|
| 106 |
+
"loss": 0.7832,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.9042954031650339,
|
| 111 |
+
"grad_norm": 0.6290099294011746,
|
| 112 |
+
"learning_rate": 5e-06,
|
| 113 |
+
"loss": 0.7818,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.9645817633760362,
|
| 118 |
+
"grad_norm": 0.683050547264518,
|
| 119 |
+
"learning_rate": 5e-06,
|
| 120 |
+
"loss": 0.7807,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.9947249434815373,
|
| 125 |
+
"eval_loss": 0.7690147161483765,
|
| 126 |
+
"eval_runtime": 114.9859,
|
| 127 |
+
"eval_samples_per_second": 38.866,
|
| 128 |
+
"eval_steps_per_second": 0.609,
|
| 129 |
+
"step": 165
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0248681235870385,
|
| 133 |
+
"grad_norm": 0.8375316338814506,
|
| 134 |
+
"learning_rate": 5e-06,
|
| 135 |
+
"loss": 0.7706,
|
| 136 |
+
"step": 170
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.0851544837980407,
|
| 140 |
+
"grad_norm": 0.8491856111019711,
|
| 141 |
+
"learning_rate": 5e-06,
|
| 142 |
+
"loss": 0.7207,
|
| 143 |
+
"step": 180
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.145440844009043,
|
| 147 |
+
"grad_norm": 1.00264106211233,
|
| 148 |
+
"learning_rate": 5e-06,
|
| 149 |
+
"loss": 0.7264,
|
| 150 |
+
"step": 190
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.2057272042200453,
|
| 154 |
+
"grad_norm": 0.9524591725477187,
|
| 155 |
+
"learning_rate": 5e-06,
|
| 156 |
+
"loss": 0.7223,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.2660135644310475,
|
| 161 |
+
"grad_norm": 0.8200876998100735,
|
| 162 |
+
"learning_rate": 5e-06,
|
| 163 |
+
"loss": 0.7237,
|
| 164 |
+
"step": 210
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.3262999246420497,
|
| 168 |
+
"grad_norm": 0.6972921329207209,
|
| 169 |
+
"learning_rate": 5e-06,
|
| 170 |
+
"loss": 0.7233,
|
| 171 |
+
"step": 220
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.3865862848530521,
|
| 175 |
+
"grad_norm": 1.3472894348419586,
|
| 176 |
+
"learning_rate": 5e-06,
|
| 177 |
+
"loss": 0.7175,
|
| 178 |
+
"step": 230
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.4468726450640543,
|
| 182 |
+
"grad_norm": 0.8080326013334395,
|
| 183 |
+
"learning_rate": 5e-06,
|
| 184 |
+
"loss": 0.7203,
|
| 185 |
+
"step": 240
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.5071590052750565,
|
| 189 |
+
"grad_norm": 0.6690842132293575,
|
| 190 |
+
"learning_rate": 5e-06,
|
| 191 |
+
"loss": 0.7184,
|
| 192 |
+
"step": 250
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.5674453654860587,
|
| 196 |
+
"grad_norm": 0.6680726643782731,
|
| 197 |
+
"learning_rate": 5e-06,
|
| 198 |
+
"loss": 0.7193,
|
| 199 |
+
"step": 260
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.627731725697061,
|
| 203 |
+
"grad_norm": 0.6582407897638277,
|
| 204 |
+
"learning_rate": 5e-06,
|
| 205 |
+
"loss": 0.7171,
|
| 206 |
+
"step": 270
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.6880180859080633,
|
| 210 |
+
"grad_norm": 0.6058749327234759,
|
| 211 |
+
"learning_rate": 5e-06,
|
| 212 |
+
"loss": 0.718,
|
| 213 |
+
"step": 280
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.7483044461190655,
|
| 217 |
+
"grad_norm": 0.6771950284954183,
|
| 218 |
+
"learning_rate": 5e-06,
|
| 219 |
+
"loss": 0.717,
|
| 220 |
+
"step": 290
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.8085908063300677,
|
| 224 |
+
"grad_norm": 0.620278984859575,
|
| 225 |
+
"learning_rate": 5e-06,
|
| 226 |
+
"loss": 0.721,
|
| 227 |
+
"step": 300
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.8688771665410702,
|
| 231 |
+
"grad_norm": 0.7376417188133093,
|
| 232 |
+
"learning_rate": 5e-06,
|
| 233 |
+
"loss": 0.7187,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.9291635267520724,
|
| 238 |
+
"grad_norm": 0.841815742734191,
|
| 239 |
+
"learning_rate": 5e-06,
|
| 240 |
+
"loss": 0.7168,
|
| 241 |
+
"step": 320
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.9894498869630746,
|
| 245 |
+
"grad_norm": 0.9694154154724743,
|
| 246 |
+
"learning_rate": 5e-06,
|
| 247 |
+
"loss": 0.7139,
|
| 248 |
+
"step": 330
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.995478522984175,
|
| 252 |
+
"eval_loss": 0.7520068287849426,
|
| 253 |
+
"eval_runtime": 112.311,
|
| 254 |
+
"eval_samples_per_second": 39.791,
|
| 255 |
+
"eval_steps_per_second": 0.623,
|
| 256 |
+
"step": 331
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.049736247174077,
|
| 260 |
+
"grad_norm": 1.3857092066503702,
|
| 261 |
+
"learning_rate": 5e-06,
|
| 262 |
+
"loss": 0.6895,
|
| 263 |
+
"step": 340
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.110022607385079,
|
| 267 |
+
"grad_norm": 0.9308939279408932,
|
| 268 |
+
"learning_rate": 5e-06,
|
| 269 |
+
"loss": 0.661,
|
| 270 |
+
"step": 350
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.1703089675960814,
|
| 274 |
+
"grad_norm": 1.7049059921843441,
|
| 275 |
+
"learning_rate": 5e-06,
|
| 276 |
+
"loss": 0.6596,
|
| 277 |
+
"step": 360
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.230595327807084,
|
| 281 |
+
"grad_norm": 1.1664692858604568,
|
| 282 |
+
"learning_rate": 5e-06,
|
| 283 |
+
"loss": 0.6626,
|
| 284 |
+
"step": 370
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.290881688018086,
|
| 288 |
+
"grad_norm": 0.8099811209522968,
|
| 289 |
+
"learning_rate": 5e-06,
|
| 290 |
+
"loss": 0.6609,
|
| 291 |
+
"step": 380
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.351168048229088,
|
| 295 |
+
"grad_norm": 0.7769380652338045,
|
| 296 |
+
"learning_rate": 5e-06,
|
| 297 |
+
"loss": 0.6633,
|
| 298 |
+
"step": 390
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.4114544084400906,
|
| 302 |
+
"grad_norm": 1.0202300923789958,
|
| 303 |
+
"learning_rate": 5e-06,
|
| 304 |
+
"loss": 0.667,
|
| 305 |
+
"step": 400
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.4717407686510926,
|
| 309 |
+
"grad_norm": 0.9120675026610565,
|
| 310 |
+
"learning_rate": 5e-06,
|
| 311 |
+
"loss": 0.6654,
|
| 312 |
+
"step": 410
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.532027128862095,
|
| 316 |
+
"grad_norm": 0.8057600298686861,
|
| 317 |
+
"learning_rate": 5e-06,
|
| 318 |
+
"loss": 0.6643,
|
| 319 |
+
"step": 420
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.5923134890730974,
|
| 323 |
+
"grad_norm": 0.6688016636812687,
|
| 324 |
+
"learning_rate": 5e-06,
|
| 325 |
+
"loss": 0.6657,
|
| 326 |
+
"step": 430
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.6525998492840994,
|
| 330 |
+
"grad_norm": 0.7419224625762498,
|
| 331 |
+
"learning_rate": 5e-06,
|
| 332 |
+
"loss": 0.6669,
|
| 333 |
+
"step": 440
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.712886209495102,
|
| 337 |
+
"grad_norm": 0.7294710798953756,
|
| 338 |
+
"learning_rate": 5e-06,
|
| 339 |
+
"loss": 0.664,
|
| 340 |
+
"step": 450
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.7731725697061043,
|
| 344 |
+
"grad_norm": 0.7239503892331308,
|
| 345 |
+
"learning_rate": 5e-06,
|
| 346 |
+
"loss": 0.6628,
|
| 347 |
+
"step": 460
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.8334589299171062,
|
| 351 |
+
"grad_norm": 0.6873637028133951,
|
| 352 |
+
"learning_rate": 5e-06,
|
| 353 |
+
"loss": 0.6634,
|
| 354 |
+
"step": 470
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.8937452901281087,
|
| 358 |
+
"grad_norm": 0.7070671521244221,
|
| 359 |
+
"learning_rate": 5e-06,
|
| 360 |
+
"loss": 0.67,
|
| 361 |
+
"step": 480
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.9540316503391106,
|
| 365 |
+
"grad_norm": 0.6467492636416056,
|
| 366 |
+
"learning_rate": 5e-06,
|
| 367 |
+
"loss": 0.6642,
|
| 368 |
+
"step": 490
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.9962321024868124,
|
| 372 |
+
"eval_loss": 0.7524702548980713,
|
| 373 |
+
"eval_runtime": 115.8914,
|
| 374 |
+
"eval_samples_per_second": 38.562,
|
| 375 |
+
"eval_steps_per_second": 0.604,
|
| 376 |
+
"step": 497
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 3.014318010550113,
|
| 380 |
+
"grad_norm": 1.2600144566497484,
|
| 381 |
+
"learning_rate": 5e-06,
|
| 382 |
+
"loss": 0.6595,
|
| 383 |
+
"step": 500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 3.0746043707611155,
|
| 387 |
+
"grad_norm": 0.9439871932745402,
|
| 388 |
+
"learning_rate": 5e-06,
|
| 389 |
+
"loss": 0.6121,
|
| 390 |
+
"step": 510
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 3.1348907309721175,
|
| 394 |
+
"grad_norm": 0.9269396364377557,
|
| 395 |
+
"learning_rate": 5e-06,
|
| 396 |
+
"loss": 0.6091,
|
| 397 |
+
"step": 520
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 3.19517709118312,
|
| 401 |
+
"grad_norm": 1.4073687560739623,
|
| 402 |
+
"learning_rate": 5e-06,
|
| 403 |
+
"loss": 0.6146,
|
| 404 |
+
"step": 530
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.255463451394122,
|
| 408 |
+
"grad_norm": 1.0909582280984875,
|
| 409 |
+
"learning_rate": 5e-06,
|
| 410 |
+
"loss": 0.6135,
|
| 411 |
+
"step": 540
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 3.3157498116051243,
|
| 415 |
+
"grad_norm": 1.6855484277386785,
|
| 416 |
+
"learning_rate": 5e-06,
|
| 417 |
+
"loss": 0.6181,
|
| 418 |
+
"step": 550
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 3.3760361718161267,
|
| 422 |
+
"grad_norm": 1.3960879287332164,
|
| 423 |
+
"learning_rate": 5e-06,
|
| 424 |
+
"loss": 0.6204,
|
| 425 |
+
"step": 560
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 3.4363225320271287,
|
| 429 |
+
"grad_norm": 1.3111215484703225,
|
| 430 |
+
"learning_rate": 5e-06,
|
| 431 |
+
"loss": 0.6166,
|
| 432 |
+
"step": 570
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 3.496608892238131,
|
| 436 |
+
"grad_norm": 1.4456626678129865,
|
| 437 |
+
"learning_rate": 5e-06,
|
| 438 |
+
"loss": 0.6187,
|
| 439 |
+
"step": 580
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 3.5568952524491335,
|
| 443 |
+
"grad_norm": 0.9267972418344134,
|
| 444 |
+
"learning_rate": 5e-06,
|
| 445 |
+
"loss": 0.6198,
|
| 446 |
+
"step": 590
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 3.6171816126601355,
|
| 450 |
+
"grad_norm": 1.5516401845259948,
|
| 451 |
+
"learning_rate": 5e-06,
|
| 452 |
+
"loss": 0.6216,
|
| 453 |
+
"step": 600
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 3.677467972871138,
|
| 457 |
+
"grad_norm": 1.5014637761221297,
|
| 458 |
+
"learning_rate": 5e-06,
|
| 459 |
+
"loss": 0.618,
|
| 460 |
+
"step": 610
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 3.7377543330821403,
|
| 464 |
+
"grad_norm": 1.1059687872166368,
|
| 465 |
+
"learning_rate": 5e-06,
|
| 466 |
+
"loss": 0.6185,
|
| 467 |
+
"step": 620
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 3.7980406932931423,
|
| 471 |
+
"grad_norm": 1.2950526783062104,
|
| 472 |
+
"learning_rate": 5e-06,
|
| 473 |
+
"loss": 0.6207,
|
| 474 |
+
"step": 630
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.8583270535041447,
|
| 478 |
+
"grad_norm": 0.9542402867171873,
|
| 479 |
+
"learning_rate": 5e-06,
|
| 480 |
+
"loss": 0.618,
|
| 481 |
+
"step": 640
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 3.918613413715147,
|
| 485 |
+
"grad_norm": 0.7924967732032556,
|
| 486 |
+
"learning_rate": 5e-06,
|
| 487 |
+
"loss": 0.6163,
|
| 488 |
+
"step": 650
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.978899773926149,
|
| 492 |
+
"grad_norm": 0.9450387587222794,
|
| 493 |
+
"learning_rate": 5e-06,
|
| 494 |
+
"loss": 0.6186,
|
| 495 |
+
"step": 660
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 3.99698568198945,
|
| 499 |
+
"eval_loss": 0.7615257501602173,
|
| 500 |
+
"eval_runtime": 111.9191,
|
| 501 |
+
"eval_samples_per_second": 39.931,
|
| 502 |
+
"eval_steps_per_second": 0.625,
|
| 503 |
+
"step": 663
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 4.039186134137151,
|
| 507 |
+
"grad_norm": 0.9971689199182937,
|
| 508 |
+
"learning_rate": 5e-06,
|
| 509 |
+
"loss": 0.5958,
|
| 510 |
+
"step": 670
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 4.099472494348154,
|
| 514 |
+
"grad_norm": 0.7965799505534403,
|
| 515 |
+
"learning_rate": 5e-06,
|
| 516 |
+
"loss": 0.5627,
|
| 517 |
+
"step": 680
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 4.159758854559156,
|
| 521 |
+
"grad_norm": 0.8634698321641859,
|
| 522 |
+
"learning_rate": 5e-06,
|
| 523 |
+
"loss": 0.5642,
|
| 524 |
+
"step": 690
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.220045214770158,
|
| 528 |
+
"grad_norm": 0.8538005391887241,
|
| 529 |
+
"learning_rate": 5e-06,
|
| 530 |
+
"loss": 0.5721,
|
| 531 |
+
"step": 700
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.280331574981161,
|
| 535 |
+
"grad_norm": 0.8087517000342034,
|
| 536 |
+
"learning_rate": 5e-06,
|
| 537 |
+
"loss": 0.5702,
|
| 538 |
+
"step": 710
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.340617935192163,
|
| 542 |
+
"grad_norm": 0.7416451232136964,
|
| 543 |
+
"learning_rate": 5e-06,
|
| 544 |
+
"loss": 0.5735,
|
| 545 |
+
"step": 720
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 4.400904295403165,
|
| 549 |
+
"grad_norm": 0.683331985209307,
|
| 550 |
+
"learning_rate": 5e-06,
|
| 551 |
+
"loss": 0.576,
|
| 552 |
+
"step": 730
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 4.461190655614168,
|
| 556 |
+
"grad_norm": 0.7531824862918711,
|
| 557 |
+
"learning_rate": 5e-06,
|
| 558 |
+
"loss": 0.5747,
|
| 559 |
+
"step": 740
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 4.52147701582517,
|
| 563 |
+
"grad_norm": 0.6679489381777426,
|
| 564 |
+
"learning_rate": 5e-06,
|
| 565 |
+
"loss": 0.5788,
|
| 566 |
+
"step": 750
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 4.581763376036172,
|
| 570 |
+
"grad_norm": 0.8184328898914945,
|
| 571 |
+
"learning_rate": 5e-06,
|
| 572 |
+
"loss": 0.5773,
|
| 573 |
+
"step": 760
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 4.642049736247174,
|
| 577 |
+
"grad_norm": 0.7709343247764259,
|
| 578 |
+
"learning_rate": 5e-06,
|
| 579 |
+
"loss": 0.5747,
|
| 580 |
+
"step": 770
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 4.702336096458176,
|
| 584 |
+
"grad_norm": 0.8008305213299817,
|
| 585 |
+
"learning_rate": 5e-06,
|
| 586 |
+
"loss": 0.5789,
|
| 587 |
+
"step": 780
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 4.762622456669178,
|
| 591 |
+
"grad_norm": 0.770151242560135,
|
| 592 |
+
"learning_rate": 5e-06,
|
| 593 |
+
"loss": 0.5808,
|
| 594 |
+
"step": 790
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 4.822908816880181,
|
| 598 |
+
"grad_norm": 0.8007315199653617,
|
| 599 |
+
"learning_rate": 5e-06,
|
| 600 |
+
"loss": 0.5766,
|
| 601 |
+
"step": 800
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 4.883195177091183,
|
| 605 |
+
"grad_norm": 0.8153924743627864,
|
| 606 |
+
"learning_rate": 5e-06,
|
| 607 |
+
"loss": 0.5794,
|
| 608 |
+
"step": 810
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 4.943481537302185,
|
| 612 |
+
"grad_norm": 0.7155989123366884,
|
| 613 |
+
"learning_rate": 5e-06,
|
| 614 |
+
"loss": 0.5777,
|
| 615 |
+
"step": 820
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 4.997739261492088,
|
| 619 |
+
"eval_loss": 0.7784610390663147,
|
| 620 |
+
"eval_runtime": 113.9486,
|
| 621 |
+
"eval_samples_per_second": 39.219,
|
| 622 |
+
"eval_steps_per_second": 0.614,
|
| 623 |
+
"step": 829
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 5.003767897513188,
|
| 627 |
+
"grad_norm": 1.5968784870292558,
|
| 628 |
+
"learning_rate": 5e-06,
|
| 629 |
+
"loss": 0.5864,
|
| 630 |
+
"step": 830
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 5.06405425772419,
|
| 634 |
+
"grad_norm": 1.3160101230444552,
|
| 635 |
+
"learning_rate": 5e-06,
|
| 636 |
+
"loss": 0.5136,
|
| 637 |
+
"step": 840
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 5.124340617935192,
|
| 641 |
+
"grad_norm": 0.9218240884563489,
|
| 642 |
+
"learning_rate": 5e-06,
|
| 643 |
+
"loss": 0.514,
|
| 644 |
+
"step": 850
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 5.184626978146194,
|
| 648 |
+
"grad_norm": 0.943292128668918,
|
| 649 |
+
"learning_rate": 5e-06,
|
| 650 |
+
"loss": 0.5129,
|
| 651 |
+
"step": 860
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 5.244913338357197,
|
| 655 |
+
"grad_norm": 0.9210116431359558,
|
| 656 |
+
"learning_rate": 5e-06,
|
| 657 |
+
"loss": 0.5147,
|
| 658 |
+
"step": 870
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 5.305199698568199,
|
| 662 |
+
"grad_norm": 0.9123827846726384,
|
| 663 |
+
"learning_rate": 5e-06,
|
| 664 |
+
"loss": 0.5152,
|
| 665 |
+
"step": 880
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 5.365486058779201,
|
| 669 |
+
"grad_norm": 0.8967759483580899,
|
| 670 |
+
"learning_rate": 5e-06,
|
| 671 |
+
"loss": 0.5246,
|
| 672 |
+
"step": 890
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 5.425772418990204,
|
| 676 |
+
"grad_norm": 1.0833892045281441,
|
| 677 |
+
"learning_rate": 5e-06,
|
| 678 |
+
"loss": 0.5173,
|
| 679 |
+
"step": 900
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 5.486058779201206,
|
| 683 |
+
"grad_norm": 0.8452587501323132,
|
| 684 |
+
"learning_rate": 5e-06,
|
| 685 |
+
"loss": 0.5205,
|
| 686 |
+
"step": 910
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 5.546345139412208,
|
| 690 |
+
"grad_norm": 0.9036749110113976,
|
| 691 |
+
"learning_rate": 5e-06,
|
| 692 |
+
"loss": 0.5213,
|
| 693 |
+
"step": 920
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 5.6066314996232105,
|
| 697 |
+
"grad_norm": 0.8270647016134544,
|
| 698 |
+
"learning_rate": 5e-06,
|
| 699 |
+
"loss": 0.5214,
|
| 700 |
+
"step": 930
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 5.6669178598342125,
|
| 704 |
+
"grad_norm": 0.760029827349666,
|
| 705 |
+
"learning_rate": 5e-06,
|
| 706 |
+
"loss": 0.5233,
|
| 707 |
+
"step": 940
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 5.7272042200452145,
|
| 711 |
+
"grad_norm": 1.1680570931837515,
|
| 712 |
+
"learning_rate": 5e-06,
|
| 713 |
+
"loss": 0.5284,
|
| 714 |
+
"step": 950
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 5.787490580256217,
|
| 718 |
+
"grad_norm": 1.0453134822125045,
|
| 719 |
+
"learning_rate": 5e-06,
|
| 720 |
+
"loss": 0.5273,
|
| 721 |
+
"step": 960
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 5.847776940467219,
|
| 725 |
+
"grad_norm": 0.8063546838482688,
|
| 726 |
+
"learning_rate": 5e-06,
|
| 727 |
+
"loss": 0.5265,
|
| 728 |
+
"step": 970
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 5.908063300678221,
|
| 732 |
+
"grad_norm": 0.8921967199529918,
|
| 733 |
+
"learning_rate": 5e-06,
|
| 734 |
+
"loss": 0.5316,
|
| 735 |
+
"step": 980
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 5.968349660889224,
|
| 739 |
+
"grad_norm": 0.8942899369042694,
|
| 740 |
+
"learning_rate": 5e-06,
|
| 741 |
+
"loss": 0.5287,
|
| 742 |
+
"step": 990
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 5.998492840994725,
|
| 746 |
+
"eval_loss": 0.815434992313385,
|
| 747 |
+
"eval_runtime": 116.1895,
|
| 748 |
+
"eval_samples_per_second": 38.463,
|
| 749 |
+
"eval_steps_per_second": 0.602,
|
| 750 |
+
"step": 995
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 6.028636021100226,
|
| 754 |
+
"grad_norm": 1.6908391333792119,
|
| 755 |
+
"learning_rate": 5e-06,
|
| 756 |
+
"loss": 0.5023,
|
| 757 |
+
"step": 1000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 6.088922381311228,
|
| 761 |
+
"grad_norm": 1.2916624414612379,
|
| 762 |
+
"learning_rate": 5e-06,
|
| 763 |
+
"loss": 0.4578,
|
| 764 |
+
"step": 1010
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 6.149208741522231,
|
| 768 |
+
"grad_norm": 1.1040568786409888,
|
| 769 |
+
"learning_rate": 5e-06,
|
| 770 |
+
"loss": 0.454,
|
| 771 |
+
"step": 1020
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 6.209495101733233,
|
| 775 |
+
"grad_norm": 1.0463754179790778,
|
| 776 |
+
"learning_rate": 5e-06,
|
| 777 |
+
"loss": 0.4592,
|
| 778 |
+
"step": 1030
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 6.269781461944235,
|
| 782 |
+
"grad_norm": 1.0493678707628407,
|
| 783 |
+
"learning_rate": 5e-06,
|
| 784 |
+
"loss": 0.4582,
|
| 785 |
+
"step": 1040
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 6.330067822155238,
|
| 789 |
+
"grad_norm": 1.105648007217362,
|
| 790 |
+
"learning_rate": 5e-06,
|
| 791 |
+
"loss": 0.4593,
|
| 792 |
+
"step": 1050
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 6.39035418236624,
|
| 796 |
+
"grad_norm": 1.0291790248523935,
|
| 797 |
+
"learning_rate": 5e-06,
|
| 798 |
+
"loss": 0.4629,
|
| 799 |
+
"step": 1060
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 6.450640542577242,
|
| 803 |
+
"grad_norm": 1.2514460402035084,
|
| 804 |
+
"learning_rate": 5e-06,
|
| 805 |
+
"loss": 0.4634,
|
| 806 |
+
"step": 1070
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 6.510926902788244,
|
| 810 |
+
"grad_norm": 0.8518373738036685,
|
| 811 |
+
"learning_rate": 5e-06,
|
| 812 |
+
"loss": 0.4654,
|
| 813 |
+
"step": 1080
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 6.571213262999247,
|
| 817 |
+
"grad_norm": 1.071998502866085,
|
| 818 |
+
"learning_rate": 5e-06,
|
| 819 |
+
"loss": 0.4663,
|
| 820 |
+
"step": 1090
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 6.6314996232102486,
|
| 824 |
+
"grad_norm": 0.9000534182028347,
|
| 825 |
+
"learning_rate": 5e-06,
|
| 826 |
+
"loss": 0.4658,
|
| 827 |
+
"step": 1100
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 6.691785983421251,
|
| 831 |
+
"grad_norm": 0.8995635602376101,
|
| 832 |
+
"learning_rate": 5e-06,
|
| 833 |
+
"loss": 0.469,
|
| 834 |
+
"step": 1110
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 6.752072343632253,
|
| 838 |
+
"grad_norm": 0.9011803423576781,
|
| 839 |
+
"learning_rate": 5e-06,
|
| 840 |
+
"loss": 0.4715,
|
| 841 |
+
"step": 1120
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 6.812358703843255,
|
| 845 |
+
"grad_norm": 0.9140285189147836,
|
| 846 |
+
"learning_rate": 5e-06,
|
| 847 |
+
"loss": 0.4763,
|
| 848 |
+
"step": 1130
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 6.872645064054257,
|
| 852 |
+
"grad_norm": 1.0476263187239303,
|
| 853 |
+
"learning_rate": 5e-06,
|
| 854 |
+
"loss": 0.4781,
|
| 855 |
+
"step": 1140
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 6.93293142426526,
|
| 859 |
+
"grad_norm": 1.1369535585860062,
|
| 860 |
+
"learning_rate": 5e-06,
|
| 861 |
+
"loss": 0.4702,
|
| 862 |
+
"step": 1150
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 6.993217784476262,
|
| 866 |
+
"grad_norm": 1.2720345806831712,
|
| 867 |
+
"learning_rate": 5e-06,
|
| 868 |
+
"loss": 0.473,
|
| 869 |
+
"step": 1160
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 6.999246420497363,
|
| 873 |
+
"eval_loss": 0.8709951639175415,
|
| 874 |
+
"eval_runtime": 116.2672,
|
| 875 |
+
"eval_samples_per_second": 38.437,
|
| 876 |
+
"eval_steps_per_second": 0.602,
|
| 877 |
+
"step": 1161
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 7.053504144687264,
|
| 881 |
+
"grad_norm": 1.6175687599813067,
|
| 882 |
+
"learning_rate": 5e-06,
|
| 883 |
+
"loss": 0.4154,
|
| 884 |
+
"step": 1170
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 7.113790504898267,
|
| 888 |
+
"grad_norm": 1.2621822187556238,
|
| 889 |
+
"learning_rate": 5e-06,
|
| 890 |
+
"loss": 0.3985,
|
| 891 |
+
"step": 1180
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 7.174076865109269,
|
| 895 |
+
"grad_norm": 1.3109771889362003,
|
| 896 |
+
"learning_rate": 5e-06,
|
| 897 |
+
"loss": 0.3937,
|
| 898 |
+
"step": 1190
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 7.234363225320271,
|
| 902 |
+
"grad_norm": 1.1476426236895934,
|
| 903 |
+
"learning_rate": 5e-06,
|
| 904 |
+
"loss": 0.3986,
|
| 905 |
+
"step": 1200
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 7.294649585531274,
|
| 909 |
+
"grad_norm": 1.1513686323462657,
|
| 910 |
+
"learning_rate": 5e-06,
|
| 911 |
+
"loss": 0.4003,
|
| 912 |
+
"step": 1210
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 7.354935945742276,
|
| 916 |
+
"grad_norm": 1.2318284334824612,
|
| 917 |
+
"learning_rate": 5e-06,
|
| 918 |
+
"loss": 0.402,
|
| 919 |
+
"step": 1220
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 7.415222305953278,
|
| 923 |
+
"grad_norm": 1.3293034219686575,
|
| 924 |
+
"learning_rate": 5e-06,
|
| 925 |
+
"loss": 0.4041,
|
| 926 |
+
"step": 1230
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 7.475508666164281,
|
| 930 |
+
"grad_norm": 1.3126587055332364,
|
| 931 |
+
"learning_rate": 5e-06,
|
| 932 |
+
"loss": 0.4028,
|
| 933 |
+
"step": 1240
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 7.535795026375283,
|
| 937 |
+
"grad_norm": 1.4541660160673635,
|
| 938 |
+
"learning_rate": 5e-06,
|
| 939 |
+
"loss": 0.4046,
|
| 940 |
+
"step": 1250
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 7.596081386586285,
|
| 944 |
+
"grad_norm": 1.1677113248358406,
|
| 945 |
+
"learning_rate": 5e-06,
|
| 946 |
+
"loss": 0.4064,
|
| 947 |
+
"step": 1260
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 7.6563677467972875,
|
| 951 |
+
"grad_norm": 1.2339252526811149,
|
| 952 |
+
"learning_rate": 5e-06,
|
| 953 |
+
"loss": 0.4057,
|
| 954 |
+
"step": 1270
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 7.7166541070082895,
|
| 958 |
+
"grad_norm": 0.9795667261721808,
|
| 959 |
+
"learning_rate": 5e-06,
|
| 960 |
+
"loss": 0.4105,
|
| 961 |
+
"step": 1280
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 7.776940467219291,
|
| 965 |
+
"grad_norm": 1.3122595751880688,
|
| 966 |
+
"learning_rate": 5e-06,
|
| 967 |
+
"loss": 0.4115,
|
| 968 |
+
"step": 1290
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 7.837226827430294,
|
| 972 |
+
"grad_norm": 1.3363743108253807,
|
| 973 |
+
"learning_rate": 5e-06,
|
| 974 |
+
"loss": 0.4094,
|
| 975 |
+
"step": 1300
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 7.897513187641296,
|
| 979 |
+
"grad_norm": 1.2278606177387434,
|
| 980 |
+
"learning_rate": 5e-06,
|
| 981 |
+
"loss": 0.411,
|
| 982 |
+
"step": 1310
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 7.957799547852298,
|
| 986 |
+
"grad_norm": 1.099485555308857,
|
| 987 |
+
"learning_rate": 5e-06,
|
| 988 |
+
"loss": 0.4134,
|
| 989 |
+
"step": 1320
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 8.0,
|
| 993 |
+
"eval_loss": 0.9475046396255493,
|
| 994 |
+
"eval_runtime": 118.217,
|
| 995 |
+
"eval_samples_per_second": 37.803,
|
| 996 |
+
"eval_steps_per_second": 0.592,
|
| 997 |
+
"step": 1327
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 8.0180859080633,
|
| 1001 |
+
"grad_norm": 2.747551612605114,
|
| 1002 |
+
"learning_rate": 5e-06,
|
| 1003 |
+
"loss": 0.3977,
|
| 1004 |
+
"step": 1330
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 8.078372268274302,
|
| 1008 |
+
"grad_norm": 1.7260940558261295,
|
| 1009 |
+
"learning_rate": 5e-06,
|
| 1010 |
+
"loss": 0.3441,
|
| 1011 |
+
"step": 1340
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 8.138658628485306,
|
| 1015 |
+
"grad_norm": 1.2842675938679606,
|
| 1016 |
+
"learning_rate": 5e-06,
|
| 1017 |
+
"loss": 0.3397,
|
| 1018 |
+
"step": 1350
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 8.198944988696308,
|
| 1022 |
+
"grad_norm": 1.1398592198842168,
|
| 1023 |
+
"learning_rate": 5e-06,
|
| 1024 |
+
"loss": 0.3442,
|
| 1025 |
+
"step": 1360
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 8.25923134890731,
|
| 1029 |
+
"grad_norm": 1.153557070127385,
|
| 1030 |
+
"learning_rate": 5e-06,
|
| 1031 |
+
"loss": 0.3454,
|
| 1032 |
+
"step": 1370
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 8.319517709118312,
|
| 1036 |
+
"grad_norm": 1.3384373413827295,
|
| 1037 |
+
"learning_rate": 5e-06,
|
| 1038 |
+
"loss": 0.3438,
|
| 1039 |
+
"step": 1380
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 8.379804069329314,
|
| 1043 |
+
"grad_norm": 1.3554470516864743,
|
| 1044 |
+
"learning_rate": 5e-06,
|
| 1045 |
+
"loss": 0.3493,
|
| 1046 |
+
"step": 1390
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 8.440090429540316,
|
| 1050 |
+
"grad_norm": 1.325212009488347,
|
| 1051 |
+
"learning_rate": 5e-06,
|
| 1052 |
+
"loss": 0.3528,
|
| 1053 |
+
"step": 1400
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 8.500376789751318,
|
| 1057 |
+
"grad_norm": 1.74556781346271,
|
| 1058 |
+
"learning_rate": 5e-06,
|
| 1059 |
+
"loss": 0.3548,
|
| 1060 |
+
"step": 1410
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 8.560663149962322,
|
| 1064 |
+
"grad_norm": 1.3741632805919384,
|
| 1065 |
+
"learning_rate": 5e-06,
|
| 1066 |
+
"loss": 0.3518,
|
| 1067 |
+
"step": 1420
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 8.620949510173324,
|
| 1071 |
+
"grad_norm": 1.3019952245252948,
|
| 1072 |
+
"learning_rate": 5e-06,
|
| 1073 |
+
"loss": 0.3521,
|
| 1074 |
+
"step": 1430
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 8.681235870384326,
|
| 1078 |
+
"grad_norm": 1.100365748448614,
|
| 1079 |
+
"learning_rate": 5e-06,
|
| 1080 |
+
"loss": 0.3573,
|
| 1081 |
+
"step": 1440
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 8.741522230595328,
|
| 1085 |
+
"grad_norm": 1.3911327396968252,
|
| 1086 |
+
"learning_rate": 5e-06,
|
| 1087 |
+
"loss": 0.3574,
|
| 1088 |
+
"step": 1450
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 8.80180859080633,
|
| 1092 |
+
"grad_norm": 1.3853784407579808,
|
| 1093 |
+
"learning_rate": 5e-06,
|
| 1094 |
+
"loss": 0.3606,
|
| 1095 |
+
"step": 1460
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 8.862094951017333,
|
| 1099 |
+
"grad_norm": 1.278532404930149,
|
| 1100 |
+
"learning_rate": 5e-06,
|
| 1101 |
+
"loss": 0.3567,
|
| 1102 |
+
"step": 1470
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 8.922381311228335,
|
| 1106 |
+
"grad_norm": 1.207780040603517,
|
| 1107 |
+
"learning_rate": 5e-06,
|
| 1108 |
+
"loss": 0.3565,
|
| 1109 |
+
"step": 1480
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 8.982667671439337,
|
| 1113 |
+
"grad_norm": 1.262306381996002,
|
| 1114 |
+
"learning_rate": 5e-06,
|
| 1115 |
+
"loss": 0.3615,
|
| 1116 |
+
"step": 1490
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 8.994724943481538,
|
| 1120 |
+
"eval_loss": 1.0203213691711426,
|
| 1121 |
+
"eval_runtime": 114.5597,
|
| 1122 |
+
"eval_samples_per_second": 39.01,
|
| 1123 |
+
"eval_steps_per_second": 0.611,
|
| 1124 |
+
"step": 1492
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 9.04295403165034,
|
| 1128 |
+
"grad_norm": 1.9088952161459727,
|
| 1129 |
+
"learning_rate": 5e-06,
|
| 1130 |
+
"loss": 0.3188,
|
| 1131 |
+
"step": 1500
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 9.103240391861341,
|
| 1135 |
+
"grad_norm": 1.5762821812746035,
|
| 1136 |
+
"learning_rate": 5e-06,
|
| 1137 |
+
"loss": 0.2921,
|
| 1138 |
+
"step": 1510
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 9.163526752072343,
|
| 1142 |
+
"grad_norm": 1.892762439817433,
|
| 1143 |
+
"learning_rate": 5e-06,
|
| 1144 |
+
"loss": 0.2866,
|
| 1145 |
+
"step": 1520
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 9.223813112283345,
|
| 1149 |
+
"grad_norm": 1.5993468972103952,
|
| 1150 |
+
"learning_rate": 5e-06,
|
| 1151 |
+
"loss": 0.2913,
|
| 1152 |
+
"step": 1530
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 9.284099472494349,
|
| 1156 |
+
"grad_norm": 1.7180835806783503,
|
| 1157 |
+
"learning_rate": 5e-06,
|
| 1158 |
+
"loss": 0.2925,
|
| 1159 |
+
"step": 1540
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 9.34438583270535,
|
| 1163 |
+
"grad_norm": 1.4283006573501846,
|
| 1164 |
+
"learning_rate": 5e-06,
|
| 1165 |
+
"loss": 0.2936,
|
| 1166 |
+
"step": 1550
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 9.404672192916353,
|
| 1170 |
+
"grad_norm": 1.472590439034633,
|
| 1171 |
+
"learning_rate": 5e-06,
|
| 1172 |
+
"loss": 0.2957,
|
| 1173 |
+
"step": 1560
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 9.464958553127355,
|
| 1177 |
+
"grad_norm": 1.6627373496351208,
|
| 1178 |
+
"learning_rate": 5e-06,
|
| 1179 |
+
"loss": 0.2968,
|
| 1180 |
+
"step": 1570
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 9.525244913338357,
|
| 1184 |
+
"grad_norm": 1.6479857681742245,
|
| 1185 |
+
"learning_rate": 5e-06,
|
| 1186 |
+
"loss": 0.2968,
|
| 1187 |
+
"step": 1580
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 9.585531273549359,
|
| 1191 |
+
"grad_norm": 1.451039229532527,
|
| 1192 |
+
"learning_rate": 5e-06,
|
| 1193 |
+
"loss": 0.2969,
|
| 1194 |
+
"step": 1590
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 9.645817633760362,
|
| 1198 |
+
"grad_norm": 1.338137019241527,
|
| 1199 |
+
"learning_rate": 5e-06,
|
| 1200 |
+
"loss": 0.2999,
|
| 1201 |
+
"step": 1600
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 9.706103993971364,
|
| 1205 |
+
"grad_norm": 1.4364529449975805,
|
| 1206 |
+
"learning_rate": 5e-06,
|
| 1207 |
+
"loss": 0.301,
|
| 1208 |
+
"step": 1610
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 9.766390354182366,
|
| 1212 |
+
"grad_norm": 1.6181472883899526,
|
| 1213 |
+
"learning_rate": 5e-06,
|
| 1214 |
+
"loss": 0.3021,
|
| 1215 |
+
"step": 1620
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 9.826676714393368,
|
| 1219 |
+
"grad_norm": 1.2929981294290136,
|
| 1220 |
+
"learning_rate": 5e-06,
|
| 1221 |
+
"loss": 0.3047,
|
| 1222 |
+
"step": 1630
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 9.88696307460437,
|
| 1226 |
+
"grad_norm": 1.5569082204346147,
|
| 1227 |
+
"learning_rate": 5e-06,
|
| 1228 |
+
"loss": 0.3039,
|
| 1229 |
+
"step": 1640
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 9.947249434815372,
|
| 1233 |
+
"grad_norm": 1.2611067636631264,
|
| 1234 |
+
"learning_rate": 5e-06,
|
| 1235 |
+
"loss": 0.3057,
|
| 1236 |
+
"step": 1650
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 9.995478522984175,
|
| 1240 |
+
"eval_loss": 1.1177444458007812,
|
| 1241 |
+
"eval_runtime": 116.3468,
|
| 1242 |
+
"eval_samples_per_second": 38.411,
|
| 1243 |
+
"eval_steps_per_second": 0.602,
|
| 1244 |
+
"step": 1658
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 10.007535795026376,
|
| 1248 |
+
"grad_norm": 1.793363117063337,
|
| 1249 |
+
"learning_rate": 5e-06,
|
| 1250 |
+
"loss": 0.3015,
|
| 1251 |
+
"step": 1660
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 10.067822155237378,
|
| 1255 |
+
"grad_norm": 1.9015543562249504,
|
| 1256 |
+
"learning_rate": 5e-06,
|
| 1257 |
+
"loss": 0.2446,
|
| 1258 |
+
"step": 1670
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 10.12810851544838,
|
| 1262 |
+
"grad_norm": 1.6166742540525376,
|
| 1263 |
+
"learning_rate": 5e-06,
|
| 1264 |
+
"loss": 0.2417,
|
| 1265 |
+
"step": 1680
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 10.188394875659382,
|
| 1269 |
+
"grad_norm": 1.5826959730874415,
|
| 1270 |
+
"learning_rate": 5e-06,
|
| 1271 |
+
"loss": 0.2407,
|
| 1272 |
+
"step": 1690
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 10.248681235870384,
|
| 1276 |
+
"grad_norm": 1.313614154478162,
|
| 1277 |
+
"learning_rate": 5e-06,
|
| 1278 |
+
"loss": 0.2421,
|
| 1279 |
+
"step": 1700
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 10.308967596081386,
|
| 1283 |
+
"grad_norm": 1.8551338463863283,
|
| 1284 |
+
"learning_rate": 5e-06,
|
| 1285 |
+
"loss": 0.2432,
|
| 1286 |
+
"step": 1710
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 10.369253956292388,
|
| 1290 |
+
"grad_norm": 1.3937043583045259,
|
| 1291 |
+
"learning_rate": 5e-06,
|
| 1292 |
+
"loss": 0.2426,
|
| 1293 |
+
"step": 1720
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 10.429540316503392,
|
| 1297 |
+
"grad_norm": 1.4908947612613312,
|
| 1298 |
+
"learning_rate": 5e-06,
|
| 1299 |
+
"loss": 0.2458,
|
| 1300 |
+
"step": 1730
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 10.489826676714394,
|
| 1304 |
+
"grad_norm": 1.3513472181146375,
|
| 1305 |
+
"learning_rate": 5e-06,
|
| 1306 |
+
"loss": 0.2468,
|
| 1307 |
+
"step": 1740
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 10.550113036925396,
|
| 1311 |
+
"grad_norm": 1.2442859437459592,
|
| 1312 |
+
"learning_rate": 5e-06,
|
| 1313 |
+
"loss": 0.2472,
|
| 1314 |
+
"step": 1750
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 10.610399397136398,
|
| 1318 |
+
"grad_norm": 1.61049205004398,
|
| 1319 |
+
"learning_rate": 5e-06,
|
| 1320 |
+
"loss": 0.2497,
|
| 1321 |
+
"step": 1760
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 10.6706857573474,
|
| 1325 |
+
"grad_norm": 1.4687600663257643,
|
| 1326 |
+
"learning_rate": 5e-06,
|
| 1327 |
+
"loss": 0.2513,
|
| 1328 |
+
"step": 1770
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 10.730972117558402,
|
| 1332 |
+
"grad_norm": 1.3544154919254643,
|
| 1333 |
+
"learning_rate": 5e-06,
|
| 1334 |
+
"loss": 0.2534,
|
| 1335 |
+
"step": 1780
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 10.791258477769405,
|
| 1339 |
+
"grad_norm": 1.4310611416705883,
|
| 1340 |
+
"learning_rate": 5e-06,
|
| 1341 |
+
"loss": 0.2534,
|
| 1342 |
+
"step": 1790
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 10.851544837980407,
|
| 1346 |
+
"grad_norm": 1.5722579222869695,
|
| 1347 |
+
"learning_rate": 5e-06,
|
| 1348 |
+
"loss": 0.2543,
|
| 1349 |
+
"step": 1800
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 10.91183119819141,
|
| 1353 |
+
"grad_norm": 1.520704900113862,
|
| 1354 |
+
"learning_rate": 5e-06,
|
| 1355 |
+
"loss": 0.254,
|
| 1356 |
+
"step": 1810
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 10.972117558402411,
|
| 1360 |
+
"grad_norm": 1.4393868378454422,
|
| 1361 |
+
"learning_rate": 5e-06,
|
| 1362 |
+
"loss": 0.2565,
|
| 1363 |
+
"step": 1820
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 10.996232102486813,
|
| 1367 |
+
"eval_loss": 1.236780047416687,
|
| 1368 |
+
"eval_runtime": 115.9111,
|
| 1369 |
+
"eval_samples_per_second": 38.555,
|
| 1370 |
+
"eval_steps_per_second": 0.604,
|
| 1371 |
+
"step": 1824
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 11.032403918613413,
|
| 1375 |
+
"grad_norm": 2.415959560977028,
|
| 1376 |
+
"learning_rate": 5e-06,
|
| 1377 |
+
"loss": 0.2285,
|
| 1378 |
+
"step": 1830
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 11.092690278824415,
|
| 1382 |
+
"grad_norm": 1.5442333094800837,
|
| 1383 |
+
"learning_rate": 5e-06,
|
| 1384 |
+
"loss": 0.1975,
|
| 1385 |
+
"step": 1840
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 11.152976639035419,
|
| 1389 |
+
"grad_norm": 2.1161682250412257,
|
| 1390 |
+
"learning_rate": 5e-06,
|
| 1391 |
+
"loss": 0.1968,
|
| 1392 |
+
"step": 1850
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 11.213262999246421,
|
| 1396 |
+
"grad_norm": 1.6062184104077026,
|
| 1397 |
+
"learning_rate": 5e-06,
|
| 1398 |
+
"loss": 0.1967,
|
| 1399 |
+
"step": 1860
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 11.273549359457423,
|
| 1403 |
+
"grad_norm": 1.6876217090206025,
|
| 1404 |
+
"learning_rate": 5e-06,
|
| 1405 |
+
"loss": 0.1937,
|
| 1406 |
+
"step": 1870
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 11.333835719668425,
|
| 1410 |
+
"grad_norm": 1.756341727895367,
|
| 1411 |
+
"learning_rate": 5e-06,
|
| 1412 |
+
"loss": 0.1966,
|
| 1413 |
+
"step": 1880
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 11.394122079879427,
|
| 1417 |
+
"grad_norm": 1.5508699967732422,
|
| 1418 |
+
"learning_rate": 5e-06,
|
| 1419 |
+
"loss": 0.1997,
|
| 1420 |
+
"step": 1890
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 11.454408440090429,
|
| 1424 |
+
"grad_norm": 1.5669766694757912,
|
| 1425 |
+
"learning_rate": 5e-06,
|
| 1426 |
+
"loss": 0.2023,
|
| 1427 |
+
"step": 1900
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 11.51469480030143,
|
| 1431 |
+
"grad_norm": 1.4502558429874435,
|
| 1432 |
+
"learning_rate": 5e-06,
|
| 1433 |
+
"loss": 0.202,
|
| 1434 |
+
"step": 1910
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 11.574981160512435,
|
| 1438 |
+
"grad_norm": 1.432868245038754,
|
| 1439 |
+
"learning_rate": 5e-06,
|
| 1440 |
+
"loss": 0.2027,
|
| 1441 |
+
"step": 1920
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 11.635267520723437,
|
| 1445 |
+
"grad_norm": 1.944951997696116,
|
| 1446 |
+
"learning_rate": 5e-06,
|
| 1447 |
+
"loss": 0.2051,
|
| 1448 |
+
"step": 1930
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 11.695553880934439,
|
| 1452 |
+
"grad_norm": 1.5628919140219797,
|
| 1453 |
+
"learning_rate": 5e-06,
|
| 1454 |
+
"loss": 0.2047,
|
| 1455 |
+
"step": 1940
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 11.75584024114544,
|
| 1459 |
+
"grad_norm": 1.8554418586433572,
|
| 1460 |
+
"learning_rate": 5e-06,
|
| 1461 |
+
"loss": 0.2036,
|
| 1462 |
+
"step": 1950
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 11.816126601356443,
|
| 1466 |
+
"grad_norm": 1.5083234117594833,
|
| 1467 |
+
"learning_rate": 5e-06,
|
| 1468 |
+
"loss": 0.2057,
|
| 1469 |
+
"step": 1960
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 11.876412961567446,
|
| 1473 |
+
"grad_norm": 1.6850324704140989,
|
| 1474 |
+
"learning_rate": 5e-06,
|
| 1475 |
+
"loss": 0.2084,
|
| 1476 |
+
"step": 1970
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 11.936699321778448,
|
| 1480 |
+
"grad_norm": 1.518863675506513,
|
| 1481 |
+
"learning_rate": 5e-06,
|
| 1482 |
+
"loss": 0.2081,
|
| 1483 |
+
"step": 1980
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 11.99698568198945,
|
| 1487 |
+
"grad_norm": 2.025449919349873,
|
| 1488 |
+
"learning_rate": 5e-06,
|
| 1489 |
+
"loss": 0.2099,
|
| 1490 |
+
"step": 1990
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 11.99698568198945,
|
| 1494 |
+
"eval_loss": 1.3552354574203491,
|
| 1495 |
+
"eval_runtime": 116.1042,
|
| 1496 |
+
"eval_samples_per_second": 38.491,
|
| 1497 |
+
"eval_steps_per_second": 0.603,
|
| 1498 |
+
"step": 1990
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 12.057272042200452,
|
| 1502 |
+
"grad_norm": 1.9707275891443514,
|
| 1503 |
+
"learning_rate": 5e-06,
|
| 1504 |
+
"loss": 0.161,
|
| 1505 |
+
"step": 2000
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 12.117558402411454,
|
| 1509 |
+
"grad_norm": 1.60811233115537,
|
| 1510 |
+
"learning_rate": 5e-06,
|
| 1511 |
+
"loss": 0.1527,
|
| 1512 |
+
"step": 2010
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 12.177844762622456,
|
| 1516 |
+
"grad_norm": 1.6751301942391925,
|
| 1517 |
+
"learning_rate": 5e-06,
|
| 1518 |
+
"loss": 0.1541,
|
| 1519 |
+
"step": 2020
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 12.238131122833458,
|
| 1523 |
+
"grad_norm": 1.4492067853763235,
|
| 1524 |
+
"learning_rate": 5e-06,
|
| 1525 |
+
"loss": 0.1559,
|
| 1526 |
+
"step": 2030
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 12.298417483044462,
|
| 1530 |
+
"grad_norm": 1.9142452633048677,
|
| 1531 |
+
"learning_rate": 5e-06,
|
| 1532 |
+
"loss": 0.1576,
|
| 1533 |
+
"step": 2040
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 12.358703843255464,
|
| 1537 |
+
"grad_norm": 1.4385285667791092,
|
| 1538 |
+
"learning_rate": 5e-06,
|
| 1539 |
+
"loss": 0.1589,
|
| 1540 |
+
"step": 2050
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 12.418990203466466,
|
| 1544 |
+
"grad_norm": 1.4928747098321213,
|
| 1545 |
+
"learning_rate": 5e-06,
|
| 1546 |
+
"loss": 0.1608,
|
| 1547 |
+
"step": 2060
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 12.479276563677468,
|
| 1551 |
+
"grad_norm": 1.5019521655912937,
|
| 1552 |
+
"learning_rate": 5e-06,
|
| 1553 |
+
"loss": 0.1605,
|
| 1554 |
+
"step": 2070
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 12.53956292388847,
|
| 1558 |
+
"grad_norm": 1.874845955363987,
|
| 1559 |
+
"learning_rate": 5e-06,
|
| 1560 |
+
"loss": 0.1612,
|
| 1561 |
+
"step": 2080
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 12.599849284099472,
|
| 1565 |
+
"grad_norm": 1.885802643703012,
|
| 1566 |
+
"learning_rate": 5e-06,
|
| 1567 |
+
"loss": 0.1614,
|
| 1568 |
+
"step": 2090
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 12.660135644310476,
|
| 1572 |
+
"grad_norm": 1.515486697883317,
|
| 1573 |
+
"learning_rate": 5e-06,
|
| 1574 |
+
"loss": 0.1622,
|
| 1575 |
+
"step": 2100
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 12.720422004521478,
|
| 1579 |
+
"grad_norm": 1.7537896738178245,
|
| 1580 |
+
"learning_rate": 5e-06,
|
| 1581 |
+
"loss": 0.1644,
|
| 1582 |
+
"step": 2110
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 12.78070836473248,
|
| 1586 |
+
"grad_norm": 1.6471681441773147,
|
| 1587 |
+
"learning_rate": 5e-06,
|
| 1588 |
+
"loss": 0.1647,
|
| 1589 |
+
"step": 2120
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 12.840994724943481,
|
| 1593 |
+
"grad_norm": 1.590934554816475,
|
| 1594 |
+
"learning_rate": 5e-06,
|
| 1595 |
+
"loss": 0.1648,
|
| 1596 |
+
"step": 2130
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 12.901281085154483,
|
| 1600 |
+
"grad_norm": 1.7730249951617123,
|
| 1601 |
+
"learning_rate": 5e-06,
|
| 1602 |
+
"loss": 0.1656,
|
| 1603 |
+
"step": 2140
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 12.961567445365485,
|
| 1607 |
+
"grad_norm": 1.6645957171669876,
|
| 1608 |
+
"learning_rate": 5e-06,
|
| 1609 |
+
"loss": 0.1676,
|
| 1610 |
+
"step": 2150
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 12.997739261492088,
|
| 1614 |
+
"eval_loss": 1.5070669651031494,
|
| 1615 |
+
"eval_runtime": 116.1666,
|
| 1616 |
+
"eval_samples_per_second": 38.471,
|
| 1617 |
+
"eval_steps_per_second": 0.603,
|
| 1618 |
+
"step": 2156
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 13.02185380557649,
|
| 1622 |
+
"grad_norm": 2.462408072338029,
|
| 1623 |
+
"learning_rate": 5e-06,
|
| 1624 |
+
"loss": 0.1517,
|
| 1625 |
+
"step": 2160
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 13.082140165787491,
|
| 1629 |
+
"grad_norm": 2.4318360950844804,
|
| 1630 |
+
"learning_rate": 5e-06,
|
| 1631 |
+
"loss": 0.1195,
|
| 1632 |
+
"step": 2170
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 13.142426525998493,
|
| 1636 |
+
"grad_norm": 1.7426593578459117,
|
| 1637 |
+
"learning_rate": 5e-06,
|
| 1638 |
+
"loss": 0.1201,
|
| 1639 |
+
"step": 2180
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 13.202712886209495,
|
| 1643 |
+
"grad_norm": 2.2383839547509696,
|
| 1644 |
+
"learning_rate": 5e-06,
|
| 1645 |
+
"loss": 0.1191,
|
| 1646 |
+
"step": 2190
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 13.262999246420497,
|
| 1650 |
+
"grad_norm": 1.5553317913023317,
|
| 1651 |
+
"learning_rate": 5e-06,
|
| 1652 |
+
"loss": 0.118,
|
| 1653 |
+
"step": 2200
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 13.323285606631499,
|
| 1657 |
+
"grad_norm": 1.4669913420252614,
|
| 1658 |
+
"learning_rate": 5e-06,
|
| 1659 |
+
"loss": 0.119,
|
| 1660 |
+
"step": 2210
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 13.383571966842501,
|
| 1664 |
+
"grad_norm": 1.8080283494554128,
|
| 1665 |
+
"learning_rate": 5e-06,
|
| 1666 |
+
"loss": 0.1196,
|
| 1667 |
+
"step": 2220
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 13.443858327053505,
|
| 1671 |
+
"grad_norm": 1.646647327329277,
|
| 1672 |
+
"learning_rate": 5e-06,
|
| 1673 |
+
"loss": 0.1233,
|
| 1674 |
+
"step": 2230
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 13.504144687264507,
|
| 1678 |
+
"grad_norm": 1.8533611593149397,
|
| 1679 |
+
"learning_rate": 5e-06,
|
| 1680 |
+
"loss": 0.1241,
|
| 1681 |
+
"step": 2240
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 13.564431047475509,
|
| 1685 |
+
"grad_norm": 1.6777607818909321,
|
| 1686 |
+
"learning_rate": 5e-06,
|
| 1687 |
+
"loss": 0.126,
|
| 1688 |
+
"step": 2250
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 13.62471740768651,
|
| 1692 |
+
"grad_norm": 1.7745416454251923,
|
| 1693 |
+
"learning_rate": 5e-06,
|
| 1694 |
+
"loss": 0.1266,
|
| 1695 |
+
"step": 2260
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 13.685003767897513,
|
| 1699 |
+
"grad_norm": 1.7096006593564836,
|
| 1700 |
+
"learning_rate": 5e-06,
|
| 1701 |
+
"loss": 0.1282,
|
| 1702 |
+
"step": 2270
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 13.745290128108515,
|
| 1706 |
+
"grad_norm": 1.8638678064551115,
|
| 1707 |
+
"learning_rate": 5e-06,
|
| 1708 |
+
"loss": 0.1267,
|
| 1709 |
+
"step": 2280
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 13.805576488319518,
|
| 1713 |
+
"grad_norm": 1.4262510400727582,
|
| 1714 |
+
"learning_rate": 5e-06,
|
| 1715 |
+
"loss": 0.1273,
|
| 1716 |
+
"step": 2290
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 13.86586284853052,
|
| 1720 |
+
"grad_norm": 1.8828592364046501,
|
| 1721 |
+
"learning_rate": 5e-06,
|
| 1722 |
+
"loss": 0.1269,
|
| 1723 |
+
"step": 2300
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 13.926149208741522,
|
| 1727 |
+
"grad_norm": 1.6583448018945106,
|
| 1728 |
+
"learning_rate": 5e-06,
|
| 1729 |
+
"loss": 0.1273,
|
| 1730 |
+
"step": 2310
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 13.986435568952524,
|
| 1734 |
+
"grad_norm": 1.548491818827999,
|
| 1735 |
+
"learning_rate": 5e-06,
|
| 1736 |
+
"loss": 0.1283,
|
| 1737 |
+
"step": 2320
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 13.998492840994725,
|
| 1741 |
+
"eval_loss": 1.6323552131652832,
|
| 1742 |
+
"eval_runtime": 115.7608,
|
| 1743 |
+
"eval_samples_per_second": 38.605,
|
| 1744 |
+
"eval_steps_per_second": 0.605,
|
| 1745 |
+
"step": 2322
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 14.046721929163526,
|
| 1749 |
+
"grad_norm": 2.0937924453069203,
|
| 1750 |
+
"learning_rate": 5e-06,
|
| 1751 |
+
"loss": 0.1006,
|
| 1752 |
+
"step": 2330
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 14.107008289374528,
|
| 1756 |
+
"grad_norm": 2.139024110314688,
|
| 1757 |
+
"learning_rate": 5e-06,
|
| 1758 |
+
"loss": 0.0897,
|
| 1759 |
+
"step": 2340
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 14.167294649585532,
|
| 1763 |
+
"grad_norm": 1.5948191216888472,
|
| 1764 |
+
"learning_rate": 5e-06,
|
| 1765 |
+
"loss": 0.092,
|
| 1766 |
+
"step": 2350
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 14.227581009796534,
|
| 1770 |
+
"grad_norm": 1.6158036371008326,
|
| 1771 |
+
"learning_rate": 5e-06,
|
| 1772 |
+
"loss": 0.0916,
|
| 1773 |
+
"step": 2360
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 14.287867370007536,
|
| 1777 |
+
"grad_norm": 1.6380184320632138,
|
| 1778 |
+
"learning_rate": 5e-06,
|
| 1779 |
+
"loss": 0.0946,
|
| 1780 |
+
"step": 2370
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 14.348153730218538,
|
| 1784 |
+
"grad_norm": 1.5125683894811608,
|
| 1785 |
+
"learning_rate": 5e-06,
|
| 1786 |
+
"loss": 0.0926,
|
| 1787 |
+
"step": 2380
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 14.40844009042954,
|
| 1791 |
+
"grad_norm": 1.8291017955611948,
|
| 1792 |
+
"learning_rate": 5e-06,
|
| 1793 |
+
"loss": 0.0967,
|
| 1794 |
+
"step": 2390
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 14.468726450640542,
|
| 1798 |
+
"grad_norm": 1.6743144432735446,
|
| 1799 |
+
"learning_rate": 5e-06,
|
| 1800 |
+
"loss": 0.097,
|
| 1801 |
+
"step": 2400
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 14.529012810851544,
|
| 1805 |
+
"grad_norm": 1.7276135456842643,
|
| 1806 |
+
"learning_rate": 5e-06,
|
| 1807 |
+
"loss": 0.096,
|
| 1808 |
+
"step": 2410
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 14.589299171062548,
|
| 1812 |
+
"grad_norm": 1.734533902590376,
|
| 1813 |
+
"learning_rate": 5e-06,
|
| 1814 |
+
"loss": 0.0977,
|
| 1815 |
+
"step": 2420
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 14.64958553127355,
|
| 1819 |
+
"grad_norm": 1.7106037137785368,
|
| 1820 |
+
"learning_rate": 5e-06,
|
| 1821 |
+
"loss": 0.0974,
|
| 1822 |
+
"step": 2430
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 14.709871891484552,
|
| 1826 |
+
"grad_norm": 1.848167736152076,
|
| 1827 |
+
"learning_rate": 5e-06,
|
| 1828 |
+
"loss": 0.098,
|
| 1829 |
+
"step": 2440
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 14.770158251695554,
|
| 1833 |
+
"grad_norm": 2.0729946870310907,
|
| 1834 |
+
"learning_rate": 5e-06,
|
| 1835 |
+
"loss": 0.0988,
|
| 1836 |
+
"step": 2450
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 14.830444611906556,
|
| 1840 |
+
"grad_norm": 1.6522773246662863,
|
| 1841 |
+
"learning_rate": 5e-06,
|
| 1842 |
+
"loss": 0.0996,
|
| 1843 |
+
"step": 2460
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 14.890730972117558,
|
| 1847 |
+
"grad_norm": 1.8165186575808512,
|
| 1848 |
+
"learning_rate": 5e-06,
|
| 1849 |
+
"loss": 0.1018,
|
| 1850 |
+
"step": 2470
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 14.951017332328561,
|
| 1854 |
+
"grad_norm": 1.8324940323277443,
|
| 1855 |
+
"learning_rate": 5e-06,
|
| 1856 |
+
"loss": 0.1022,
|
| 1857 |
+
"step": 2480
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 14.999246420497363,
|
| 1861 |
+
"eval_loss": 1.7541841268539429,
|
| 1862 |
+
"eval_runtime": 116.4567,
|
| 1863 |
+
"eval_samples_per_second": 38.375,
|
| 1864 |
+
"eval_steps_per_second": 0.601,
|
| 1865 |
+
"step": 2488
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 15.011303692539563,
|
| 1869 |
+
"grad_norm": 1.7122346572238332,
|
| 1870 |
+
"learning_rate": 5e-06,
|
| 1871 |
+
"loss": 0.0969,
|
| 1872 |
+
"step": 2490
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 15.071590052750565,
|
| 1876 |
+
"grad_norm": 1.7869038836513413,
|
| 1877 |
+
"learning_rate": 5e-06,
|
| 1878 |
+
"loss": 0.07,
|
| 1879 |
+
"step": 2500
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 15.131876412961567,
|
| 1883 |
+
"grad_norm": 1.5692589038411329,
|
| 1884 |
+
"learning_rate": 5e-06,
|
| 1885 |
+
"loss": 0.069,
|
| 1886 |
+
"step": 2510
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 15.19216277317257,
|
| 1890 |
+
"grad_norm": 1.5568543176798626,
|
| 1891 |
+
"learning_rate": 5e-06,
|
| 1892 |
+
"loss": 0.0714,
|
| 1893 |
+
"step": 2520
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 15.252449133383571,
|
| 1897 |
+
"grad_norm": 1.8979966448489778,
|
| 1898 |
+
"learning_rate": 5e-06,
|
| 1899 |
+
"loss": 0.0721,
|
| 1900 |
+
"step": 2530
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 15.312735493594575,
|
| 1904 |
+
"grad_norm": 1.6079155193057422,
|
| 1905 |
+
"learning_rate": 5e-06,
|
| 1906 |
+
"loss": 0.0734,
|
| 1907 |
+
"step": 2540
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 15.373021853805577,
|
| 1911 |
+
"grad_norm": 1.5682954069117738,
|
| 1912 |
+
"learning_rate": 5e-06,
|
| 1913 |
+
"loss": 0.0731,
|
| 1914 |
+
"step": 2550
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 15.433308214016579,
|
| 1918 |
+
"grad_norm": 1.607771288791899,
|
| 1919 |
+
"learning_rate": 5e-06,
|
| 1920 |
+
"loss": 0.0736,
|
| 1921 |
+
"step": 2560
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 15.493594574227581,
|
| 1925 |
+
"grad_norm": 1.592532966261963,
|
| 1926 |
+
"learning_rate": 5e-06,
|
| 1927 |
+
"loss": 0.0751,
|
| 1928 |
+
"step": 2570
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 15.553880934438583,
|
| 1932 |
+
"grad_norm": 2.106042369619277,
|
| 1933 |
+
"learning_rate": 5e-06,
|
| 1934 |
+
"loss": 0.0765,
|
| 1935 |
+
"step": 2580
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 15.614167294649585,
|
| 1939 |
+
"grad_norm": 1.9643387382353286,
|
| 1940 |
+
"learning_rate": 5e-06,
|
| 1941 |
+
"loss": 0.0764,
|
| 1942 |
+
"step": 2590
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 15.674453654860589,
|
| 1946 |
+
"grad_norm": 2.356472326588705,
|
| 1947 |
+
"learning_rate": 5e-06,
|
| 1948 |
+
"loss": 0.0766,
|
| 1949 |
+
"step": 2600
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 15.73474001507159,
|
| 1953 |
+
"grad_norm": 1.7285204414699735,
|
| 1954 |
+
"learning_rate": 5e-06,
|
| 1955 |
+
"loss": 0.0768,
|
| 1956 |
+
"step": 2610
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 15.795026375282593,
|
| 1960 |
+
"grad_norm": 1.656141946179993,
|
| 1961 |
+
"learning_rate": 5e-06,
|
| 1962 |
+
"loss": 0.0781,
|
| 1963 |
+
"step": 2620
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 15.855312735493595,
|
| 1967 |
+
"grad_norm": 1.751656627220462,
|
| 1968 |
+
"learning_rate": 5e-06,
|
| 1969 |
+
"loss": 0.0778,
|
| 1970 |
+
"step": 2630
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 15.915599095704597,
|
| 1974 |
+
"grad_norm": 1.9120339283286005,
|
| 1975 |
+
"learning_rate": 5e-06,
|
| 1976 |
+
"loss": 0.0777,
|
| 1977 |
+
"step": 2640
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 15.975885455915598,
|
| 1981 |
+
"grad_norm": 1.7091309765672063,
|
| 1982 |
+
"learning_rate": 5e-06,
|
| 1983 |
+
"loss": 0.0779,
|
| 1984 |
+
"step": 2650
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 16.0,
|
| 1988 |
+
"eval_loss": 1.8729431629180908,
|
| 1989 |
+
"eval_runtime": 130.3121,
|
| 1990 |
+
"eval_samples_per_second": 34.295,
|
| 1991 |
+
"eval_steps_per_second": 0.537,
|
| 1992 |
+
"step": 2654
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 16.0361718161266,
|
| 1996 |
+
"grad_norm": 1.6873009916146906,
|
| 1997 |
+
"learning_rate": 5e-06,
|
| 1998 |
+
"loss": 0.0634,
|
| 1999 |
+
"step": 2660
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 16.096458176337602,
|
| 2003 |
+
"grad_norm": 1.4499459070972593,
|
| 2004 |
+
"learning_rate": 5e-06,
|
| 2005 |
+
"loss": 0.0517,
|
| 2006 |
+
"step": 2670
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 16.156744536548604,
|
| 2010 |
+
"grad_norm": 1.518251208097316,
|
| 2011 |
+
"learning_rate": 5e-06,
|
| 2012 |
+
"loss": 0.0532,
|
| 2013 |
+
"step": 2680
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 16.217030896759606,
|
| 2017 |
+
"grad_norm": 1.7901282259982598,
|
| 2018 |
+
"learning_rate": 5e-06,
|
| 2019 |
+
"loss": 0.0547,
|
| 2020 |
+
"step": 2690
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 16.277317256970612,
|
| 2024 |
+
"grad_norm": 1.6296841948609884,
|
| 2025 |
+
"learning_rate": 5e-06,
|
| 2026 |
+
"loss": 0.0553,
|
| 2027 |
+
"step": 2700
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 16.337603617181614,
|
| 2031 |
+
"grad_norm": 1.9077348260850149,
|
| 2032 |
+
"learning_rate": 5e-06,
|
| 2033 |
+
"loss": 0.0551,
|
| 2034 |
+
"step": 2710
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 16.397889977392616,
|
| 2038 |
+
"grad_norm": 1.6275792847492503,
|
| 2039 |
+
"learning_rate": 5e-06,
|
| 2040 |
+
"loss": 0.0556,
|
| 2041 |
+
"step": 2720
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 16.458176337603618,
|
| 2045 |
+
"grad_norm": 1.8060616409571004,
|
| 2046 |
+
"learning_rate": 5e-06,
|
| 2047 |
+
"loss": 0.0569,
|
| 2048 |
+
"step": 2730
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 16.51846269781462,
|
| 2052 |
+
"grad_norm": 1.6503072239622354,
|
| 2053 |
+
"learning_rate": 5e-06,
|
| 2054 |
+
"loss": 0.0568,
|
| 2055 |
+
"step": 2740
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 16.578749058025622,
|
| 2059 |
+
"grad_norm": 1.5506661680891511,
|
| 2060 |
+
"learning_rate": 5e-06,
|
| 2061 |
+
"loss": 0.0572,
|
| 2062 |
+
"step": 2750
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 16.639035418236624,
|
| 2066 |
+
"grad_norm": 1.6807035773937569,
|
| 2067 |
+
"learning_rate": 5e-06,
|
| 2068 |
+
"loss": 0.0583,
|
| 2069 |
+
"step": 2760
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 16.699321778447626,
|
| 2073 |
+
"grad_norm": 1.794347099575113,
|
| 2074 |
+
"learning_rate": 5e-06,
|
| 2075 |
+
"loss": 0.0589,
|
| 2076 |
+
"step": 2770
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 16.759608138658628,
|
| 2080 |
+
"grad_norm": 1.6112070974863837,
|
| 2081 |
+
"learning_rate": 5e-06,
|
| 2082 |
+
"loss": 0.0591,
|
| 2083 |
+
"step": 2780
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 16.81989449886963,
|
| 2087 |
+
"grad_norm": 1.5309535448790481,
|
| 2088 |
+
"learning_rate": 5e-06,
|
| 2089 |
+
"loss": 0.0595,
|
| 2090 |
+
"step": 2790
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 16.88018085908063,
|
| 2094 |
+
"grad_norm": 1.6962829999060645,
|
| 2095 |
+
"learning_rate": 5e-06,
|
| 2096 |
+
"loss": 0.0595,
|
| 2097 |
+
"step": 2800
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 16.940467219291634,
|
| 2101 |
+
"grad_norm": 1.5160770218318527,
|
| 2102 |
+
"learning_rate": 5e-06,
|
| 2103 |
+
"loss": 0.0607,
|
| 2104 |
+
"step": 2810
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 16.994724943481536,
|
| 2108 |
+
"eval_loss": 1.9862152338027954,
|
| 2109 |
+
"eval_runtime": 135.1067,
|
| 2110 |
+
"eval_samples_per_second": 33.078,
|
| 2111 |
+
"eval_steps_per_second": 0.518,
|
| 2112 |
+
"step": 2819
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 17.00075357950264,
|
| 2116 |
+
"grad_norm": 2.340332137108714,
|
| 2117 |
+
"learning_rate": 5e-06,
|
| 2118 |
+
"loss": 0.0615,
|
| 2119 |
+
"step": 2820
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 17.06103993971364,
|
| 2123 |
+
"grad_norm": 1.4236109373520875,
|
| 2124 |
+
"learning_rate": 5e-06,
|
| 2125 |
+
"loss": 0.0397,
|
| 2126 |
+
"step": 2830
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 17.121326299924643,
|
| 2130 |
+
"grad_norm": 1.5339925411690258,
|
| 2131 |
+
"learning_rate": 5e-06,
|
| 2132 |
+
"loss": 0.0409,
|
| 2133 |
+
"step": 2840
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 17.181612660135645,
|
| 2137 |
+
"grad_norm": 1.5223509157958206,
|
| 2138 |
+
"learning_rate": 5e-06,
|
| 2139 |
+
"loss": 0.0416,
|
| 2140 |
+
"step": 2850
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 17.241899020346647,
|
| 2144 |
+
"grad_norm": 1.7444832258175693,
|
| 2145 |
+
"learning_rate": 5e-06,
|
| 2146 |
+
"loss": 0.0425,
|
| 2147 |
+
"step": 2860
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 17.30218538055765,
|
| 2151 |
+
"grad_norm": 1.7829806133150694,
|
| 2152 |
+
"learning_rate": 5e-06,
|
| 2153 |
+
"loss": 0.0426,
|
| 2154 |
+
"step": 2870
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 17.36247174076865,
|
| 2158 |
+
"grad_norm": 1.7776363684895005,
|
| 2159 |
+
"learning_rate": 5e-06,
|
| 2160 |
+
"loss": 0.0437,
|
| 2161 |
+
"step": 2880
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 17.422758100979653,
|
| 2165 |
+
"grad_norm": 1.449976018411429,
|
| 2166 |
+
"learning_rate": 5e-06,
|
| 2167 |
+
"loss": 0.0434,
|
| 2168 |
+
"step": 2890
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 17.483044461190655,
|
| 2172 |
+
"grad_norm": 1.4737183761321249,
|
| 2173 |
+
"learning_rate": 5e-06,
|
| 2174 |
+
"loss": 0.0443,
|
| 2175 |
+
"step": 2900
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 17.543330821401657,
|
| 2179 |
+
"grad_norm": 1.5570813152529674,
|
| 2180 |
+
"learning_rate": 5e-06,
|
| 2181 |
+
"loss": 0.0439,
|
| 2182 |
+
"step": 2910
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 17.60361718161266,
|
| 2186 |
+
"grad_norm": 1.6635463270221298,
|
| 2187 |
+
"learning_rate": 5e-06,
|
| 2188 |
+
"loss": 0.0451,
|
| 2189 |
+
"step": 2920
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 17.66390354182366,
|
| 2193 |
+
"grad_norm": 1.5633000807042863,
|
| 2194 |
+
"learning_rate": 5e-06,
|
| 2195 |
+
"loss": 0.0452,
|
| 2196 |
+
"step": 2930
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 17.724189902034666,
|
| 2200 |
+
"grad_norm": 1.5756718088358965,
|
| 2201 |
+
"learning_rate": 5e-06,
|
| 2202 |
+
"loss": 0.0453,
|
| 2203 |
+
"step": 2940
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 17.78447626224567,
|
| 2207 |
+
"grad_norm": 1.5086652524908912,
|
| 2208 |
+
"learning_rate": 5e-06,
|
| 2209 |
+
"loss": 0.0461,
|
| 2210 |
+
"step": 2950
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 17.84476262245667,
|
| 2214 |
+
"grad_norm": 1.543593492657027,
|
| 2215 |
+
"learning_rate": 5e-06,
|
| 2216 |
+
"loss": 0.0471,
|
| 2217 |
+
"step": 2960
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 17.905048982667672,
|
| 2221 |
+
"grad_norm": 1.8850354234893694,
|
| 2222 |
+
"learning_rate": 5e-06,
|
| 2223 |
+
"loss": 0.0473,
|
| 2224 |
+
"step": 2970
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 17.965335342878674,
|
| 2228 |
+
"grad_norm": 1.5531948749022353,
|
| 2229 |
+
"learning_rate": 5e-06,
|
| 2230 |
+
"loss": 0.0481,
|
| 2231 |
+
"step": 2980
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 17.995478522984175,
|
| 2235 |
+
"eval_loss": 2.0547046661376953,
|
| 2236 |
+
"eval_runtime": 131.4004,
|
| 2237 |
+
"eval_samples_per_second": 34.011,
|
| 2238 |
+
"eval_steps_per_second": 0.533,
|
| 2239 |
+
"step": 2985
|
| 2240 |
+
},
|
| 2241 |
+
{
|
| 2242 |
+
"epoch": 18.025621703089676,
|
| 2243 |
+
"grad_norm": 1.5738867889136725,
|
| 2244 |
+
"learning_rate": 5e-06,
|
| 2245 |
+
"loss": 0.041,
|
| 2246 |
+
"step": 2990
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 18.08590806330068,
|
| 2250 |
+
"grad_norm": 1.7413824765739192,
|
| 2251 |
+
"learning_rate": 5e-06,
|
| 2252 |
+
"loss": 0.0319,
|
| 2253 |
+
"step": 3000
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 18.14619442351168,
|
| 2257 |
+
"grad_norm": 1.349257585093222,
|
| 2258 |
+
"learning_rate": 5e-06,
|
| 2259 |
+
"loss": 0.0321,
|
| 2260 |
+
"step": 3010
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 18.206480783722682,
|
| 2264 |
+
"grad_norm": 1.5030527061221075,
|
| 2265 |
+
"learning_rate": 5e-06,
|
| 2266 |
+
"loss": 0.033,
|
| 2267 |
+
"step": 3020
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 18.266767143933684,
|
| 2271 |
+
"grad_norm": 1.3839757017227572,
|
| 2272 |
+
"learning_rate": 5e-06,
|
| 2273 |
+
"loss": 0.0338,
|
| 2274 |
+
"step": 3030
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 18.327053504144686,
|
| 2278 |
+
"grad_norm": 1.3639549234993298,
|
| 2279 |
+
"learning_rate": 5e-06,
|
| 2280 |
+
"loss": 0.034,
|
| 2281 |
+
"step": 3040
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 18.38733986435569,
|
| 2285 |
+
"grad_norm": 1.491804900184454,
|
| 2286 |
+
"learning_rate": 5e-06,
|
| 2287 |
+
"loss": 0.0344,
|
| 2288 |
+
"step": 3050
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 18.44762622456669,
|
| 2292 |
+
"grad_norm": 1.7182953986203733,
|
| 2293 |
+
"learning_rate": 5e-06,
|
| 2294 |
+
"loss": 0.0345,
|
| 2295 |
+
"step": 3060
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 18.507912584777696,
|
| 2299 |
+
"grad_norm": 1.5927159416873042,
|
| 2300 |
+
"learning_rate": 5e-06,
|
| 2301 |
+
"loss": 0.035,
|
| 2302 |
+
"step": 3070
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 18.568198944988698,
|
| 2306 |
+
"grad_norm": 1.48048633519207,
|
| 2307 |
+
"learning_rate": 5e-06,
|
| 2308 |
+
"loss": 0.0351,
|
| 2309 |
+
"step": 3080
|
| 2310 |
+
},
|
| 2311 |
+
{
|
| 2312 |
+
"epoch": 18.6284853051997,
|
| 2313 |
+
"grad_norm": 1.5724972628180884,
|
| 2314 |
+
"learning_rate": 5e-06,
|
| 2315 |
+
"loss": 0.0357,
|
| 2316 |
+
"step": 3090
|
| 2317 |
+
},
|
| 2318 |
+
{
|
| 2319 |
+
"epoch": 18.6887716654107,
|
| 2320 |
+
"grad_norm": 1.460368142482016,
|
| 2321 |
+
"learning_rate": 5e-06,
|
| 2322 |
+
"loss": 0.0367,
|
| 2323 |
+
"step": 3100
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"epoch": 18.749058025621704,
|
| 2327 |
+
"grad_norm": 1.5605571994020253,
|
| 2328 |
+
"learning_rate": 5e-06,
|
| 2329 |
+
"loss": 0.037,
|
| 2330 |
+
"step": 3110
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 18.809344385832706,
|
| 2334 |
+
"grad_norm": 1.809142213762153,
|
| 2335 |
+
"learning_rate": 5e-06,
|
| 2336 |
+
"loss": 0.0364,
|
| 2337 |
+
"step": 3120
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 18.869630746043708,
|
| 2341 |
+
"grad_norm": 1.6290220037965493,
|
| 2342 |
+
"learning_rate": 5e-06,
|
| 2343 |
+
"loss": 0.0367,
|
| 2344 |
+
"step": 3130
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 18.92991710625471,
|
| 2348 |
+
"grad_norm": 1.441383521222635,
|
| 2349 |
+
"learning_rate": 5e-06,
|
| 2350 |
+
"loss": 0.0373,
|
| 2351 |
+
"step": 3140
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 18.99020346646571,
|
| 2355 |
+
"grad_norm": 1.4650505575255777,
|
| 2356 |
+
"learning_rate": 5e-06,
|
| 2357 |
+
"loss": 0.038,
|
| 2358 |
+
"step": 3150
|
| 2359 |
+
},
|
| 2360 |
+
{
|
| 2361 |
+
"epoch": 18.99623210248681,
|
| 2362 |
+
"eval_loss": 2.1350691318511963,
|
| 2363 |
+
"eval_runtime": 116.4178,
|
| 2364 |
+
"eval_samples_per_second": 38.388,
|
| 2365 |
+
"eval_steps_per_second": 0.601,
|
| 2366 |
+
"step": 3151
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 19.050489826676714,
|
| 2370 |
+
"grad_norm": 1.2364462769526054,
|
| 2371 |
+
"learning_rate": 5e-06,
|
| 2372 |
+
"loss": 0.0273,
|
| 2373 |
+
"step": 3160
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 19.110776186887716,
|
| 2377 |
+
"grad_norm": 1.2847338125252712,
|
| 2378 |
+
"learning_rate": 5e-06,
|
| 2379 |
+
"loss": 0.0263,
|
| 2380 |
+
"step": 3170
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 19.171062547098717,
|
| 2384 |
+
"grad_norm": 1.4364005652974698,
|
| 2385 |
+
"learning_rate": 5e-06,
|
| 2386 |
+
"loss": 0.027,
|
| 2387 |
+
"step": 3180
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 19.23134890730972,
|
| 2391 |
+
"grad_norm": 1.4324790800038558,
|
| 2392 |
+
"learning_rate": 5e-06,
|
| 2393 |
+
"loss": 0.0278,
|
| 2394 |
+
"step": 3190
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 19.291635267520725,
|
| 2398 |
+
"grad_norm": 1.3696161004609035,
|
| 2399 |
+
"learning_rate": 5e-06,
|
| 2400 |
+
"loss": 0.0268,
|
| 2401 |
+
"step": 3200
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 19.351921627731727,
|
| 2405 |
+
"grad_norm": 1.4042401515519942,
|
| 2406 |
+
"learning_rate": 5e-06,
|
| 2407 |
+
"loss": 0.0276,
|
| 2408 |
+
"step": 3210
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 19.41220798794273,
|
| 2412 |
+
"grad_norm": 1.4201103499411916,
|
| 2413 |
+
"learning_rate": 5e-06,
|
| 2414 |
+
"loss": 0.0282,
|
| 2415 |
+
"step": 3220
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 19.47249434815373,
|
| 2419 |
+
"grad_norm": 1.4787498619597446,
|
| 2420 |
+
"learning_rate": 5e-06,
|
| 2421 |
+
"loss": 0.029,
|
| 2422 |
+
"step": 3230
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 19.532780708364733,
|
| 2426 |
+
"grad_norm": 1.3042565121452792,
|
| 2427 |
+
"learning_rate": 5e-06,
|
| 2428 |
+
"loss": 0.0289,
|
| 2429 |
+
"step": 3240
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 19.593067068575735,
|
| 2433 |
+
"grad_norm": 1.5616956908222805,
|
| 2434 |
+
"learning_rate": 5e-06,
|
| 2435 |
+
"loss": 0.0291,
|
| 2436 |
+
"step": 3250
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 19.653353428786737,
|
| 2440 |
+
"grad_norm": 1.3916636056123413,
|
| 2441 |
+
"learning_rate": 5e-06,
|
| 2442 |
+
"loss": 0.0294,
|
| 2443 |
+
"step": 3260
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 19.71363978899774,
|
| 2447 |
+
"grad_norm": 1.5340949993651216,
|
| 2448 |
+
"learning_rate": 5e-06,
|
| 2449 |
+
"loss": 0.0302,
|
| 2450 |
+
"step": 3270
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 19.77392614920874,
|
| 2454 |
+
"grad_norm": 1.567955217260001,
|
| 2455 |
+
"learning_rate": 5e-06,
|
| 2456 |
+
"loss": 0.03,
|
| 2457 |
+
"step": 3280
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 19.834212509419743,
|
| 2461 |
+
"grad_norm": 1.4614834017683382,
|
| 2462 |
+
"learning_rate": 5e-06,
|
| 2463 |
+
"loss": 0.0302,
|
| 2464 |
+
"step": 3290
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 19.894498869630745,
|
| 2468 |
+
"grad_norm": 1.3384969584915662,
|
| 2469 |
+
"learning_rate": 5e-06,
|
| 2470 |
+
"loss": 0.0299,
|
| 2471 |
+
"step": 3300
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 19.954785229841747,
|
| 2475 |
+
"grad_norm": 1.6563798754782657,
|
| 2476 |
+
"learning_rate": 5e-06,
|
| 2477 |
+
"loss": 0.0306,
|
| 2478 |
+
"step": 3310
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 19.99698568198945,
|
| 2482 |
+
"eval_loss": 2.2255122661590576,
|
| 2483 |
+
"eval_runtime": 131.2328,
|
| 2484 |
+
"eval_samples_per_second": 34.054,
|
| 2485 |
+
"eval_steps_per_second": 0.533,
|
| 2486 |
+
"step": 3317
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 20.015071590052752,
|
| 2490 |
+
"grad_norm": 1.2478793413376605,
|
| 2491 |
+
"learning_rate": 5e-06,
|
| 2492 |
+
"loss": 0.0282,
|
| 2493 |
+
"step": 3320
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 20.075357950263754,
|
| 2497 |
+
"grad_norm": 1.5652614584389117,
|
| 2498 |
+
"learning_rate": 5e-06,
|
| 2499 |
+
"loss": 0.0201,
|
| 2500 |
+
"step": 3330
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 20.135644310474756,
|
| 2504 |
+
"grad_norm": 1.6731787596786802,
|
| 2505 |
+
"learning_rate": 5e-06,
|
| 2506 |
+
"loss": 0.0215,
|
| 2507 |
+
"step": 3340
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 20.195930670685758,
|
| 2511 |
+
"grad_norm": 1.2852656312690158,
|
| 2512 |
+
"learning_rate": 5e-06,
|
| 2513 |
+
"loss": 0.0215,
|
| 2514 |
+
"step": 3350
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 20.25621703089676,
|
| 2518 |
+
"grad_norm": 1.189629660977167,
|
| 2519 |
+
"learning_rate": 5e-06,
|
| 2520 |
+
"loss": 0.0218,
|
| 2521 |
+
"step": 3360
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 20.316503391107762,
|
| 2525 |
+
"grad_norm": 1.280211241755101,
|
| 2526 |
+
"learning_rate": 5e-06,
|
| 2527 |
+
"loss": 0.0225,
|
| 2528 |
+
"step": 3370
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 20.376789751318764,
|
| 2532 |
+
"grad_norm": 1.3671510362477786,
|
| 2533 |
+
"learning_rate": 5e-06,
|
| 2534 |
+
"loss": 0.0233,
|
| 2535 |
+
"step": 3380
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 20.437076111529766,
|
| 2539 |
+
"grad_norm": 1.2479455610052839,
|
| 2540 |
+
"learning_rate": 5e-06,
|
| 2541 |
+
"loss": 0.0235,
|
| 2542 |
+
"step": 3390
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 20.497362471740768,
|
| 2546 |
+
"grad_norm": 1.3487469571257409,
|
| 2547 |
+
"learning_rate": 5e-06,
|
| 2548 |
+
"loss": 0.024,
|
| 2549 |
+
"step": 3400
|
| 2550 |
+
},
|
| 2551 |
+
{
|
| 2552 |
+
"epoch": 20.55764883195177,
|
| 2553 |
+
"grad_norm": 1.296539761041376,
|
| 2554 |
+
"learning_rate": 5e-06,
|
| 2555 |
+
"loss": 0.0239,
|
| 2556 |
+
"step": 3410
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 20.617935192162772,
|
| 2560 |
+
"grad_norm": 1.3688473131734242,
|
| 2561 |
+
"learning_rate": 5e-06,
|
| 2562 |
+
"loss": 0.0243,
|
| 2563 |
+
"step": 3420
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 20.678221552373774,
|
| 2567 |
+
"grad_norm": 1.3764265527684842,
|
| 2568 |
+
"learning_rate": 5e-06,
|
| 2569 |
+
"loss": 0.0246,
|
| 2570 |
+
"step": 3430
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 20.738507912584776,
|
| 2574 |
+
"grad_norm": 1.5833602766458634,
|
| 2575 |
+
"learning_rate": 5e-06,
|
| 2576 |
+
"loss": 0.0249,
|
| 2577 |
+
"step": 3440
|
| 2578 |
+
},
|
| 2579 |
+
{
|
| 2580 |
+
"epoch": 20.79879427279578,
|
| 2581 |
+
"grad_norm": 1.528464092769155,
|
| 2582 |
+
"learning_rate": 5e-06,
|
| 2583 |
+
"loss": 0.0258,
|
| 2584 |
+
"step": 3450
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 20.859080633006784,
|
| 2588 |
+
"grad_norm": 1.605313026206358,
|
| 2589 |
+
"learning_rate": 5e-06,
|
| 2590 |
+
"loss": 0.026,
|
| 2591 |
+
"step": 3460
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 20.919366993217785,
|
| 2595 |
+
"grad_norm": 1.3923344492334147,
|
| 2596 |
+
"learning_rate": 5e-06,
|
| 2597 |
+
"loss": 0.0259,
|
| 2598 |
+
"step": 3470
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 20.979653353428787,
|
| 2602 |
+
"grad_norm": 1.4204063096393778,
|
| 2603 |
+
"learning_rate": 5e-06,
|
| 2604 |
+
"loss": 0.0256,
|
| 2605 |
+
"step": 3480
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 20.997739261492086,
|
| 2609 |
+
"eval_loss": 2.269946813583374,
|
| 2610 |
+
"eval_runtime": 116.5456,
|
| 2611 |
+
"eval_samples_per_second": 38.345,
|
| 2612 |
+
"eval_steps_per_second": 0.601,
|
| 2613 |
+
"step": 3483
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 21.03993971363979,
|
| 2617 |
+
"grad_norm": 1.310936775235027,
|
| 2618 |
+
"learning_rate": 5e-06,
|
| 2619 |
+
"loss": 0.0208,
|
| 2620 |
+
"step": 3490
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 21.10022607385079,
|
| 2624 |
+
"grad_norm": 1.1547005623677518,
|
| 2625 |
+
"learning_rate": 5e-06,
|
| 2626 |
+
"loss": 0.0186,
|
| 2627 |
+
"step": 3500
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 21.160512434061793,
|
| 2631 |
+
"grad_norm": 1.2938934451769595,
|
| 2632 |
+
"learning_rate": 5e-06,
|
| 2633 |
+
"loss": 0.0194,
|
| 2634 |
+
"step": 3510
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 21.220798794272795,
|
| 2638 |
+
"grad_norm": 1.484320867371485,
|
| 2639 |
+
"learning_rate": 5e-06,
|
| 2640 |
+
"loss": 0.0196,
|
| 2641 |
+
"step": 3520
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 21.281085154483797,
|
| 2645 |
+
"grad_norm": 1.4005288463687333,
|
| 2646 |
+
"learning_rate": 5e-06,
|
| 2647 |
+
"loss": 0.02,
|
| 2648 |
+
"step": 3530
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 21.3413715146948,
|
| 2652 |
+
"grad_norm": 1.6013274204687469,
|
| 2653 |
+
"learning_rate": 5e-06,
|
| 2654 |
+
"loss": 0.0201,
|
| 2655 |
+
"step": 3540
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 21.4016578749058,
|
| 2659 |
+
"grad_norm": 1.4800630752519932,
|
| 2660 |
+
"learning_rate": 5e-06,
|
| 2661 |
+
"loss": 0.02,
|
| 2662 |
+
"step": 3550
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 21.461944235116803,
|
| 2666 |
+
"grad_norm": 1.497381379010042,
|
| 2667 |
+
"learning_rate": 5e-06,
|
| 2668 |
+
"loss": 0.0203,
|
| 2669 |
+
"step": 3560
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 21.52223059532781,
|
| 2673 |
+
"grad_norm": 1.2261010629715117,
|
| 2674 |
+
"learning_rate": 5e-06,
|
| 2675 |
+
"loss": 0.0208,
|
| 2676 |
+
"step": 3570
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 21.58251695553881,
|
| 2680 |
+
"grad_norm": 1.353310824251755,
|
| 2681 |
+
"learning_rate": 5e-06,
|
| 2682 |
+
"loss": 0.0211,
|
| 2683 |
+
"step": 3580
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 21.642803315749813,
|
| 2687 |
+
"grad_norm": 1.2550244372657888,
|
| 2688 |
+
"learning_rate": 5e-06,
|
| 2689 |
+
"loss": 0.0217,
|
| 2690 |
+
"step": 3590
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 21.703089675960815,
|
| 2694 |
+
"grad_norm": 1.3870144386972025,
|
| 2695 |
+
"learning_rate": 5e-06,
|
| 2696 |
+
"loss": 0.0218,
|
| 2697 |
+
"step": 3600
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 21.763376036171817,
|
| 2701 |
+
"grad_norm": 1.2996006226877992,
|
| 2702 |
+
"learning_rate": 5e-06,
|
| 2703 |
+
"loss": 0.0217,
|
| 2704 |
+
"step": 3610
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 21.82366239638282,
|
| 2708 |
+
"grad_norm": 1.3533095992501387,
|
| 2709 |
+
"learning_rate": 5e-06,
|
| 2710 |
+
"loss": 0.0216,
|
| 2711 |
+
"step": 3620
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 21.88394875659382,
|
| 2715 |
+
"grad_norm": 1.3112078284304194,
|
| 2716 |
+
"learning_rate": 5e-06,
|
| 2717 |
+
"loss": 0.0217,
|
| 2718 |
+
"step": 3630
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 21.944235116804823,
|
| 2722 |
+
"grad_norm": 1.3288506152434172,
|
| 2723 |
+
"learning_rate": 5e-06,
|
| 2724 |
+
"loss": 0.0221,
|
| 2725 |
+
"step": 3640
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 21.998492840994725,
|
| 2729 |
+
"eval_loss": 2.351545572280884,
|
| 2730 |
+
"eval_runtime": 116.2788,
|
| 2731 |
+
"eval_samples_per_second": 38.433,
|
| 2732 |
+
"eval_steps_per_second": 0.602,
|
| 2733 |
+
"step": 3649
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 22.004521477015825,
|
| 2737 |
+
"grad_norm": 1.0649685278407737,
|
| 2738 |
+
"learning_rate": 5e-06,
|
| 2739 |
+
"loss": 0.0218,
|
| 2740 |
+
"step": 3650
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 22.064807837226827,
|
| 2744 |
+
"grad_norm": 1.2396804319512318,
|
| 2745 |
+
"learning_rate": 5e-06,
|
| 2746 |
+
"loss": 0.0158,
|
| 2747 |
+
"step": 3660
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 22.12509419743783,
|
| 2751 |
+
"grad_norm": 1.2203455145168154,
|
| 2752 |
+
"learning_rate": 5e-06,
|
| 2753 |
+
"loss": 0.0163,
|
| 2754 |
+
"step": 3670
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 22.18538055764883,
|
| 2758 |
+
"grad_norm": 1.1547196338255674,
|
| 2759 |
+
"learning_rate": 5e-06,
|
| 2760 |
+
"loss": 0.0169,
|
| 2761 |
+
"step": 3680
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 22.245666917859833,
|
| 2765 |
+
"grad_norm": 1.309239398813011,
|
| 2766 |
+
"learning_rate": 5e-06,
|
| 2767 |
+
"loss": 0.0173,
|
| 2768 |
+
"step": 3690
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 22.305953278070838,
|
| 2772 |
+
"grad_norm": 1.2683908516455489,
|
| 2773 |
+
"learning_rate": 5e-06,
|
| 2774 |
+
"loss": 0.0173,
|
| 2775 |
+
"step": 3700
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 22.36623963828184,
|
| 2779 |
+
"grad_norm": 1.5982780924133835,
|
| 2780 |
+
"learning_rate": 5e-06,
|
| 2781 |
+
"loss": 0.0183,
|
| 2782 |
+
"step": 3710
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 22.426525998492842,
|
| 2786 |
+
"grad_norm": 1.4555126747204115,
|
| 2787 |
+
"learning_rate": 5e-06,
|
| 2788 |
+
"loss": 0.0184,
|
| 2789 |
+
"step": 3720
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 22.486812358703844,
|
| 2793 |
+
"grad_norm": 1.3675517314579124,
|
| 2794 |
+
"learning_rate": 5e-06,
|
| 2795 |
+
"loss": 0.0186,
|
| 2796 |
+
"step": 3730
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 22.547098718914846,
|
| 2800 |
+
"grad_norm": 1.2088813887554757,
|
| 2801 |
+
"learning_rate": 5e-06,
|
| 2802 |
+
"loss": 0.0188,
|
| 2803 |
+
"step": 3740
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 22.607385079125848,
|
| 2807 |
+
"grad_norm": 1.157896013440299,
|
| 2808 |
+
"learning_rate": 5e-06,
|
| 2809 |
+
"loss": 0.0194,
|
| 2810 |
+
"step": 3750
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 22.66767143933685,
|
| 2814 |
+
"grad_norm": 1.4152388376390785,
|
| 2815 |
+
"learning_rate": 5e-06,
|
| 2816 |
+
"loss": 0.0189,
|
| 2817 |
+
"step": 3760
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 22.727957799547852,
|
| 2821 |
+
"grad_norm": 1.2554294042934306,
|
| 2822 |
+
"learning_rate": 5e-06,
|
| 2823 |
+
"loss": 0.019,
|
| 2824 |
+
"step": 3770
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 22.788244159758854,
|
| 2828 |
+
"grad_norm": 1.3329931616963409,
|
| 2829 |
+
"learning_rate": 5e-06,
|
| 2830 |
+
"loss": 0.019,
|
| 2831 |
+
"step": 3780
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 22.848530519969856,
|
| 2835 |
+
"grad_norm": 1.256565715756104,
|
| 2836 |
+
"learning_rate": 5e-06,
|
| 2837 |
+
"loss": 0.0189,
|
| 2838 |
+
"step": 3790
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 22.908816880180858,
|
| 2842 |
+
"grad_norm": 1.368011051794652,
|
| 2843 |
+
"learning_rate": 5e-06,
|
| 2844 |
+
"loss": 0.0195,
|
| 2845 |
+
"step": 3800
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 22.96910324039186,
|
| 2849 |
+
"grad_norm": 1.4311223887923525,
|
| 2850 |
+
"learning_rate": 5e-06,
|
| 2851 |
+
"loss": 0.0197,
|
| 2852 |
+
"step": 3810
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 22.99924642049736,
|
| 2856 |
+
"eval_loss": 2.359902858734131,
|
| 2857 |
+
"eval_runtime": 117.0325,
|
| 2858 |
+
"eval_samples_per_second": 38.186,
|
| 2859 |
+
"eval_steps_per_second": 0.598,
|
| 2860 |
+
"step": 3815
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 23.029389600602865,
|
| 2864 |
+
"grad_norm": 1.114496963570823,
|
| 2865 |
+
"learning_rate": 5e-06,
|
| 2866 |
+
"loss": 0.0171,
|
| 2867 |
+
"step": 3820
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 23.089675960813867,
|
| 2871 |
+
"grad_norm": 1.1656429989984474,
|
| 2872 |
+
"learning_rate": 5e-06,
|
| 2873 |
+
"loss": 0.0149,
|
| 2874 |
+
"step": 3830
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 23.14996232102487,
|
| 2878 |
+
"grad_norm": 1.3181375592191023,
|
| 2879 |
+
"learning_rate": 5e-06,
|
| 2880 |
+
"loss": 0.0152,
|
| 2881 |
+
"step": 3840
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 23.21024868123587,
|
| 2885 |
+
"grad_norm": 1.214891422163979,
|
| 2886 |
+
"learning_rate": 5e-06,
|
| 2887 |
+
"loss": 0.0157,
|
| 2888 |
+
"step": 3850
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 23.270535041446873,
|
| 2892 |
+
"grad_norm": 1.1329477739782727,
|
| 2893 |
+
"learning_rate": 5e-06,
|
| 2894 |
+
"loss": 0.0159,
|
| 2895 |
+
"step": 3860
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 23.330821401657875,
|
| 2899 |
+
"grad_norm": 1.178978239439168,
|
| 2900 |
+
"learning_rate": 5e-06,
|
| 2901 |
+
"loss": 0.0163,
|
| 2902 |
+
"step": 3870
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 23.391107761868877,
|
| 2906 |
+
"grad_norm": 1.1072042006237766,
|
| 2907 |
+
"learning_rate": 5e-06,
|
| 2908 |
+
"loss": 0.0165,
|
| 2909 |
+
"step": 3880
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 23.45139412207988,
|
| 2913 |
+
"grad_norm": 1.2281267859101168,
|
| 2914 |
+
"learning_rate": 5e-06,
|
| 2915 |
+
"loss": 0.0168,
|
| 2916 |
+
"step": 3890
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 23.51168048229088,
|
| 2920 |
+
"grad_norm": 1.1558899107856422,
|
| 2921 |
+
"learning_rate": 5e-06,
|
| 2922 |
+
"loss": 0.0174,
|
| 2923 |
+
"step": 3900
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 23.571966842501883,
|
| 2927 |
+
"grad_norm": 1.217612194947995,
|
| 2928 |
+
"learning_rate": 5e-06,
|
| 2929 |
+
"loss": 0.0173,
|
| 2930 |
+
"step": 3910
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 23.632253202712885,
|
| 2934 |
+
"grad_norm": 1.2440180440988633,
|
| 2935 |
+
"learning_rate": 5e-06,
|
| 2936 |
+
"loss": 0.0174,
|
| 2937 |
+
"step": 3920
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 23.692539562923887,
|
| 2941 |
+
"grad_norm": 1.3420865819465875,
|
| 2942 |
+
"learning_rate": 5e-06,
|
| 2943 |
+
"loss": 0.0174,
|
| 2944 |
+
"step": 3930
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 23.75282592313489,
|
| 2948 |
+
"grad_norm": 1.3593190067484822,
|
| 2949 |
+
"learning_rate": 5e-06,
|
| 2950 |
+
"loss": 0.018,
|
| 2951 |
+
"step": 3940
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 23.813112283345895,
|
| 2955 |
+
"grad_norm": 1.2367975594859635,
|
| 2956 |
+
"learning_rate": 5e-06,
|
| 2957 |
+
"loss": 0.0177,
|
| 2958 |
+
"step": 3950
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 23.873398643556897,
|
| 2962 |
+
"grad_norm": 1.2214254875430963,
|
| 2963 |
+
"learning_rate": 5e-06,
|
| 2964 |
+
"loss": 0.0179,
|
| 2965 |
+
"step": 3960
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 23.9336850037679,
|
| 2969 |
+
"grad_norm": 1.20689456847013,
|
| 2970 |
+
"learning_rate": 5e-06,
|
| 2971 |
+
"loss": 0.018,
|
| 2972 |
+
"step": 3970
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 23.9939713639789,
|
| 2976 |
+
"grad_norm": 1.2666923485205688,
|
| 2977 |
+
"learning_rate": 5e-06,
|
| 2978 |
+
"loss": 0.0186,
|
| 2979 |
+
"step": 3980
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 24.0,
|
| 2983 |
+
"eval_loss": 2.3888232707977295,
|
| 2984 |
+
"eval_runtime": 116.343,
|
| 2985 |
+
"eval_samples_per_second": 38.412,
|
| 2986 |
+
"eval_steps_per_second": 0.602,
|
| 2987 |
+
"step": 3981
|
| 2988 |
+
},
|
| 2989 |
+
{
|
| 2990 |
+
"epoch": 24.054257724189902,
|
| 2991 |
+
"grad_norm": 1.1066074003233175,
|
| 2992 |
+
"learning_rate": 5e-06,
|
| 2993 |
+
"loss": 0.0141,
|
| 2994 |
+
"step": 3990
|
| 2995 |
+
},
|
| 2996 |
+
{
|
| 2997 |
+
"epoch": 24.114544084400904,
|
| 2998 |
+
"grad_norm": 1.0389652033545478,
|
| 2999 |
+
"learning_rate": 5e-06,
|
| 3000 |
+
"loss": 0.0141,
|
| 3001 |
+
"step": 4000
|
| 3002 |
+
},
|
| 3003 |
+
{
|
| 3004 |
+
"epoch": 24.174830444611906,
|
| 3005 |
+
"grad_norm": 1.1258379344398495,
|
| 3006 |
+
"learning_rate": 5e-06,
|
| 3007 |
+
"loss": 0.0144,
|
| 3008 |
+
"step": 4010
|
| 3009 |
+
},
|
| 3010 |
+
{
|
| 3011 |
+
"epoch": 24.23511680482291,
|
| 3012 |
+
"grad_norm": 1.1758249329629795,
|
| 3013 |
+
"learning_rate": 5e-06,
|
| 3014 |
+
"loss": 0.0148,
|
| 3015 |
+
"step": 4020
|
| 3016 |
+
},
|
| 3017 |
+
{
|
| 3018 |
+
"epoch": 24.29540316503391,
|
| 3019 |
+
"grad_norm": 1.1741885490501298,
|
| 3020 |
+
"learning_rate": 5e-06,
|
| 3021 |
+
"loss": 0.0151,
|
| 3022 |
+
"step": 4030
|
| 3023 |
+
},
|
| 3024 |
+
{
|
| 3025 |
+
"epoch": 24.355689525244912,
|
| 3026 |
+
"grad_norm": 1.1622125971160786,
|
| 3027 |
+
"learning_rate": 5e-06,
|
| 3028 |
+
"loss": 0.0156,
|
| 3029 |
+
"step": 4040
|
| 3030 |
+
},
|
| 3031 |
+
{
|
| 3032 |
+
"epoch": 24.415975885455914,
|
| 3033 |
+
"grad_norm": 1.0899820074898727,
|
| 3034 |
+
"learning_rate": 5e-06,
|
| 3035 |
+
"loss": 0.0154,
|
| 3036 |
+
"step": 4050
|
| 3037 |
+
},
|
| 3038 |
+
{
|
| 3039 |
+
"epoch": 24.476262245666916,
|
| 3040 |
+
"grad_norm": 1.2599823712922462,
|
| 3041 |
+
"learning_rate": 5e-06,
|
| 3042 |
+
"loss": 0.0159,
|
| 3043 |
+
"step": 4060
|
| 3044 |
+
},
|
| 3045 |
+
{
|
| 3046 |
+
"epoch": 24.53654860587792,
|
| 3047 |
+
"grad_norm": 1.140178001121795,
|
| 3048 |
+
"learning_rate": 5e-06,
|
| 3049 |
+
"loss": 0.0157,
|
| 3050 |
+
"step": 4070
|
| 3051 |
+
},
|
| 3052 |
+
{
|
| 3053 |
+
"epoch": 24.596834966088924,
|
| 3054 |
+
"grad_norm": 1.1230288891139422,
|
| 3055 |
+
"learning_rate": 5e-06,
|
| 3056 |
+
"loss": 0.0158,
|
| 3057 |
+
"step": 4080
|
| 3058 |
+
},
|
| 3059 |
+
{
|
| 3060 |
+
"epoch": 24.657121326299926,
|
| 3061 |
+
"grad_norm": 1.1218784125193524,
|
| 3062 |
+
"learning_rate": 5e-06,
|
| 3063 |
+
"loss": 0.0163,
|
| 3064 |
+
"step": 4090
|
| 3065 |
+
},
|
| 3066 |
+
{
|
| 3067 |
+
"epoch": 24.717407686510928,
|
| 3068 |
+
"grad_norm": 1.1524016261830858,
|
| 3069 |
+
"learning_rate": 5e-06,
|
| 3070 |
+
"loss": 0.0164,
|
| 3071 |
+
"step": 4100
|
| 3072 |
+
},
|
| 3073 |
+
{
|
| 3074 |
+
"epoch": 24.77769404672193,
|
| 3075 |
+
"grad_norm": 1.15414666614049,
|
| 3076 |
+
"learning_rate": 5e-06,
|
| 3077 |
+
"loss": 0.0166,
|
| 3078 |
+
"step": 4110
|
| 3079 |
+
},
|
| 3080 |
+
{
|
| 3081 |
+
"epoch": 24.83798040693293,
|
| 3082 |
+
"grad_norm": 1.3244431296269448,
|
| 3083 |
+
"learning_rate": 5e-06,
|
| 3084 |
+
"loss": 0.0169,
|
| 3085 |
+
"step": 4120
|
| 3086 |
+
},
|
| 3087 |
+
{
|
| 3088 |
+
"epoch": 24.868123587038433,
|
| 3089 |
+
"eval_loss": 2.4057514667510986,
|
| 3090 |
+
"eval_runtime": 113.5854,
|
| 3091 |
+
"eval_samples_per_second": 39.345,
|
| 3092 |
+
"eval_steps_per_second": 0.616,
|
| 3093 |
+
"step": 4125
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 24.868123587038433,
|
| 3097 |
+
"step": 4125,
|
| 3098 |
+
"total_flos": 6909319257784320.0,
|
| 3099 |
+
"train_loss": 0.2661740447047985,
|
| 3100 |
+
"train_runtime": 139979.5045,
|
| 3101 |
+
"train_samples_per_second": 15.162,
|
| 3102 |
+
"train_steps_per_second": 0.029
|
| 3103 |
+
}
|
| 3104 |
+
],
|
| 3105 |
+
"logging_steps": 10,
|
| 3106 |
+
"max_steps": 4125,
|
| 3107 |
+
"num_input_tokens_seen": 0,
|
| 3108 |
+
"num_train_epochs": 25,
|
| 3109 |
+
"save_steps": 500,
|
| 3110 |
+
"stateful_callbacks": {
|
| 3111 |
+
"TrainerControl": {
|
| 3112 |
+
"args": {
|
| 3113 |
+
"should_epoch_stop": false,
|
| 3114 |
+
"should_evaluate": false,
|
| 3115 |
+
"should_log": false,
|
| 3116 |
+
"should_save": true,
|
| 3117 |
+
"should_training_stop": true
|
| 3118 |
+
},
|
| 3119 |
+
"attributes": {}
|
| 3120 |
+
}
|
| 3121 |
+
},
|
| 3122 |
+
"total_flos": 6909319257784320.0,
|
| 3123 |
+
"train_batch_size": 8,
|
| 3124 |
+
"trial_name": null,
|
| 3125 |
+
"trial_params": null
|
| 3126 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|