End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +3865 -0
- training_loss.png +0 -0
README.md
CHANGED
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
4 |
base_model: secmlr/final_model
|
5 |
tags:
|
6 |
- llama-factory
|
|
|
7 |
- generated_from_trainer
|
8 |
model-index:
|
9 |
- name: best_n_no_rationale_poc_agent_withjava_final_model_agent
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# best_n_no_rationale_poc_agent_withjava_final_model_agent
|
17 |
|
18 |
-
This model is a fine-tuned version of [secmlr/final_model](https://huggingface.co/secmlr/final_model) on
|
19 |
|
20 |
## Model description
|
21 |
|
|
|
4 |
base_model: secmlr/final_model
|
5 |
tags:
|
6 |
- llama-factory
|
7 |
+
- full
|
8 |
- generated_from_trainer
|
9 |
model-index:
|
10 |
- name: best_n_no_rationale_poc_agent_withjava_final_model_agent
|
|
|
16 |
|
17 |
# best_n_no_rationale_poc_agent_withjava_final_model_agent
|
18 |
|
19 |
+
This model is a fine-tuned version of [secmlr/final_model](https://huggingface.co/secmlr/final_model) on the best_n_no_rationale_poc_agent_withjava dataset.
|
20 |
|
21 |
## Model description
|
22 |
|
all_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9849521203830367,
|
3 |
+
"total_flos": 4.3229407025404314e+17,
|
4 |
+
"train_loss": 0.7402517178774753,
|
5 |
+
"train_runtime": 59397.3306,
|
6 |
+
"train_samples_per_second": 0.443,
|
7 |
+
"train_steps_per_second": 0.009
|
8 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9849521203830367,
|
3 |
+
"total_flos": 4.3229407025404314e+17,
|
4 |
+
"train_loss": 0.7402517178774753,
|
5 |
+
"train_runtime": 59397.3306,
|
6 |
+
"train_samples_per_second": 0.443,
|
7 |
+
"train_steps_per_second": 0.009
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3865 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 2.9849521203830367,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 546,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.005471956224350205,
|
14 |
+
"grad_norm": 10.35721852671886,
|
15 |
+
"learning_rate": 0.0,
|
16 |
+
"loss": 1.8883,
|
17 |
+
"step": 1
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.01094391244870041,
|
21 |
+
"grad_norm": 9.153579967775789,
|
22 |
+
"learning_rate": 1.8181818181818183e-07,
|
23 |
+
"loss": 1.6762,
|
24 |
+
"step": 2
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.016415868673050615,
|
28 |
+
"grad_norm": 8.595961036647338,
|
29 |
+
"learning_rate": 3.6363636363636366e-07,
|
30 |
+
"loss": 1.7134,
|
31 |
+
"step": 3
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.02188782489740082,
|
35 |
+
"grad_norm": 9.767137847671998,
|
36 |
+
"learning_rate": 5.454545454545455e-07,
|
37 |
+
"loss": 1.8695,
|
38 |
+
"step": 4
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.027359781121751026,
|
42 |
+
"grad_norm": 9.587445150060104,
|
43 |
+
"learning_rate": 7.272727272727273e-07,
|
44 |
+
"loss": 1.8403,
|
45 |
+
"step": 5
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.03283173734610123,
|
49 |
+
"grad_norm": 9.177008384918361,
|
50 |
+
"learning_rate": 9.090909090909091e-07,
|
51 |
+
"loss": 1.8047,
|
52 |
+
"step": 6
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.038303693570451436,
|
56 |
+
"grad_norm": 9.735262587335374,
|
57 |
+
"learning_rate": 1.090909090909091e-06,
|
58 |
+
"loss": 1.8478,
|
59 |
+
"step": 7
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.04377564979480164,
|
63 |
+
"grad_norm": 9.947195530164768,
|
64 |
+
"learning_rate": 1.2727272727272728e-06,
|
65 |
+
"loss": 1.9031,
|
66 |
+
"step": 8
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.049247606019151846,
|
70 |
+
"grad_norm": 8.893018941795928,
|
71 |
+
"learning_rate": 1.4545454545454546e-06,
|
72 |
+
"loss": 1.8532,
|
73 |
+
"step": 9
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.05471956224350205,
|
77 |
+
"grad_norm": 7.7444186622176066,
|
78 |
+
"learning_rate": 1.6363636363636365e-06,
|
79 |
+
"loss": 1.7478,
|
80 |
+
"step": 10
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.060191518467852256,
|
84 |
+
"grad_norm": 6.797207399484385,
|
85 |
+
"learning_rate": 1.8181818181818183e-06,
|
86 |
+
"loss": 1.8029,
|
87 |
+
"step": 11
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.06566347469220246,
|
91 |
+
"grad_norm": 6.272817111572408,
|
92 |
+
"learning_rate": 2.0000000000000003e-06,
|
93 |
+
"loss": 1.5675,
|
94 |
+
"step": 12
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.07113543091655267,
|
98 |
+
"grad_norm": 5.803486984123274,
|
99 |
+
"learning_rate": 2.181818181818182e-06,
|
100 |
+
"loss": 1.5805,
|
101 |
+
"step": 13
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.07660738714090287,
|
105 |
+
"grad_norm": 4.929571426424171,
|
106 |
+
"learning_rate": 2.363636363636364e-06,
|
107 |
+
"loss": 1.4595,
|
108 |
+
"step": 14
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.08207934336525308,
|
112 |
+
"grad_norm": 4.180209755084977,
|
113 |
+
"learning_rate": 2.5454545454545456e-06,
|
114 |
+
"loss": 1.5726,
|
115 |
+
"step": 15
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.08755129958960328,
|
119 |
+
"grad_norm": 3.6628700301414185,
|
120 |
+
"learning_rate": 2.7272727272727272e-06,
|
121 |
+
"loss": 1.4351,
|
122 |
+
"step": 16
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.09302325581395349,
|
126 |
+
"grad_norm": 3.308235213719836,
|
127 |
+
"learning_rate": 2.9090909090909093e-06,
|
128 |
+
"loss": 1.3786,
|
129 |
+
"step": 17
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.09849521203830369,
|
133 |
+
"grad_norm": 3.5573667308014936,
|
134 |
+
"learning_rate": 3.090909090909091e-06,
|
135 |
+
"loss": 1.4688,
|
136 |
+
"step": 18
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.1039671682626539,
|
140 |
+
"grad_norm": 2.909229460614563,
|
141 |
+
"learning_rate": 3.272727272727273e-06,
|
142 |
+
"loss": 1.3943,
|
143 |
+
"step": 19
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.1094391244870041,
|
147 |
+
"grad_norm": 2.7422131458495964,
|
148 |
+
"learning_rate": 3.454545454545455e-06,
|
149 |
+
"loss": 1.3837,
|
150 |
+
"step": 20
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.11491108071135431,
|
154 |
+
"grad_norm": 2.6686714630813446,
|
155 |
+
"learning_rate": 3.6363636363636366e-06,
|
156 |
+
"loss": 1.2383,
|
157 |
+
"step": 21
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.12038303693570451,
|
161 |
+
"grad_norm": 2.537185931531279,
|
162 |
+
"learning_rate": 3.818181818181819e-06,
|
163 |
+
"loss": 1.2007,
|
164 |
+
"step": 22
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.12585499316005472,
|
168 |
+
"grad_norm": 2.8025222991868093,
|
169 |
+
"learning_rate": 4.000000000000001e-06,
|
170 |
+
"loss": 1.3034,
|
171 |
+
"step": 23
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.13132694938440492,
|
175 |
+
"grad_norm": 2.702541995531342,
|
176 |
+
"learning_rate": 4.181818181818182e-06,
|
177 |
+
"loss": 1.2581,
|
178 |
+
"step": 24
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.13679890560875513,
|
182 |
+
"grad_norm": 2.7630783359767417,
|
183 |
+
"learning_rate": 4.363636363636364e-06,
|
184 |
+
"loss": 1.2718,
|
185 |
+
"step": 25
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.14227086183310533,
|
189 |
+
"grad_norm": 2.4995926175272,
|
190 |
+
"learning_rate": 4.5454545454545455e-06,
|
191 |
+
"loss": 1.2783,
|
192 |
+
"step": 26
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.14774281805745554,
|
196 |
+
"grad_norm": 2.04117236736765,
|
197 |
+
"learning_rate": 4.727272727272728e-06,
|
198 |
+
"loss": 1.2178,
|
199 |
+
"step": 27
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.15321477428180574,
|
203 |
+
"grad_norm": 2.0163615923695084,
|
204 |
+
"learning_rate": 4.90909090909091e-06,
|
205 |
+
"loss": 1.2405,
|
206 |
+
"step": 28
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.15868673050615595,
|
210 |
+
"grad_norm": 1.884651209109897,
|
211 |
+
"learning_rate": 5.090909090909091e-06,
|
212 |
+
"loss": 1.1669,
|
213 |
+
"step": 29
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.16415868673050615,
|
217 |
+
"grad_norm": 1.7458906060669555,
|
218 |
+
"learning_rate": 5.272727272727273e-06,
|
219 |
+
"loss": 1.1599,
|
220 |
+
"step": 30
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.16963064295485636,
|
224 |
+
"grad_norm": 1.8792292725188067,
|
225 |
+
"learning_rate": 5.4545454545454545e-06,
|
226 |
+
"loss": 1.2979,
|
227 |
+
"step": 31
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.17510259917920656,
|
231 |
+
"grad_norm": 1.758309878878688,
|
232 |
+
"learning_rate": 5.636363636363636e-06,
|
233 |
+
"loss": 1.1704,
|
234 |
+
"step": 32
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.18057455540355677,
|
238 |
+
"grad_norm": 1.8180147632994625,
|
239 |
+
"learning_rate": 5.8181818181818185e-06,
|
240 |
+
"loss": 1.2306,
|
241 |
+
"step": 33
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.18604651162790697,
|
245 |
+
"grad_norm": 1.6677866975645874,
|
246 |
+
"learning_rate": 6e-06,
|
247 |
+
"loss": 1.1128,
|
248 |
+
"step": 34
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.19151846785225718,
|
252 |
+
"grad_norm": 1.6378208683813074,
|
253 |
+
"learning_rate": 6.181818181818182e-06,
|
254 |
+
"loss": 1.1292,
|
255 |
+
"step": 35
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.19699042407660738,
|
259 |
+
"grad_norm": 1.6748036142282632,
|
260 |
+
"learning_rate": 6.363636363636364e-06,
|
261 |
+
"loss": 1.1514,
|
262 |
+
"step": 36
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.2024623803009576,
|
266 |
+
"grad_norm": 1.5719011117183284,
|
267 |
+
"learning_rate": 6.545454545454546e-06,
|
268 |
+
"loss": 1.0348,
|
269 |
+
"step": 37
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.2079343365253078,
|
273 |
+
"grad_norm": 1.705586999155045,
|
274 |
+
"learning_rate": 6.7272727272727275e-06,
|
275 |
+
"loss": 1.0895,
|
276 |
+
"step": 38
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.213406292749658,
|
280 |
+
"grad_norm": 1.501654311195198,
|
281 |
+
"learning_rate": 6.90909090909091e-06,
|
282 |
+
"loss": 1.0314,
|
283 |
+
"step": 39
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.2188782489740082,
|
287 |
+
"grad_norm": 1.5318447039140235,
|
288 |
+
"learning_rate": 7.0909090909090916e-06,
|
289 |
+
"loss": 1.0393,
|
290 |
+
"step": 40
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.2243502051983584,
|
294 |
+
"grad_norm": 1.5330990400515416,
|
295 |
+
"learning_rate": 7.272727272727273e-06,
|
296 |
+
"loss": 1.0725,
|
297 |
+
"step": 41
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.22982216142270862,
|
301 |
+
"grad_norm": 1.5747637040267966,
|
302 |
+
"learning_rate": 7.454545454545456e-06,
|
303 |
+
"loss": 1.1064,
|
304 |
+
"step": 42
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.23529411764705882,
|
308 |
+
"grad_norm": 1.6579634988946377,
|
309 |
+
"learning_rate": 7.636363636363638e-06,
|
310 |
+
"loss": 1.1167,
|
311 |
+
"step": 43
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.24076607387140903,
|
315 |
+
"grad_norm": 1.4573180490811242,
|
316 |
+
"learning_rate": 7.81818181818182e-06,
|
317 |
+
"loss": 1.0057,
|
318 |
+
"step": 44
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.24623803009575923,
|
322 |
+
"grad_norm": 1.5087523436369001,
|
323 |
+
"learning_rate": 8.000000000000001e-06,
|
324 |
+
"loss": 1.0419,
|
325 |
+
"step": 45
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.25170998632010944,
|
329 |
+
"grad_norm": 1.4184742171640738,
|
330 |
+
"learning_rate": 8.181818181818183e-06,
|
331 |
+
"loss": 1.0261,
|
332 |
+
"step": 46
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.25718194254445964,
|
336 |
+
"grad_norm": 1.4793061661317761,
|
337 |
+
"learning_rate": 8.363636363636365e-06,
|
338 |
+
"loss": 1.0732,
|
339 |
+
"step": 47
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.26265389876880985,
|
343 |
+
"grad_norm": 1.4483388558986123,
|
344 |
+
"learning_rate": 8.545454545454546e-06,
|
345 |
+
"loss": 1.0268,
|
346 |
+
"step": 48
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 0.26812585499316005,
|
350 |
+
"grad_norm": 1.3959861971253678,
|
351 |
+
"learning_rate": 8.727272727272728e-06,
|
352 |
+
"loss": 0.9602,
|
353 |
+
"step": 49
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 0.27359781121751026,
|
357 |
+
"grad_norm": 1.4654503788815159,
|
358 |
+
"learning_rate": 8.90909090909091e-06,
|
359 |
+
"loss": 1.0214,
|
360 |
+
"step": 50
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.27906976744186046,
|
364 |
+
"grad_norm": 1.3375520902585056,
|
365 |
+
"learning_rate": 9.090909090909091e-06,
|
366 |
+
"loss": 0.9672,
|
367 |
+
"step": 51
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 0.28454172366621067,
|
371 |
+
"grad_norm": 1.4572373659872022,
|
372 |
+
"learning_rate": 9.272727272727273e-06,
|
373 |
+
"loss": 1.0336,
|
374 |
+
"step": 52
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 0.29001367989056087,
|
378 |
+
"grad_norm": 1.3309853940134408,
|
379 |
+
"learning_rate": 9.454545454545456e-06,
|
380 |
+
"loss": 0.921,
|
381 |
+
"step": 53
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.2954856361149111,
|
385 |
+
"grad_norm": 1.3696989025706146,
|
386 |
+
"learning_rate": 9.636363636363638e-06,
|
387 |
+
"loss": 0.9601,
|
388 |
+
"step": 54
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 0.3009575923392613,
|
392 |
+
"grad_norm": 1.5697276495556032,
|
393 |
+
"learning_rate": 9.81818181818182e-06,
|
394 |
+
"loss": 1.0595,
|
395 |
+
"step": 55
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 0.3064295485636115,
|
399 |
+
"grad_norm": 1.3272312418787842,
|
400 |
+
"learning_rate": 1e-05,
|
401 |
+
"loss": 0.9707,
|
402 |
+
"step": 56
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.3119015047879617,
|
406 |
+
"grad_norm": 1.4251214405189137,
|
407 |
+
"learning_rate": 9.999897652959688e-06,
|
408 |
+
"loss": 1.0395,
|
409 |
+
"step": 57
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 0.3173734610123119,
|
413 |
+
"grad_norm": 1.5692127414245156,
|
414 |
+
"learning_rate": 9.999590616028718e-06,
|
415 |
+
"loss": 0.9859,
|
416 |
+
"step": 58
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 0.3228454172366621,
|
420 |
+
"grad_norm": 1.3732222598153316,
|
421 |
+
"learning_rate": 9.999078901776819e-06,
|
422 |
+
"loss": 0.8967,
|
423 |
+
"step": 59
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.3283173734610123,
|
427 |
+
"grad_norm": 1.2388204623597197,
|
428 |
+
"learning_rate": 9.998362531152967e-06,
|
429 |
+
"loss": 0.8705,
|
430 |
+
"step": 60
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 0.3337893296853625,
|
434 |
+
"grad_norm": 1.3101231933303101,
|
435 |
+
"learning_rate": 9.997441533484527e-06,
|
436 |
+
"loss": 0.9106,
|
437 |
+
"step": 61
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 0.3392612859097127,
|
441 |
+
"grad_norm": 1.4133686606978433,
|
442 |
+
"learning_rate": 9.996315946476054e-06,
|
443 |
+
"loss": 0.9483,
|
444 |
+
"step": 62
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.3447332421340629,
|
448 |
+
"grad_norm": 1.4050099822399718,
|
449 |
+
"learning_rate": 9.994985816207745e-06,
|
450 |
+
"loss": 0.9165,
|
451 |
+
"step": 63
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 0.35020519835841313,
|
455 |
+
"grad_norm": 1.2811009758457281,
|
456 |
+
"learning_rate": 9.993451197133558e-06,
|
457 |
+
"loss": 0.9197,
|
458 |
+
"step": 64
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 0.35567715458276333,
|
462 |
+
"grad_norm": 1.3918961064264443,
|
463 |
+
"learning_rate": 9.991712152078988e-06,
|
464 |
+
"loss": 0.9584,
|
465 |
+
"step": 65
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.36114911080711354,
|
469 |
+
"grad_norm": 1.278041046459637,
|
470 |
+
"learning_rate": 9.989768752238473e-06,
|
471 |
+
"loss": 0.8481,
|
472 |
+
"step": 66
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 0.36662106703146374,
|
476 |
+
"grad_norm": 1.4164659056438145,
|
477 |
+
"learning_rate": 9.987621077172504e-06,
|
478 |
+
"loss": 0.9649,
|
479 |
+
"step": 67
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 0.37209302325581395,
|
483 |
+
"grad_norm": 1.4262605111640658,
|
484 |
+
"learning_rate": 9.985269214804358e-06,
|
485 |
+
"loss": 1.0005,
|
486 |
+
"step": 68
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 0.37756497948016415,
|
490 |
+
"grad_norm": 1.3473955130378172,
|
491 |
+
"learning_rate": 9.982713261416493e-06,
|
492 |
+
"loss": 0.9358,
|
493 |
+
"step": 69
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 0.38303693570451436,
|
497 |
+
"grad_norm": 1.309971844519936,
|
498 |
+
"learning_rate": 9.979953321646615e-06,
|
499 |
+
"loss": 0.9322,
|
500 |
+
"step": 70
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 0.38850889192886456,
|
504 |
+
"grad_norm": 1.2854972333406363,
|
505 |
+
"learning_rate": 9.976989508483395e-06,
|
506 |
+
"loss": 0.8887,
|
507 |
+
"step": 71
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.39398084815321477,
|
511 |
+
"grad_norm": 1.2580467092774625,
|
512 |
+
"learning_rate": 9.97382194326183e-06,
|
513 |
+
"loss": 0.8773,
|
514 |
+
"step": 72
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 0.399452804377565,
|
518 |
+
"grad_norm": 1.375744374898363,
|
519 |
+
"learning_rate": 9.970450755658293e-06,
|
520 |
+
"loss": 0.9411,
|
521 |
+
"step": 73
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 0.4049247606019152,
|
525 |
+
"grad_norm": 1.3308096424393216,
|
526 |
+
"learning_rate": 9.96687608368521e-06,
|
527 |
+
"loss": 0.9381,
|
528 |
+
"step": 74
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 0.4103967168262654,
|
532 |
+
"grad_norm": 1.3484391473936597,
|
533 |
+
"learning_rate": 9.963098073685425e-06,
|
534 |
+
"loss": 0.9509,
|
535 |
+
"step": 75
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 0.4158686730506156,
|
539 |
+
"grad_norm": 1.4391584068557033,
|
540 |
+
"learning_rate": 9.95911688032619e-06,
|
541 |
+
"loss": 0.9841,
|
542 |
+
"step": 76
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 0.4213406292749658,
|
546 |
+
"grad_norm": 1.4045023186509025,
|
547 |
+
"learning_rate": 9.954932666592852e-06,
|
548 |
+
"loss": 0.9244,
|
549 |
+
"step": 77
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.426812585499316,
|
553 |
+
"grad_norm": 1.1699025954332016,
|
554 |
+
"learning_rate": 9.950545603782163e-06,
|
555 |
+
"loss": 0.8224,
|
556 |
+
"step": 78
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 0.4322845417236662,
|
560 |
+
"grad_norm": 1.4395940015493598,
|
561 |
+
"learning_rate": 9.945955871495284e-06,
|
562 |
+
"loss": 0.9675,
|
563 |
+
"step": 79
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 0.4377564979480164,
|
567 |
+
"grad_norm": 1.394141429502428,
|
568 |
+
"learning_rate": 9.94116365763042e-06,
|
569 |
+
"loss": 0.931,
|
570 |
+
"step": 80
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 0.4432284541723666,
|
574 |
+
"grad_norm": 1.2867524014600322,
|
575 |
+
"learning_rate": 9.936169158375132e-06,
|
576 |
+
"loss": 0.9014,
|
577 |
+
"step": 81
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 0.4487004103967168,
|
581 |
+
"grad_norm": 1.4449589187576641,
|
582 |
+
"learning_rate": 9.93097257819831e-06,
|
583 |
+
"loss": 1.0038,
|
584 |
+
"step": 82
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 0.454172366621067,
|
588 |
+
"grad_norm": 1.3699175592657065,
|
589 |
+
"learning_rate": 9.925574129841793e-06,
|
590 |
+
"loss": 0.9125,
|
591 |
+
"step": 83
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.45964432284541723,
|
595 |
+
"grad_norm": 1.4034783366132408,
|
596 |
+
"learning_rate": 9.919974034311665e-06,
|
597 |
+
"loss": 1.1419,
|
598 |
+
"step": 84
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 0.46511627906976744,
|
602 |
+
"grad_norm": 1.3583019034444384,
|
603 |
+
"learning_rate": 9.914172520869207e-06,
|
604 |
+
"loss": 0.9291,
|
605 |
+
"step": 85
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 0.47058823529411764,
|
609 |
+
"grad_norm": 1.2923333604930634,
|
610 |
+
"learning_rate": 9.90816982702151e-06,
|
611 |
+
"loss": 1.0433,
|
612 |
+
"step": 86
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 0.47606019151846785,
|
616 |
+
"grad_norm": 1.3321288080713247,
|
617 |
+
"learning_rate": 9.901966198511756e-06,
|
618 |
+
"loss": 0.8418,
|
619 |
+
"step": 87
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 0.48153214774281805,
|
623 |
+
"grad_norm": 1.351154257007612,
|
624 |
+
"learning_rate": 9.895561889309152e-06,
|
625 |
+
"loss": 0.8594,
|
626 |
+
"step": 88
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 0.48700410396716826,
|
630 |
+
"grad_norm": 1.2894835310975192,
|
631 |
+
"learning_rate": 9.888957161598532e-06,
|
632 |
+
"loss": 0.8793,
|
633 |
+
"step": 89
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.49247606019151846,
|
637 |
+
"grad_norm": 1.2666013479461309,
|
638 |
+
"learning_rate": 9.882152285769633e-06,
|
639 |
+
"loss": 0.8542,
|
640 |
+
"step": 90
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 0.49794801641586867,
|
644 |
+
"grad_norm": 1.3295038462724125,
|
645 |
+
"learning_rate": 9.875147540406012e-06,
|
646 |
+
"loss": 0.877,
|
647 |
+
"step": 91
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 0.5034199726402189,
|
651 |
+
"grad_norm": 1.4321066386634147,
|
652 |
+
"learning_rate": 9.867943212273653e-06,
|
653 |
+
"loss": 0.9977,
|
654 |
+
"step": 92
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 0.5088919288645691,
|
658 |
+
"grad_norm": 1.4216470891810016,
|
659 |
+
"learning_rate": 9.86053959630922e-06,
|
660 |
+
"loss": 0.9423,
|
661 |
+
"step": 93
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 0.5143638850889193,
|
665 |
+
"grad_norm": 1.2905384460560352,
|
666 |
+
"learning_rate": 9.852936995607986e-06,
|
667 |
+
"loss": 0.8744,
|
668 |
+
"step": 94
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 0.5198358413132695,
|
672 |
+
"grad_norm": 1.35646383730229,
|
673 |
+
"learning_rate": 9.845135721411425e-06,
|
674 |
+
"loss": 1.0106,
|
675 |
+
"step": 95
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 0.5253077975376197,
|
679 |
+
"grad_norm": 1.3119338491945538,
|
680 |
+
"learning_rate": 9.837136093094465e-06,
|
681 |
+
"loss": 0.9134,
|
682 |
+
"step": 96
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 0.53077975376197,
|
686 |
+
"grad_norm": 1.311870742321416,
|
687 |
+
"learning_rate": 9.828938438152418e-06,
|
688 |
+
"loss": 0.9009,
|
689 |
+
"step": 97
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 0.5362517099863201,
|
693 |
+
"grad_norm": 1.4372169411893168,
|
694 |
+
"learning_rate": 9.820543092187574e-06,
|
695 |
+
"loss": 0.9185,
|
696 |
+
"step": 98
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 0.5417236662106704,
|
700 |
+
"grad_norm": 1.3641071917754564,
|
701 |
+
"learning_rate": 9.811950398895457e-06,
|
702 |
+
"loss": 0.8773,
|
703 |
+
"step": 99
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 0.5471956224350205,
|
707 |
+
"grad_norm": 1.2729859663889775,
|
708 |
+
"learning_rate": 9.803160710050757e-06,
|
709 |
+
"loss": 0.9434,
|
710 |
+
"step": 100
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 0.5526675786593708,
|
714 |
+
"grad_norm": 1.3891437673950484,
|
715 |
+
"learning_rate": 9.79417438549293e-06,
|
716 |
+
"loss": 0.9904,
|
717 |
+
"step": 101
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 0.5581395348837209,
|
721 |
+
"grad_norm": 1.5015289964425491,
|
722 |
+
"learning_rate": 9.784991793111466e-06,
|
723 |
+
"loss": 0.9558,
|
724 |
+
"step": 102
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 0.5636114911080712,
|
728 |
+
"grad_norm": 1.2961636683011992,
|
729 |
+
"learning_rate": 9.775613308830825e-06,
|
730 |
+
"loss": 0.8844,
|
731 |
+
"step": 103
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 0.5690834473324213,
|
735 |
+
"grad_norm": 1.189881928766638,
|
736 |
+
"learning_rate": 9.766039316595048e-06,
|
737 |
+
"loss": 0.8812,
|
738 |
+
"step": 104
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.5745554035567716,
|
742 |
+
"grad_norm": 1.265576937966586,
|
743 |
+
"learning_rate": 9.756270208352048e-06,
|
744 |
+
"loss": 0.8349,
|
745 |
+
"step": 105
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.5800273597811217,
|
749 |
+
"grad_norm": 1.4272067437995613,
|
750 |
+
"learning_rate": 9.746306384037547e-06,
|
751 |
+
"loss": 0.9154,
|
752 |
+
"step": 106
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.585499316005472,
|
756 |
+
"grad_norm": 1.318753086910619,
|
757 |
+
"learning_rate": 9.736148251558719e-06,
|
758 |
+
"loss": 0.8915,
|
759 |
+
"step": 107
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.5909712722298222,
|
763 |
+
"grad_norm": 1.405278350886593,
|
764 |
+
"learning_rate": 9.725796226777478e-06,
|
765 |
+
"loss": 0.889,
|
766 |
+
"step": 108
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.5964432284541724,
|
770 |
+
"grad_norm": 1.3295796442633006,
|
771 |
+
"learning_rate": 9.715250733493466e-06,
|
772 |
+
"loss": 0.8909,
|
773 |
+
"step": 109
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.6019151846785226,
|
777 |
+
"grad_norm": 1.2515511449024315,
|
778 |
+
"learning_rate": 9.704512203426695e-06,
|
779 |
+
"loss": 0.8007,
|
780 |
+
"step": 110
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.6073871409028728,
|
784 |
+
"grad_norm": 1.338859170748462,
|
785 |
+
"learning_rate": 9.69358107619987e-06,
|
786 |
+
"loss": 0.8406,
|
787 |
+
"step": 111
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 0.612859097127223,
|
791 |
+
"grad_norm": 1.3869981352978968,
|
792 |
+
"learning_rate": 9.682457799320398e-06,
|
793 |
+
"loss": 0.9031,
|
794 |
+
"step": 112
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 0.6183310533515732,
|
798 |
+
"grad_norm": 1.282383836342225,
|
799 |
+
"learning_rate": 9.671142828162067e-06,
|
800 |
+
"loss": 0.8727,
|
801 |
+
"step": 113
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.6238030095759234,
|
805 |
+
"grad_norm": 1.4622746227313144,
|
806 |
+
"learning_rate": 9.659636625946403e-06,
|
807 |
+
"loss": 0.9793,
|
808 |
+
"step": 114
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 0.6292749658002736,
|
812 |
+
"grad_norm": 1.3250537140327348,
|
813 |
+
"learning_rate": 9.6479396637237e-06,
|
814 |
+
"loss": 0.8545,
|
815 |
+
"step": 115
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 0.6347469220246238,
|
819 |
+
"grad_norm": 1.3729794244714189,
|
820 |
+
"learning_rate": 9.636052420353744e-06,
|
821 |
+
"loss": 0.9157,
|
822 |
+
"step": 116
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 0.640218878248974,
|
826 |
+
"grad_norm": 1.4257039054662717,
|
827 |
+
"learning_rate": 9.623975382486206e-06,
|
828 |
+
"loss": 1.085,
|
829 |
+
"step": 117
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 0.6456908344733242,
|
833 |
+
"grad_norm": 1.396725871761606,
|
834 |
+
"learning_rate": 9.61170904454072e-06,
|
835 |
+
"loss": 0.9142,
|
836 |
+
"step": 118
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 0.6511627906976745,
|
840 |
+
"grad_norm": 1.3662875900817737,
|
841 |
+
"learning_rate": 9.599253908686637e-06,
|
842 |
+
"loss": 0.8722,
|
843 |
+
"step": 119
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 0.6566347469220246,
|
847 |
+
"grad_norm": 1.412280837869338,
|
848 |
+
"learning_rate": 9.586610484822473e-06,
|
849 |
+
"loss": 0.9536,
|
850 |
+
"step": 120
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 0.6621067031463749,
|
854 |
+
"grad_norm": 1.4012864642846434,
|
855 |
+
"learning_rate": 9.57377929055504e-06,
|
856 |
+
"loss": 0.9002,
|
857 |
+
"step": 121
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 0.667578659370725,
|
861 |
+
"grad_norm": 1.3456409823375266,
|
862 |
+
"learning_rate": 9.56076085117823e-06,
|
863 |
+
"loss": 0.8881,
|
864 |
+
"step": 122
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 0.6730506155950753,
|
868 |
+
"grad_norm": 1.317365944067738,
|
869 |
+
"learning_rate": 9.547555699651546e-06,
|
870 |
+
"loss": 0.8939,
|
871 |
+
"step": 123
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 0.6785225718194254,
|
875 |
+
"grad_norm": 1.2804552485881544,
|
876 |
+
"learning_rate": 9.534164376578257e-06,
|
877 |
+
"loss": 0.8814,
|
878 |
+
"step": 124
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 0.6839945280437757,
|
882 |
+
"grad_norm": 1.4025346265827956,
|
883 |
+
"learning_rate": 9.520587430183275e-06,
|
884 |
+
"loss": 0.8962,
|
885 |
+
"step": 125
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 0.6894664842681258,
|
889 |
+
"grad_norm": 1.3713569063571371,
|
890 |
+
"learning_rate": 9.506825416290712e-06,
|
891 |
+
"loss": 0.8379,
|
892 |
+
"step": 126
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 0.6949384404924761,
|
896 |
+
"grad_norm": 1.3271468439938563,
|
897 |
+
"learning_rate": 9.492878898301126e-06,
|
898 |
+
"loss": 0.825,
|
899 |
+
"step": 127
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 0.7004103967168263,
|
903 |
+
"grad_norm": 1.367623422591148,
|
904 |
+
"learning_rate": 9.47874844716845e-06,
|
905 |
+
"loss": 0.8565,
|
906 |
+
"step": 128
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 0.7058823529411765,
|
910 |
+
"grad_norm": 1.4176894490822682,
|
911 |
+
"learning_rate": 9.464434641376626e-06,
|
912 |
+
"loss": 0.9507,
|
913 |
+
"step": 129
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 0.7113543091655267,
|
917 |
+
"grad_norm": 1.287419108983359,
|
918 |
+
"learning_rate": 9.449938066915918e-06,
|
919 |
+
"loss": 0.8671,
|
920 |
+
"step": 130
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 0.7168262653898769,
|
924 |
+
"grad_norm": 1.5338023811645511,
|
925 |
+
"learning_rate": 9.435259317258922e-06,
|
926 |
+
"loss": 0.8903,
|
927 |
+
"step": 131
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 0.7222982216142271,
|
931 |
+
"grad_norm": 1.33948451169851,
|
932 |
+
"learning_rate": 9.42039899333627e-06,
|
933 |
+
"loss": 0.9351,
|
934 |
+
"step": 132
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 0.7277701778385773,
|
938 |
+
"grad_norm": 1.47805293101446,
|
939 |
+
"learning_rate": 9.40535770351203e-06,
|
940 |
+
"loss": 0.9344,
|
941 |
+
"step": 133
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 0.7332421340629275,
|
945 |
+
"grad_norm": 1.3344811523812627,
|
946 |
+
"learning_rate": 9.390136063558803e-06,
|
947 |
+
"loss": 0.8538,
|
948 |
+
"step": 134
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 0.7387140902872777,
|
952 |
+
"grad_norm": 1.2538675374089643,
|
953 |
+
"learning_rate": 9.374734696632508e-06,
|
954 |
+
"loss": 0.851,
|
955 |
+
"step": 135
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 0.7441860465116279,
|
959 |
+
"grad_norm": 1.355445116330821,
|
960 |
+
"learning_rate": 9.359154233246872e-06,
|
961 |
+
"loss": 0.9007,
|
962 |
+
"step": 136
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 0.7496580027359782,
|
966 |
+
"grad_norm": 1.5186977614121933,
|
967 |
+
"learning_rate": 9.34339531124762e-06,
|
968 |
+
"loss": 0.8761,
|
969 |
+
"step": 137
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 0.7551299589603283,
|
973 |
+
"grad_norm": 1.40745828136921,
|
974 |
+
"learning_rate": 9.327458575786365e-06,
|
975 |
+
"loss": 0.9599,
|
976 |
+
"step": 138
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 0.7606019151846786,
|
980 |
+
"grad_norm": 1.3718619422077885,
|
981 |
+
"learning_rate": 9.311344679294186e-06,
|
982 |
+
"loss": 0.9159,
|
983 |
+
"step": 139
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 0.7660738714090287,
|
987 |
+
"grad_norm": 1.3494245404539824,
|
988 |
+
"learning_rate": 9.295054281454931e-06,
|
989 |
+
"loss": 0.8942,
|
990 |
+
"step": 140
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 0.771545827633379,
|
994 |
+
"grad_norm": 1.4052370328097452,
|
995 |
+
"learning_rate": 9.278588049178203e-06,
|
996 |
+
"loss": 0.9061,
|
997 |
+
"step": 141
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 0.7770177838577291,
|
1001 |
+
"grad_norm": 1.3316617433442073,
|
1002 |
+
"learning_rate": 9.261946656572056e-06,
|
1003 |
+
"loss": 0.8498,
|
1004 |
+
"step": 142
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 0.7824897400820794,
|
1008 |
+
"grad_norm": 1.366916406757608,
|
1009 |
+
"learning_rate": 9.2451307849154e-06,
|
1010 |
+
"loss": 0.9037,
|
1011 |
+
"step": 143
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 0.7879616963064295,
|
1015 |
+
"grad_norm": 1.2903864487223162,
|
1016 |
+
"learning_rate": 9.228141122630114e-06,
|
1017 |
+
"loss": 0.8793,
|
1018 |
+
"step": 144
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 0.7934336525307798,
|
1022 |
+
"grad_norm": 1.3932457100420763,
|
1023 |
+
"learning_rate": 9.21097836525286e-06,
|
1024 |
+
"loss": 0.9432,
|
1025 |
+
"step": 145
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 0.79890560875513,
|
1029 |
+
"grad_norm": 1.5137405657021465,
|
1030 |
+
"learning_rate": 9.193643215406605e-06,
|
1031 |
+
"loss": 0.862,
|
1032 |
+
"step": 146
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 0.8043775649794802,
|
1036 |
+
"grad_norm": 1.3653166193627995,
|
1037 |
+
"learning_rate": 9.176136382771861e-06,
|
1038 |
+
"loss": 0.8631,
|
1039 |
+
"step": 147
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 0.8098495212038304,
|
1043 |
+
"grad_norm": 1.4506283043215642,
|
1044 |
+
"learning_rate": 9.15845858405763e-06,
|
1045 |
+
"loss": 1.0209,
|
1046 |
+
"step": 148
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 0.8153214774281806,
|
1050 |
+
"grad_norm": 1.335415236914369,
|
1051 |
+
"learning_rate": 9.140610542972064e-06,
|
1052 |
+
"loss": 0.9388,
|
1053 |
+
"step": 149
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 0.8207934336525308,
|
1057 |
+
"grad_norm": 1.3544652215536117,
|
1058 |
+
"learning_rate": 9.122592990192836e-06,
|
1059 |
+
"loss": 0.8924,
|
1060 |
+
"step": 150
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 0.826265389876881,
|
1064 |
+
"grad_norm": 1.4422771448220417,
|
1065 |
+
"learning_rate": 9.104406663337222e-06,
|
1066 |
+
"loss": 0.8725,
|
1067 |
+
"step": 151
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 0.8317373461012312,
|
1071 |
+
"grad_norm": 1.3481241477062407,
|
1072 |
+
"learning_rate": 9.086052306931919e-06,
|
1073 |
+
"loss": 0.8792,
|
1074 |
+
"step": 152
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 0.8372093023255814,
|
1078 |
+
"grad_norm": 1.4349960850309402,
|
1079 |
+
"learning_rate": 9.067530672382544e-06,
|
1080 |
+
"loss": 0.8701,
|
1081 |
+
"step": 153
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 0.8426812585499316,
|
1085 |
+
"grad_norm": 1.4297585081861366,
|
1086 |
+
"learning_rate": 9.04884251794289e-06,
|
1087 |
+
"loss": 0.8282,
|
1088 |
+
"step": 154
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 0.8481532147742818,
|
1092 |
+
"grad_norm": 1.3910006203559417,
|
1093 |
+
"learning_rate": 9.029988608683877e-06,
|
1094 |
+
"loss": 0.8931,
|
1095 |
+
"step": 155
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 0.853625170998632,
|
1099 |
+
"grad_norm": 1.4471745500262192,
|
1100 |
+
"learning_rate": 9.010969716462228e-06,
|
1101 |
+
"loss": 0.8944,
|
1102 |
+
"step": 156
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 0.8590971272229823,
|
1106 |
+
"grad_norm": 1.5363876296675514,
|
1107 |
+
"learning_rate": 8.991786619888873e-06,
|
1108 |
+
"loss": 0.8643,
|
1109 |
+
"step": 157
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 0.8645690834473324,
|
1113 |
+
"grad_norm": 1.5499000970863974,
|
1114 |
+
"learning_rate": 8.972440104297078e-06,
|
1115 |
+
"loss": 0.9268,
|
1116 |
+
"step": 158
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 0.8700410396716827,
|
1120 |
+
"grad_norm": 1.4863820608676213,
|
1121 |
+
"learning_rate": 8.952930961710284e-06,
|
1122 |
+
"loss": 0.8852,
|
1123 |
+
"step": 159
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 0.8755129958960328,
|
1127 |
+
"grad_norm": 1.3313314949898831,
|
1128 |
+
"learning_rate": 8.933259990809698e-06,
|
1129 |
+
"loss": 0.9337,
|
1130 |
+
"step": 160
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 0.8809849521203831,
|
1134 |
+
"grad_norm": 1.49818548160592,
|
1135 |
+
"learning_rate": 8.913427996901574e-06,
|
1136 |
+
"loss": 0.9081,
|
1137 |
+
"step": 161
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 0.8864569083447332,
|
1141 |
+
"grad_norm": 1.462278193593804,
|
1142 |
+
"learning_rate": 8.893435791884268e-06,
|
1143 |
+
"loss": 0.7921,
|
1144 |
+
"step": 162
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 0.8919288645690835,
|
1148 |
+
"grad_norm": 1.3626875771991256,
|
1149 |
+
"learning_rate": 8.873284194214983e-06,
|
1150 |
+
"loss": 0.8238,
|
1151 |
+
"step": 163
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 0.8974008207934336,
|
1155 |
+
"grad_norm": 1.2683693736678303,
|
1156 |
+
"learning_rate": 8.852974028876273e-06,
|
1157 |
+
"loss": 0.8774,
|
1158 |
+
"step": 164
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 0.9028727770177839,
|
1162 |
+
"grad_norm": 1.3492304042005043,
|
1163 |
+
"learning_rate": 8.832506127342261e-06,
|
1164 |
+
"loss": 0.8396,
|
1165 |
+
"step": 165
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 0.908344733242134,
|
1169 |
+
"grad_norm": 1.2400818250532626,
|
1170 |
+
"learning_rate": 8.811881327544604e-06,
|
1171 |
+
"loss": 0.829,
|
1172 |
+
"step": 166
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 0.9138166894664843,
|
1176 |
+
"grad_norm": 1.3920072983751164,
|
1177 |
+
"learning_rate": 8.79110047383819e-06,
|
1178 |
+
"loss": 0.8345,
|
1179 |
+
"step": 167
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 0.9192886456908345,
|
1183 |
+
"grad_norm": 1.5341288288004693,
|
1184 |
+
"learning_rate": 8.770164416966565e-06,
|
1185 |
+
"loss": 0.9128,
|
1186 |
+
"step": 168
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 0.9247606019151847,
|
1190 |
+
"grad_norm": 1.3803948083641058,
|
1191 |
+
"learning_rate": 8.749074014027116e-06,
|
1192 |
+
"loss": 0.8803,
|
1193 |
+
"step": 169
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 0.9302325581395349,
|
1197 |
+
"grad_norm": 1.3951333336999596,
|
1198 |
+
"learning_rate": 8.727830128435965e-06,
|
1199 |
+
"loss": 0.8813,
|
1200 |
+
"step": 170
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 0.9357045143638851,
|
1204 |
+
"grad_norm": 1.5299036284742302,
|
1205 |
+
"learning_rate": 8.706433629892643e-06,
|
1206 |
+
"loss": 0.8473,
|
1207 |
+
"step": 171
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 0.9411764705882353,
|
1211 |
+
"grad_norm": 1.4178298744964428,
|
1212 |
+
"learning_rate": 8.684885394344469e-06,
|
1213 |
+
"loss": 0.8233,
|
1214 |
+
"step": 172
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 0.9466484268125855,
|
1218 |
+
"grad_norm": 1.4465269313848104,
|
1219 |
+
"learning_rate": 8.663186303950694e-06,
|
1220 |
+
"loss": 0.8639,
|
1221 |
+
"step": 173
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 0.9521203830369357,
|
1225 |
+
"grad_norm": 1.4518576453167145,
|
1226 |
+
"learning_rate": 8.64133724704639e-06,
|
1227 |
+
"loss": 0.8479,
|
1228 |
+
"step": 174
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 0.957592339261286,
|
1232 |
+
"grad_norm": 1.4475079213628008,
|
1233 |
+
"learning_rate": 8.61933911810608e-06,
|
1234 |
+
"loss": 0.772,
|
1235 |
+
"step": 175
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 0.9630642954856361,
|
1239 |
+
"grad_norm": 1.4662882290270267,
|
1240 |
+
"learning_rate": 8.597192817707121e-06,
|
1241 |
+
"loss": 0.8885,
|
1242 |
+
"step": 176
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 0.9685362517099864,
|
1246 |
+
"grad_norm": 1.3414157064655197,
|
1247 |
+
"learning_rate": 8.574899252492833e-06,
|
1248 |
+
"loss": 0.8218,
|
1249 |
+
"step": 177
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 0.9740082079343365,
|
1253 |
+
"grad_norm": 1.347041000686313,
|
1254 |
+
"learning_rate": 8.552459335135382e-06,
|
1255 |
+
"loss": 0.8549,
|
1256 |
+
"step": 178
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 0.9794801641586868,
|
1260 |
+
"grad_norm": 1.4274303795160541,
|
1261 |
+
"learning_rate": 8.529873984298418e-06,
|
1262 |
+
"loss": 0.893,
|
1263 |
+
"step": 179
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 0.9849521203830369,
|
1267 |
+
"grad_norm": 1.3201028415081821,
|
1268 |
+
"learning_rate": 8.507144124599469e-06,
|
1269 |
+
"loss": 0.7664,
|
1270 |
+
"step": 180
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 0.9904240766073872,
|
1274 |
+
"grad_norm": 1.5434924259568004,
|
1275 |
+
"learning_rate": 8.484270686572078e-06,
|
1276 |
+
"loss": 0.8556,
|
1277 |
+
"step": 181
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 0.9958960328317373,
|
1281 |
+
"grad_norm": 1.4140319742585339,
|
1282 |
+
"learning_rate": 8.46125460662772e-06,
|
1283 |
+
"loss": 0.8664,
|
1284 |
+
"step": 182
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 1.0,
|
1288 |
+
"grad_norm": 1.4140319742585339,
|
1289 |
+
"learning_rate": 8.438096827017461e-06,
|
1290 |
+
"loss": 0.8435,
|
1291 |
+
"step": 183
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 1.0054719562243501,
|
1295 |
+
"grad_norm": 1.713669355579185,
|
1296 |
+
"learning_rate": 8.41479829579338e-06,
|
1297 |
+
"loss": 0.6684,
|
1298 |
+
"step": 184
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 1.0109439124487003,
|
1302 |
+
"grad_norm": 1.3053699773113805,
|
1303 |
+
"learning_rate": 8.391359966769765e-06,
|
1304 |
+
"loss": 0.6753,
|
1305 |
+
"step": 185
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 1.0164158686730507,
|
1309 |
+
"grad_norm": 1.2915343576520686,
|
1310 |
+
"learning_rate": 8.367782799484057e-06,
|
1311 |
+
"loss": 0.724,
|
1312 |
+
"step": 186
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 1.0218878248974008,
|
1316 |
+
"grad_norm": 1.3705945528435555,
|
1317 |
+
"learning_rate": 8.344067759157572e-06,
|
1318 |
+
"loss": 0.6771,
|
1319 |
+
"step": 187
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 1.027359781121751,
|
1323 |
+
"grad_norm": 1.329823752209595,
|
1324 |
+
"learning_rate": 8.320215816655987e-06,
|
1325 |
+
"loss": 0.7047,
|
1326 |
+
"step": 188
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 1.0328317373461013,
|
1330 |
+
"grad_norm": 1.8441048694209619,
|
1331 |
+
"learning_rate": 8.296227948449589e-06,
|
1332 |
+
"loss": 0.7324,
|
1333 |
+
"step": 189
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 1.0383036935704515,
|
1337 |
+
"grad_norm": 1.3355378928572972,
|
1338 |
+
"learning_rate": 8.272105136573304e-06,
|
1339 |
+
"loss": 0.6813,
|
1340 |
+
"step": 190
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 1.0437756497948016,
|
1344 |
+
"grad_norm": 1.3153848906746584,
|
1345 |
+
"learning_rate": 8.247848368586491e-06,
|
1346 |
+
"loss": 0.7278,
|
1347 |
+
"step": 191
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 1.0492476060191518,
|
1351 |
+
"grad_norm": 1.3367252138888372,
|
1352 |
+
"learning_rate": 8.223458637532516e-06,
|
1353 |
+
"loss": 0.6951,
|
1354 |
+
"step": 192
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 1.054719562243502,
|
1358 |
+
"grad_norm": 1.561179528729577,
|
1359 |
+
"learning_rate": 8.198936941898092e-06,
|
1360 |
+
"loss": 0.634,
|
1361 |
+
"step": 193
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 1.0601915184678523,
|
1365 |
+
"grad_norm": 1.3962450107109083,
|
1366 |
+
"learning_rate": 8.174284285572408e-06,
|
1367 |
+
"loss": 0.6791,
|
1368 |
+
"step": 194
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 1.0656634746922025,
|
1372 |
+
"grad_norm": 1.3499529359765832,
|
1373 |
+
"learning_rate": 8.14950167780603e-06,
|
1374 |
+
"loss": 0.6872,
|
1375 |
+
"step": 195
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 1.0711354309165526,
|
1379 |
+
"grad_norm": 1.421181424986065,
|
1380 |
+
"learning_rate": 8.12459013316958e-06,
|
1381 |
+
"loss": 0.6869,
|
1382 |
+
"step": 196
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 1.076607387140903,
|
1386 |
+
"grad_norm": 1.5673918401772315,
|
1387 |
+
"learning_rate": 8.099550671512202e-06,
|
1388 |
+
"loss": 0.7162,
|
1389 |
+
"step": 197
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 1.0820793433652531,
|
1393 |
+
"grad_norm": 1.4609331044020695,
|
1394 |
+
"learning_rate": 8.074384317919812e-06,
|
1395 |
+
"loss": 0.6265,
|
1396 |
+
"step": 198
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 1.0875512995896033,
|
1400 |
+
"grad_norm": 1.3348736952463767,
|
1401 |
+
"learning_rate": 8.049092102673135e-06,
|
1402 |
+
"loss": 0.6755,
|
1403 |
+
"step": 199
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 1.0930232558139534,
|
1407 |
+
"grad_norm": 1.4635980596688503,
|
1408 |
+
"learning_rate": 8.023675061205519e-06,
|
1409 |
+
"loss": 0.7468,
|
1410 |
+
"step": 200
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 1.0984952120383036,
|
1414 |
+
"grad_norm": 1.430054757096999,
|
1415 |
+
"learning_rate": 7.99813423406055e-06,
|
1416 |
+
"loss": 0.6451,
|
1417 |
+
"step": 201
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 1.103967168262654,
|
1421 |
+
"grad_norm": 1.4136438116646735,
|
1422 |
+
"learning_rate": 7.972470666849456e-06,
|
1423 |
+
"loss": 0.6455,
|
1424 |
+
"step": 202
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 1.109439124487004,
|
1428 |
+
"grad_norm": 1.4627384137295982,
|
1429 |
+
"learning_rate": 7.946685410208297e-06,
|
1430 |
+
"loss": 0.6776,
|
1431 |
+
"step": 203
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 1.1149110807113543,
|
1435 |
+
"grad_norm": 1.4215864940794252,
|
1436 |
+
"learning_rate": 7.920779519754949e-06,
|
1437 |
+
"loss": 0.6216,
|
1438 |
+
"step": 204
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 1.1203830369357046,
|
1442 |
+
"grad_norm": 1.3645813012632697,
|
1443 |
+
"learning_rate": 7.894754056045902e-06,
|
1444 |
+
"loss": 0.7176,
|
1445 |
+
"step": 205
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 1.1258549931600548,
|
1449 |
+
"grad_norm": 1.4372059540489919,
|
1450 |
+
"learning_rate": 7.868610084532828e-06,
|
1451 |
+
"loss": 0.7148,
|
1452 |
+
"step": 206
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 1.131326949384405,
|
1456 |
+
"grad_norm": 1.524978523277985,
|
1457 |
+
"learning_rate": 7.842348675518968e-06,
|
1458 |
+
"loss": 0.6616,
|
1459 |
+
"step": 207
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 1.136798905608755,
|
1463 |
+
"grad_norm": 1.3444970409298955,
|
1464 |
+
"learning_rate": 7.81597090411532e-06,
|
1465 |
+
"loss": 0.9283,
|
1466 |
+
"step": 208
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 1.1422708618331052,
|
1470 |
+
"grad_norm": 1.4174330989035344,
|
1471 |
+
"learning_rate": 7.789477850196615e-06,
|
1472 |
+
"loss": 0.6953,
|
1473 |
+
"step": 209
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 1.1477428180574556,
|
1477 |
+
"grad_norm": 1.444706767187531,
|
1478 |
+
"learning_rate": 7.762870598357114e-06,
|
1479 |
+
"loss": 0.6362,
|
1480 |
+
"step": 210
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 1.1532147742818057,
|
1484 |
+
"grad_norm": 1.4213314960182526,
|
1485 |
+
"learning_rate": 7.736150237866214e-06,
|
1486 |
+
"loss": 0.6879,
|
1487 |
+
"step": 211
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 1.158686730506156,
|
1491 |
+
"grad_norm": 1.430848865627854,
|
1492 |
+
"learning_rate": 7.709317862623833e-06,
|
1493 |
+
"loss": 0.6706,
|
1494 |
+
"step": 212
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 1.1641586867305063,
|
1498 |
+
"grad_norm": 1.4413260835226156,
|
1499 |
+
"learning_rate": 7.682374571115651e-06,
|
1500 |
+
"loss": 0.685,
|
1501 |
+
"step": 213
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 1.1696306429548564,
|
1505 |
+
"grad_norm": 1.3398531392616837,
|
1506 |
+
"learning_rate": 7.655321466368126e-06,
|
1507 |
+
"loss": 0.6801,
|
1508 |
+
"step": 214
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 1.1751025991792066,
|
1512 |
+
"grad_norm": 1.3769559986838098,
|
1513 |
+
"learning_rate": 7.628159655903335e-06,
|
1514 |
+
"loss": 0.6827,
|
1515 |
+
"step": 215
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 1.1805745554035567,
|
1519 |
+
"grad_norm": 1.5881264885919462,
|
1520 |
+
"learning_rate": 7.600890251693645e-06,
|
1521 |
+
"loss": 0.7006,
|
1522 |
+
"step": 216
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 1.1860465116279069,
|
1526 |
+
"grad_norm": 1.3625178539942244,
|
1527 |
+
"learning_rate": 7.5735143701161806e-06,
|
1528 |
+
"loss": 0.6949,
|
1529 |
+
"step": 217
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 1.1915184678522572,
|
1533 |
+
"grad_norm": 1.3500719795951182,
|
1534 |
+
"learning_rate": 7.546033131907123e-06,
|
1535 |
+
"loss": 0.7234,
|
1536 |
+
"step": 218
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 1.1969904240766074,
|
1540 |
+
"grad_norm": 1.3910329268683956,
|
1541 |
+
"learning_rate": 7.518447662115831e-06,
|
1542 |
+
"loss": 0.6441,
|
1543 |
+
"step": 219
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 1.2024623803009575,
|
1547 |
+
"grad_norm": 1.3754700565582025,
|
1548 |
+
"learning_rate": 7.490759090058779e-06,
|
1549 |
+
"loss": 0.6887,
|
1550 |
+
"step": 220
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 1.207934336525308,
|
1554 |
+
"grad_norm": 1.4425634456711103,
|
1555 |
+
"learning_rate": 7.462968549273326e-06,
|
1556 |
+
"loss": 0.6521,
|
1557 |
+
"step": 221
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 1.213406292749658,
|
1561 |
+
"grad_norm": 1.3004808641883596,
|
1562 |
+
"learning_rate": 7.435077177471315e-06,
|
1563 |
+
"loss": 0.7551,
|
1564 |
+
"step": 222
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 1.2188782489740082,
|
1568 |
+
"grad_norm": 1.5033009665031583,
|
1569 |
+
"learning_rate": 7.407086116492484e-06,
|
1570 |
+
"loss": 0.7039,
|
1571 |
+
"step": 223
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 1.2243502051983584,
|
1575 |
+
"grad_norm": 1.5037724031410729,
|
1576 |
+
"learning_rate": 7.378996512257735e-06,
|
1577 |
+
"loss": 0.673,
|
1578 |
+
"step": 224
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 1.2298221614227085,
|
1582 |
+
"grad_norm": 1.4425918228147547,
|
1583 |
+
"learning_rate": 7.3508095147222095e-06,
|
1584 |
+
"loss": 0.7001,
|
1585 |
+
"step": 225
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 1.2352941176470589,
|
1589 |
+
"grad_norm": 1.3948172696844245,
|
1590 |
+
"learning_rate": 7.322526277828215e-06,
|
1591 |
+
"loss": 0.6754,
|
1592 |
+
"step": 226
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 1.240766073871409,
|
1596 |
+
"grad_norm": 1.4343282628258913,
|
1597 |
+
"learning_rate": 7.294147959457989e-06,
|
1598 |
+
"loss": 0.6546,
|
1599 |
+
"step": 227
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 1.2462380300957592,
|
1603 |
+
"grad_norm": 1.3631132979844476,
|
1604 |
+
"learning_rate": 7.265675721386285e-06,
|
1605 |
+
"loss": 0.6755,
|
1606 |
+
"step": 228
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 1.2517099863201095,
|
1610 |
+
"grad_norm": 1.3712692071490096,
|
1611 |
+
"learning_rate": 7.237110729232826e-06,
|
1612 |
+
"loss": 0.7518,
|
1613 |
+
"step": 229
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 1.2571819425444597,
|
1617 |
+
"grad_norm": 1.4772756426462441,
|
1618 |
+
"learning_rate": 7.208454152414571e-06,
|
1619 |
+
"loss": 0.6776,
|
1620 |
+
"step": 230
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 1.2626538987688098,
|
1624 |
+
"grad_norm": 1.4696793742280645,
|
1625 |
+
"learning_rate": 7.179707164097852e-06,
|
1626 |
+
"loss": 0.7275,
|
1627 |
+
"step": 231
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 1.26812585499316,
|
1631 |
+
"grad_norm": 1.5474770696875386,
|
1632 |
+
"learning_rate": 7.1508709411503364e-06,
|
1633 |
+
"loss": 0.6374,
|
1634 |
+
"step": 232
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 1.2735978112175101,
|
1638 |
+
"grad_norm": 1.3364576808992332,
|
1639 |
+
"learning_rate": 7.12194666409285e-06,
|
1640 |
+
"loss": 0.6862,
|
1641 |
+
"step": 233
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 1.2790697674418605,
|
1645 |
+
"grad_norm": 1.4636731811623045,
|
1646 |
+
"learning_rate": 7.092935517051059e-06,
|
1647 |
+
"loss": 0.7242,
|
1648 |
+
"step": 234
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 1.2845417236662107,
|
1652 |
+
"grad_norm": 1.499332510440258,
|
1653 |
+
"learning_rate": 7.063838687706971e-06,
|
1654 |
+
"loss": 0.6882,
|
1655 |
+
"step": 235
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 1.2900136798905608,
|
1659 |
+
"grad_norm": 1.3368622966591606,
|
1660 |
+
"learning_rate": 7.034657367250337e-06,
|
1661 |
+
"loss": 0.6715,
|
1662 |
+
"step": 236
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 1.2954856361149112,
|
1666 |
+
"grad_norm": 1.3701261695053024,
|
1667 |
+
"learning_rate": 7.005392750329868e-06,
|
1668 |
+
"loss": 0.6845,
|
1669 |
+
"step": 237
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 1.3009575923392613,
|
1673 |
+
"grad_norm": 1.5188057782135747,
|
1674 |
+
"learning_rate": 6.976046035004334e-06,
|
1675 |
+
"loss": 0.7223,
|
1676 |
+
"step": 238
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 1.3064295485636115,
|
1680 |
+
"grad_norm": 1.466364512074953,
|
1681 |
+
"learning_rate": 6.946618422693521e-06,
|
1682 |
+
"loss": 0.714,
|
1683 |
+
"step": 239
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 1.3119015047879616,
|
1687 |
+
"grad_norm": 1.4154655783077306,
|
1688 |
+
"learning_rate": 6.917111118129035e-06,
|
1689 |
+
"loss": 0.674,
|
1690 |
+
"step": 240
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 1.3173734610123118,
|
1694 |
+
"grad_norm": 1.349310045400673,
|
1695 |
+
"learning_rate": 6.887525329304994e-06,
|
1696 |
+
"loss": 0.6741,
|
1697 |
+
"step": 241
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 1.3228454172366622,
|
1701 |
+
"grad_norm": 1.4002081915684452,
|
1702 |
+
"learning_rate": 6.8578622674285636e-06,
|
1703 |
+
"loss": 0.6539,
|
1704 |
+
"step": 242
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 1.3283173734610123,
|
1708 |
+
"grad_norm": 1.2842182847691381,
|
1709 |
+
"learning_rate": 6.828123146870384e-06,
|
1710 |
+
"loss": 0.6332,
|
1711 |
+
"step": 243
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 1.3337893296853625,
|
1715 |
+
"grad_norm": 1.489698740247267,
|
1716 |
+
"learning_rate": 6.798309185114839e-06,
|
1717 |
+
"loss": 0.7209,
|
1718 |
+
"step": 244
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 1.3392612859097128,
|
1722 |
+
"grad_norm": 1.515325663423874,
|
1723 |
+
"learning_rate": 6.76842160271023e-06,
|
1724 |
+
"loss": 0.6882,
|
1725 |
+
"step": 245
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 1.344733242134063,
|
1729 |
+
"grad_norm": 1.4630372402247052,
|
1730 |
+
"learning_rate": 6.738461623218795e-06,
|
1731 |
+
"loss": 0.6617,
|
1732 |
+
"step": 246
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 1.3502051983584131,
|
1736 |
+
"grad_norm": 1.53340581982168,
|
1737 |
+
"learning_rate": 6.708430473166628e-06,
|
1738 |
+
"loss": 0.7117,
|
1739 |
+
"step": 247
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 1.3556771545827633,
|
1743 |
+
"grad_norm": 1.7714471643819227,
|
1744 |
+
"learning_rate": 6.678329381993459e-06,
|
1745 |
+
"loss": 0.7377,
|
1746 |
+
"step": 248
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 1.3611491108071134,
|
1750 |
+
"grad_norm": 1.4372943101035542,
|
1751 |
+
"learning_rate": 6.648159582002323e-06,
|
1752 |
+
"loss": 0.6555,
|
1753 |
+
"step": 249
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 1.3666210670314638,
|
1757 |
+
"grad_norm": 1.4675367254022897,
|
1758 |
+
"learning_rate": 6.617922308309115e-06,
|
1759 |
+
"loss": 0.6755,
|
1760 |
+
"step": 250
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 1.372093023255814,
|
1764 |
+
"grad_norm": 1.4445829677301385,
|
1765 |
+
"learning_rate": 6.5876187987920224e-06,
|
1766 |
+
"loss": 0.7938,
|
1767 |
+
"step": 251
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 1.377564979480164,
|
1771 |
+
"grad_norm": 1.313511597032194,
|
1772 |
+
"learning_rate": 6.557250294040849e-06,
|
1773 |
+
"loss": 0.6595,
|
1774 |
+
"step": 252
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 1.3830369357045145,
|
1778 |
+
"grad_norm": 1.2940515217479573,
|
1779 |
+
"learning_rate": 6.526818037306229e-06,
|
1780 |
+
"loss": 0.6572,
|
1781 |
+
"step": 253
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 1.3885088919288646,
|
1785 |
+
"grad_norm": 1.4219460360392615,
|
1786 |
+
"learning_rate": 6.496323274448721e-06,
|
1787 |
+
"loss": 0.6567,
|
1788 |
+
"step": 254
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 1.3939808481532148,
|
1792 |
+
"grad_norm": 1.316969471801699,
|
1793 |
+
"learning_rate": 6.46576725388782e-06,
|
1794 |
+
"loss": 0.6738,
|
1795 |
+
"step": 255
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 1.399452804377565,
|
1799 |
+
"grad_norm": 1.4211635603950936,
|
1800 |
+
"learning_rate": 6.435151226550829e-06,
|
1801 |
+
"loss": 0.6373,
|
1802 |
+
"step": 256
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 1.404924760601915,
|
1806 |
+
"grad_norm": 1.3942664871826713,
|
1807 |
+
"learning_rate": 6.404476445821664e-06,
|
1808 |
+
"loss": 0.6516,
|
1809 |
+
"step": 257
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 1.4103967168262654,
|
1813 |
+
"grad_norm": 1.3705051299722797,
|
1814 |
+
"learning_rate": 6.373744167489532e-06,
|
1815 |
+
"loss": 0.7037,
|
1816 |
+
"step": 258
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 1.4158686730506156,
|
1820 |
+
"grad_norm": 1.4662043274329581,
|
1821 |
+
"learning_rate": 6.342955649697523e-06,
|
1822 |
+
"loss": 0.6833,
|
1823 |
+
"step": 259
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 1.4213406292749657,
|
1827 |
+
"grad_norm": 1.3232763115495254,
|
1828 |
+
"learning_rate": 6.312112152891108e-06,
|
1829 |
+
"loss": 0.7986,
|
1830 |
+
"step": 260
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 1.426812585499316,
|
1834 |
+
"grad_norm": 1.439560375340796,
|
1835 |
+
"learning_rate": 6.28121493976653e-06,
|
1836 |
+
"loss": 0.7614,
|
1837 |
+
"step": 261
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 1.4322845417236663,
|
1841 |
+
"grad_norm": 1.3569771740898653,
|
1842 |
+
"learning_rate": 6.250265275219116e-06,
|
1843 |
+
"loss": 0.6659,
|
1844 |
+
"step": 262
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 1.4377564979480164,
|
1848 |
+
"grad_norm": 1.416406770109305,
|
1849 |
+
"learning_rate": 6.219264426291495e-06,
|
1850 |
+
"loss": 0.7138,
|
1851 |
+
"step": 263
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 1.4432284541723666,
|
1855 |
+
"grad_norm": 1.5155691945596343,
|
1856 |
+
"learning_rate": 6.188213662121716e-06,
|
1857 |
+
"loss": 0.6818,
|
1858 |
+
"step": 264
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 1.4487004103967167,
|
1862 |
+
"grad_norm": 1.3869462794071394,
|
1863 |
+
"learning_rate": 6.157114253891307e-06,
|
1864 |
+
"loss": 0.7512,
|
1865 |
+
"step": 265
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 1.454172366621067,
|
1869 |
+
"grad_norm": 1.301395951476852,
|
1870 |
+
"learning_rate": 6.125967474773223e-06,
|
1871 |
+
"loss": 0.8031,
|
1872 |
+
"step": 266
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 1.4596443228454172,
|
1876 |
+
"grad_norm": 1.3983861170684568,
|
1877 |
+
"learning_rate": 6.094774599879727e-06,
|
1878 |
+
"loss": 0.6597,
|
1879 |
+
"step": 267
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 1.4651162790697674,
|
1883 |
+
"grad_norm": 1.3540672829133027,
|
1884 |
+
"learning_rate": 6.063536906210188e-06,
|
1885 |
+
"loss": 0.688,
|
1886 |
+
"step": 268
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 1.4705882352941178,
|
1890 |
+
"grad_norm": 1.3775967615458706,
|
1891 |
+
"learning_rate": 6.032255672598803e-06,
|
1892 |
+
"loss": 0.6852,
|
1893 |
+
"step": 269
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 1.476060191518468,
|
1897 |
+
"grad_norm": 1.5182115850591373,
|
1898 |
+
"learning_rate": 6.000932179662245e-06,
|
1899 |
+
"loss": 0.6753,
|
1900 |
+
"step": 270
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 1.481532147742818,
|
1904 |
+
"grad_norm": 1.5147658031887823,
|
1905 |
+
"learning_rate": 5.969567709747228e-06,
|
1906 |
+
"loss": 0.6402,
|
1907 |
+
"step": 271
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 1.4870041039671682,
|
1911 |
+
"grad_norm": 1.471719828466369,
|
1912 |
+
"learning_rate": 5.938163546878024e-06,
|
1913 |
+
"loss": 0.7131,
|
1914 |
+
"step": 272
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 1.4924760601915183,
|
1918 |
+
"grad_norm": 1.4935611505696107,
|
1919 |
+
"learning_rate": 5.906720976703877e-06,
|
1920 |
+
"loss": 0.636,
|
1921 |
+
"step": 273
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 1.4979480164158687,
|
1925 |
+
"grad_norm": 1.2991699624432158,
|
1926 |
+
"learning_rate": 5.875241286446391e-06,
|
1927 |
+
"loss": 0.7288,
|
1928 |
+
"step": 274
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 1.5034199726402189,
|
1932 |
+
"grad_norm": 1.5118176765605635,
|
1933 |
+
"learning_rate": 5.843725764846813e-06,
|
1934 |
+
"loss": 0.6836,
|
1935 |
+
"step": 275
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 1.508891928864569,
|
1939 |
+
"grad_norm": 1.3787465053311305,
|
1940 |
+
"learning_rate": 5.812175702113287e-06,
|
1941 |
+
"loss": 0.7349,
|
1942 |
+
"step": 276
|
1943 |
+
},
|
1944 |
+
{
|
1945 |
+
"epoch": 1.5143638850889194,
|
1946 |
+
"grad_norm": 1.613237980109123,
|
1947 |
+
"learning_rate": 5.7805923898680305e-06,
|
1948 |
+
"loss": 0.6599,
|
1949 |
+
"step": 277
|
1950 |
+
},
|
1951 |
+
{
|
1952 |
+
"epoch": 1.5198358413132695,
|
1953 |
+
"grad_norm": 1.5235130098212175,
|
1954 |
+
"learning_rate": 5.748977121094457e-06,
|
1955 |
+
"loss": 0.6736,
|
1956 |
+
"step": 278
|
1957 |
+
},
|
1958 |
+
{
|
1959 |
+
"epoch": 1.5253077975376197,
|
1960 |
+
"grad_norm": 1.4115555371487662,
|
1961 |
+
"learning_rate": 5.7173311900842435e-06,
|
1962 |
+
"loss": 0.6846,
|
1963 |
+
"step": 279
|
1964 |
+
},
|
1965 |
+
{
|
1966 |
+
"epoch": 1.53077975376197,
|
1967 |
+
"grad_norm": 1.3549186291909567,
|
1968 |
+
"learning_rate": 5.685655892384336e-06,
|
1969 |
+
"loss": 0.6725,
|
1970 |
+
"step": 280
|
1971 |
+
},
|
1972 |
+
{
|
1973 |
+
"epoch": 1.53625170998632,
|
1974 |
+
"grad_norm": 1.3659205105437453,
|
1975 |
+
"learning_rate": 5.653952524743928e-06,
|
1976 |
+
"loss": 0.6242,
|
1977 |
+
"step": 281
|
1978 |
+
},
|
1979 |
+
{
|
1980 |
+
"epoch": 1.5417236662106704,
|
1981 |
+
"grad_norm": 1.3967391378047187,
|
1982 |
+
"learning_rate": 5.6222223850613535e-06,
|
1983 |
+
"loss": 0.6787,
|
1984 |
+
"step": 282
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"epoch": 1.5471956224350205,
|
1988 |
+
"grad_norm": 1.3406476101907596,
|
1989 |
+
"learning_rate": 5.590466772330968e-06,
|
1990 |
+
"loss": 0.671,
|
1991 |
+
"step": 283
|
1992 |
+
},
|
1993 |
+
{
|
1994 |
+
"epoch": 1.5526675786593707,
|
1995 |
+
"grad_norm": 1.4038150192582608,
|
1996 |
+
"learning_rate": 5.5586869865899625e-06,
|
1997 |
+
"loss": 0.7263,
|
1998 |
+
"step": 284
|
1999 |
+
},
|
2000 |
+
{
|
2001 |
+
"epoch": 1.558139534883721,
|
2002 |
+
"grad_norm": 1.4528411482788122,
|
2003 |
+
"learning_rate": 5.526884328865142e-06,
|
2004 |
+
"loss": 0.669,
|
2005 |
+
"step": 285
|
2006 |
+
},
|
2007 |
+
{
|
2008 |
+
"epoch": 1.5636114911080712,
|
2009 |
+
"grad_norm": 1.4490088303566078,
|
2010 |
+
"learning_rate": 5.495060101119662e-06,
|
2011 |
+
"loss": 0.7098,
|
2012 |
+
"step": 286
|
2013 |
+
},
|
2014 |
+
{
|
2015 |
+
"epoch": 1.5690834473324213,
|
2016 |
+
"grad_norm": 1.4804781168273753,
|
2017 |
+
"learning_rate": 5.4632156061997335e-06,
|
2018 |
+
"loss": 0.7326,
|
2019 |
+
"step": 287
|
2020 |
+
},
|
2021 |
+
{
|
2022 |
+
"epoch": 1.5745554035567717,
|
2023 |
+
"grad_norm": 1.4472375050612503,
|
2024 |
+
"learning_rate": 5.431352147781275e-06,
|
2025 |
+
"loss": 0.706,
|
2026 |
+
"step": 288
|
2027 |
+
},
|
2028 |
+
{
|
2029 |
+
"epoch": 1.5800273597811216,
|
2030 |
+
"grad_norm": 1.5395517101044713,
|
2031 |
+
"learning_rate": 5.399471030316554e-06,
|
2032 |
+
"loss": 0.7053,
|
2033 |
+
"step": 289
|
2034 |
+
},
|
2035 |
+
{
|
2036 |
+
"epoch": 1.585499316005472,
|
2037 |
+
"grad_norm": 1.4669088923683866,
|
2038 |
+
"learning_rate": 5.367573558980775e-06,
|
2039 |
+
"loss": 0.6155,
|
2040 |
+
"step": 290
|
2041 |
+
},
|
2042 |
+
{
|
2043 |
+
"epoch": 1.5909712722298222,
|
2044 |
+
"grad_norm": 1.4324693068796888,
|
2045 |
+
"learning_rate": 5.335661039618654e-06,
|
2046 |
+
"loss": 0.6721,
|
2047 |
+
"step": 291
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"epoch": 1.5964432284541723,
|
2051 |
+
"grad_norm": 1.528954925681748,
|
2052 |
+
"learning_rate": 5.30373477869095e-06,
|
2053 |
+
"loss": 0.6725,
|
2054 |
+
"step": 292
|
2055 |
+
},
|
2056 |
+
{
|
2057 |
+
"epoch": 1.6019151846785227,
|
2058 |
+
"grad_norm": 1.4495214882312961,
|
2059 |
+
"learning_rate": 5.271796083220991e-06,
|
2060 |
+
"loss": 0.6443,
|
2061 |
+
"step": 293
|
2062 |
+
},
|
2063 |
+
{
|
2064 |
+
"epoch": 1.6073871409028728,
|
2065 |
+
"grad_norm": 1.4230185380662494,
|
2066 |
+
"learning_rate": 5.239846260741158e-06,
|
2067 |
+
"loss": 0.6906,
|
2068 |
+
"step": 294
|
2069 |
+
},
|
2070 |
+
{
|
2071 |
+
"epoch": 1.612859097127223,
|
2072 |
+
"grad_norm": 1.4285062698458826,
|
2073 |
+
"learning_rate": 5.2078866192393575e-06,
|
2074 |
+
"loss": 0.8223,
|
2075 |
+
"step": 295
|
2076 |
+
},
|
2077 |
+
{
|
2078 |
+
"epoch": 1.6183310533515733,
|
2079 |
+
"grad_norm": 1.33767204740181,
|
2080 |
+
"learning_rate": 5.1759184671054785e-06,
|
2081 |
+
"loss": 0.7225,
|
2082 |
+
"step": 296
|
2083 |
+
},
|
2084 |
+
{
|
2085 |
+
"epoch": 1.6238030095759233,
|
2086 |
+
"grad_norm": 1.2784052839744025,
|
2087 |
+
"learning_rate": 5.14394311307782e-06,
|
2088 |
+
"loss": 0.6211,
|
2089 |
+
"step": 297
|
2090 |
+
},
|
2091 |
+
{
|
2092 |
+
"epoch": 1.6292749658002736,
|
2093 |
+
"grad_norm": 1.4187896185437836,
|
2094 |
+
"learning_rate": 5.111961866189524e-06,
|
2095 |
+
"loss": 0.685,
|
2096 |
+
"step": 298
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"epoch": 1.6347469220246238,
|
2100 |
+
"grad_norm": 1.4964383709288,
|
2101 |
+
"learning_rate": 5.079976035714976e-06,
|
2102 |
+
"loss": 0.6506,
|
2103 |
+
"step": 299
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 1.640218878248974,
|
2107 |
+
"grad_norm": 1.305429735041424,
|
2108 |
+
"learning_rate": 5.047986931116205e-06,
|
2109 |
+
"loss": 0.7135,
|
2110 |
+
"step": 300
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"epoch": 1.6456908344733243,
|
2114 |
+
"grad_norm": 1.4712236872699123,
|
2115 |
+
"learning_rate": 5.015995861989287e-06,
|
2116 |
+
"loss": 0.6955,
|
2117 |
+
"step": 301
|
2118 |
+
},
|
2119 |
+
{
|
2120 |
+
"epoch": 1.6511627906976745,
|
2121 |
+
"grad_norm": 1.5335783688145903,
|
2122 |
+
"learning_rate": 4.984004138010715e-06,
|
2123 |
+
"loss": 0.763,
|
2124 |
+
"step": 302
|
2125 |
+
},
|
2126 |
+
{
|
2127 |
+
"epoch": 1.6566347469220246,
|
2128 |
+
"grad_norm": 1.5676763885123655,
|
2129 |
+
"learning_rate": 4.952013068883796e-06,
|
2130 |
+
"loss": 0.6975,
|
2131 |
+
"step": 303
|
2132 |
+
},
|
2133 |
+
{
|
2134 |
+
"epoch": 1.662106703146375,
|
2135 |
+
"grad_norm": 1.4693538609312076,
|
2136 |
+
"learning_rate": 4.920023964285025e-06,
|
2137 |
+
"loss": 0.7714,
|
2138 |
+
"step": 304
|
2139 |
+
},
|
2140 |
+
{
|
2141 |
+
"epoch": 1.667578659370725,
|
2142 |
+
"grad_norm": 1.4652114612995766,
|
2143 |
+
"learning_rate": 4.888038133810477e-06,
|
2144 |
+
"loss": 0.6577,
|
2145 |
+
"step": 305
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 1.6730506155950753,
|
2149 |
+
"grad_norm": 1.502179493690188,
|
2150 |
+
"learning_rate": 4.8560568869221805e-06,
|
2151 |
+
"loss": 0.7077,
|
2152 |
+
"step": 306
|
2153 |
+
},
|
2154 |
+
{
|
2155 |
+
"epoch": 1.6785225718194254,
|
2156 |
+
"grad_norm": 1.3230472540659786,
|
2157 |
+
"learning_rate": 4.824081532894523e-06,
|
2158 |
+
"loss": 0.6592,
|
2159 |
+
"step": 307
|
2160 |
+
},
|
2161 |
+
{
|
2162 |
+
"epoch": 1.6839945280437756,
|
2163 |
+
"grad_norm": 1.304860132460969,
|
2164 |
+
"learning_rate": 4.7921133807606424e-06,
|
2165 |
+
"loss": 0.6854,
|
2166 |
+
"step": 308
|
2167 |
+
},
|
2168 |
+
{
|
2169 |
+
"epoch": 1.689466484268126,
|
2170 |
+
"grad_norm": 1.462115719723299,
|
2171 |
+
"learning_rate": 4.760153739258843e-06,
|
2172 |
+
"loss": 0.6921,
|
2173 |
+
"step": 309
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"epoch": 1.694938440492476,
|
2177 |
+
"grad_norm": 1.4318713042337339,
|
2178 |
+
"learning_rate": 4.7282039167790095e-06,
|
2179 |
+
"loss": 0.6688,
|
2180 |
+
"step": 310
|
2181 |
+
},
|
2182 |
+
{
|
2183 |
+
"epoch": 1.7004103967168263,
|
2184 |
+
"grad_norm": 1.348596782624225,
|
2185 |
+
"learning_rate": 4.69626522130905e-06,
|
2186 |
+
"loss": 0.7417,
|
2187 |
+
"step": 311
|
2188 |
+
},
|
2189 |
+
{
|
2190 |
+
"epoch": 1.7058823529411766,
|
2191 |
+
"grad_norm": 1.4809205032090513,
|
2192 |
+
"learning_rate": 4.664338960381349e-06,
|
2193 |
+
"loss": 0.6535,
|
2194 |
+
"step": 312
|
2195 |
+
},
|
2196 |
+
{
|
2197 |
+
"epoch": 1.7113543091655266,
|
2198 |
+
"grad_norm": 1.4532723138483248,
|
2199 |
+
"learning_rate": 4.632426441019226e-06,
|
2200 |
+
"loss": 0.7071,
|
2201 |
+
"step": 313
|
2202 |
+
},
|
2203 |
+
{
|
2204 |
+
"epoch": 1.716826265389877,
|
2205 |
+
"grad_norm": 1.1594837748428621,
|
2206 |
+
"learning_rate": 4.600528969683448e-06,
|
2207 |
+
"loss": 0.6792,
|
2208 |
+
"step": 314
|
2209 |
+
},
|
2210 |
+
{
|
2211 |
+
"epoch": 1.722298221614227,
|
2212 |
+
"grad_norm": 1.3767390209960755,
|
2213 |
+
"learning_rate": 4.568647852218725e-06,
|
2214 |
+
"loss": 0.645,
|
2215 |
+
"step": 315
|
2216 |
+
},
|
2217 |
+
{
|
2218 |
+
"epoch": 1.7277701778385772,
|
2219 |
+
"grad_norm": 1.3348338030946811,
|
2220 |
+
"learning_rate": 4.536784393800269e-06,
|
2221 |
+
"loss": 0.6749,
|
2222 |
+
"step": 316
|
2223 |
+
},
|
2224 |
+
{
|
2225 |
+
"epoch": 1.7332421340629276,
|
2226 |
+
"grad_norm": 1.3561234439417202,
|
2227 |
+
"learning_rate": 4.5049398988803395e-06,
|
2228 |
+
"loss": 0.6504,
|
2229 |
+
"step": 317
|
2230 |
+
},
|
2231 |
+
{
|
2232 |
+
"epoch": 1.7387140902872777,
|
2233 |
+
"grad_norm": 1.3932476124764381,
|
2234 |
+
"learning_rate": 4.47311567113486e-06,
|
2235 |
+
"loss": 0.7435,
|
2236 |
+
"step": 318
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"epoch": 1.744186046511628,
|
2240 |
+
"grad_norm": 1.447005955829397,
|
2241 |
+
"learning_rate": 4.441313013410039e-06,
|
2242 |
+
"loss": 0.6163,
|
2243 |
+
"step": 319
|
2244 |
+
},
|
2245 |
+
{
|
2246 |
+
"epoch": 1.7496580027359783,
|
2247 |
+
"grad_norm": 1.2947118154835413,
|
2248 |
+
"learning_rate": 4.4095332276690334e-06,
|
2249 |
+
"loss": 0.6797,
|
2250 |
+
"step": 320
|
2251 |
+
},
|
2252 |
+
{
|
2253 |
+
"epoch": 1.7551299589603282,
|
2254 |
+
"grad_norm": 1.4650842816680685,
|
2255 |
+
"learning_rate": 4.377777614938647e-06,
|
2256 |
+
"loss": 0.6554,
|
2257 |
+
"step": 321
|
2258 |
+
},
|
2259 |
+
{
|
2260 |
+
"epoch": 1.7606019151846786,
|
2261 |
+
"grad_norm": 1.313113246375332,
|
2262 |
+
"learning_rate": 4.346047475256072e-06,
|
2263 |
+
"loss": 0.6534,
|
2264 |
+
"step": 322
|
2265 |
+
},
|
2266 |
+
{
|
2267 |
+
"epoch": 1.7660738714090287,
|
2268 |
+
"grad_norm": 1.411771456085658,
|
2269 |
+
"learning_rate": 4.314344107615665e-06,
|
2270 |
+
"loss": 0.677,
|
2271 |
+
"step": 323
|
2272 |
+
},
|
2273 |
+
{
|
2274 |
+
"epoch": 1.7715458276333789,
|
2275 |
+
"grad_norm": 1.4978660719708754,
|
2276 |
+
"learning_rate": 4.282668809915758e-06,
|
2277 |
+
"loss": 0.7344,
|
2278 |
+
"step": 324
|
2279 |
+
},
|
2280 |
+
{
|
2281 |
+
"epoch": 1.7770177838577292,
|
2282 |
+
"grad_norm": 1.4125423590869164,
|
2283 |
+
"learning_rate": 4.251022878905543e-06,
|
2284 |
+
"loss": 0.6741,
|
2285 |
+
"step": 325
|
2286 |
+
},
|
2287 |
+
{
|
2288 |
+
"epoch": 1.7824897400820794,
|
2289 |
+
"grad_norm": 1.3866460898122654,
|
2290 |
+
"learning_rate": 4.219407610131971e-06,
|
2291 |
+
"loss": 0.6386,
|
2292 |
+
"step": 326
|
2293 |
+
},
|
2294 |
+
{
|
2295 |
+
"epoch": 1.7879616963064295,
|
2296 |
+
"grad_norm": 1.2523699658732943,
|
2297 |
+
"learning_rate": 4.187824297886715e-06,
|
2298 |
+
"loss": 0.6514,
|
2299 |
+
"step": 327
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"epoch": 1.79343365253078,
|
2303 |
+
"grad_norm": 1.3712795790991985,
|
2304 |
+
"learning_rate": 4.1562742351531895e-06,
|
2305 |
+
"loss": 0.754,
|
2306 |
+
"step": 328
|
2307 |
+
},
|
2308 |
+
{
|
2309 |
+
"epoch": 1.7989056087551298,
|
2310 |
+
"grad_norm": 1.4094480016901794,
|
2311 |
+
"learning_rate": 4.124758713553613e-06,
|
2312 |
+
"loss": 0.6838,
|
2313 |
+
"step": 329
|
2314 |
+
},
|
2315 |
+
{
|
2316 |
+
"epoch": 1.8043775649794802,
|
2317 |
+
"grad_norm": 1.4084554057675314,
|
2318 |
+
"learning_rate": 4.093279023296124e-06,
|
2319 |
+
"loss": 0.7424,
|
2320 |
+
"step": 330
|
2321 |
+
},
|
2322 |
+
{
|
2323 |
+
"epoch": 1.8098495212038304,
|
2324 |
+
"grad_norm": 1.5841251676153847,
|
2325 |
+
"learning_rate": 4.061836453121978e-06,
|
2326 |
+
"loss": 0.687,
|
2327 |
+
"step": 331
|
2328 |
+
},
|
2329 |
+
{
|
2330 |
+
"epoch": 1.8153214774281805,
|
2331 |
+
"grad_norm": 1.4612003859739513,
|
2332 |
+
"learning_rate": 4.030432290252771e-06,
|
2333 |
+
"loss": 0.713,
|
2334 |
+
"step": 332
|
2335 |
+
},
|
2336 |
+
{
|
2337 |
+
"epoch": 1.8207934336525309,
|
2338 |
+
"grad_norm": 1.5244348310397957,
|
2339 |
+
"learning_rate": 3.999067820337757e-06,
|
2340 |
+
"loss": 0.7235,
|
2341 |
+
"step": 333
|
2342 |
+
},
|
2343 |
+
{
|
2344 |
+
"epoch": 1.826265389876881,
|
2345 |
+
"grad_norm": 1.4567943380833739,
|
2346 |
+
"learning_rate": 3.967744327401198e-06,
|
2347 |
+
"loss": 0.7038,
|
2348 |
+
"step": 334
|
2349 |
+
},
|
2350 |
+
{
|
2351 |
+
"epoch": 1.8317373461012312,
|
2352 |
+
"grad_norm": 1.2829318258042766,
|
2353 |
+
"learning_rate": 3.936463093789813e-06,
|
2354 |
+
"loss": 0.6626,
|
2355 |
+
"step": 335
|
2356 |
+
},
|
2357 |
+
{
|
2358 |
+
"epoch": 1.8372093023255816,
|
2359 |
+
"grad_norm": 1.4181999963388858,
|
2360 |
+
"learning_rate": 3.9052254001202746e-06,
|
2361 |
+
"loss": 0.6821,
|
2362 |
+
"step": 336
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"epoch": 1.8426812585499315,
|
2366 |
+
"grad_norm": 1.2877791458716707,
|
2367 |
+
"learning_rate": 3.8740325252267785e-06,
|
2368 |
+
"loss": 0.6946,
|
2369 |
+
"step": 337
|
2370 |
+
},
|
2371 |
+
{
|
2372 |
+
"epoch": 1.8481532147742818,
|
2373 |
+
"grad_norm": 1.2942675340827412,
|
2374 |
+
"learning_rate": 3.842885746108694e-06,
|
2375 |
+
"loss": 0.5989,
|
2376 |
+
"step": 338
|
2377 |
+
},
|
2378 |
+
{
|
2379 |
+
"epoch": 1.853625170998632,
|
2380 |
+
"grad_norm": 1.3900599897561317,
|
2381 |
+
"learning_rate": 3.811786337878284e-06,
|
2382 |
+
"loss": 0.6821,
|
2383 |
+
"step": 339
|
2384 |
+
},
|
2385 |
+
{
|
2386 |
+
"epoch": 1.8590971272229821,
|
2387 |
+
"grad_norm": 1.5973639237132788,
|
2388 |
+
"learning_rate": 3.7807355737085077e-06,
|
2389 |
+
"loss": 0.6773,
|
2390 |
+
"step": 340
|
2391 |
+
},
|
2392 |
+
{
|
2393 |
+
"epoch": 1.8645690834473325,
|
2394 |
+
"grad_norm": 1.4928529764561793,
|
2395 |
+
"learning_rate": 3.749734724780885e-06,
|
2396 |
+
"loss": 0.7959,
|
2397 |
+
"step": 341
|
2398 |
+
},
|
2399 |
+
{
|
2400 |
+
"epoch": 1.8700410396716827,
|
2401 |
+
"grad_norm": 1.3479589392528368,
|
2402 |
+
"learning_rate": 3.7187850602334714e-06,
|
2403 |
+
"loss": 0.6264,
|
2404 |
+
"step": 342
|
2405 |
+
},
|
2406 |
+
{
|
2407 |
+
"epoch": 1.8755129958960328,
|
2408 |
+
"grad_norm": 1.1969994969070834,
|
2409 |
+
"learning_rate": 3.687887847108894e-06,
|
2410 |
+
"loss": 0.6781,
|
2411 |
+
"step": 343
|
2412 |
+
},
|
2413 |
+
{
|
2414 |
+
"epoch": 1.8809849521203832,
|
2415 |
+
"grad_norm": 1.347062435341254,
|
2416 |
+
"learning_rate": 3.657044350302479e-06,
|
2417 |
+
"loss": 0.662,
|
2418 |
+
"step": 344
|
2419 |
+
},
|
2420 |
+
{
|
2421 |
+
"epoch": 1.8864569083447331,
|
2422 |
+
"grad_norm": 1.3017687857340479,
|
2423 |
+
"learning_rate": 3.6262558325104697e-06,
|
2424 |
+
"loss": 0.6646,
|
2425 |
+
"step": 345
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"epoch": 1.8919288645690835,
|
2429 |
+
"grad_norm": 1.5082475864549976,
|
2430 |
+
"learning_rate": 3.5955235541783362e-06,
|
2431 |
+
"loss": 0.7089,
|
2432 |
+
"step": 346
|
2433 |
+
},
|
2434 |
+
{
|
2435 |
+
"epoch": 1.8974008207934336,
|
2436 |
+
"grad_norm": 1.4236562957628613,
|
2437 |
+
"learning_rate": 3.564848773449172e-06,
|
2438 |
+
"loss": 0.7303,
|
2439 |
+
"step": 347
|
2440 |
+
},
|
2441 |
+
{
|
2442 |
+
"epoch": 1.9028727770177838,
|
2443 |
+
"grad_norm": 1.512232444395897,
|
2444 |
+
"learning_rate": 3.5342327461121806e-06,
|
2445 |
+
"loss": 0.6073,
|
2446 |
+
"step": 348
|
2447 |
+
},
|
2448 |
+
{
|
2449 |
+
"epoch": 1.9083447332421342,
|
2450 |
+
"grad_norm": 1.362318777020872,
|
2451 |
+
"learning_rate": 3.5036767255512784e-06,
|
2452 |
+
"loss": 0.6606,
|
2453 |
+
"step": 349
|
2454 |
+
},
|
2455 |
+
{
|
2456 |
+
"epoch": 1.9138166894664843,
|
2457 |
+
"grad_norm": 1.406483908993142,
|
2458 |
+
"learning_rate": 3.473181962693773e-06,
|
2459 |
+
"loss": 0.7171,
|
2460 |
+
"step": 350
|
2461 |
+
},
|
2462 |
+
{
|
2463 |
+
"epoch": 1.9192886456908345,
|
2464 |
+
"grad_norm": 1.5780917295423231,
|
2465 |
+
"learning_rate": 3.442749705959152e-06,
|
2466 |
+
"loss": 0.6292,
|
2467 |
+
"step": 351
|
2468 |
+
},
|
2469 |
+
{
|
2470 |
+
"epoch": 1.9247606019151848,
|
2471 |
+
"grad_norm": 1.5079580554009018,
|
2472 |
+
"learning_rate": 3.412381201207979e-06,
|
2473 |
+
"loss": 0.6995,
|
2474 |
+
"step": 352
|
2475 |
+
},
|
2476 |
+
{
|
2477 |
+
"epoch": 1.9302325581395348,
|
2478 |
+
"grad_norm": 1.4509521426633267,
|
2479 |
+
"learning_rate": 3.382077691690886e-06,
|
2480 |
+
"loss": 0.7553,
|
2481 |
+
"step": 353
|
2482 |
+
},
|
2483 |
+
{
|
2484 |
+
"epoch": 1.9357045143638851,
|
2485 |
+
"grad_norm": 1.4321730201024216,
|
2486 |
+
"learning_rate": 3.3518404179976792e-06,
|
2487 |
+
"loss": 0.6731,
|
2488 |
+
"step": 354
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"epoch": 1.9411764705882353,
|
2492 |
+
"grad_norm": 1.3617282268861342,
|
2493 |
+
"learning_rate": 3.3216706180065427e-06,
|
2494 |
+
"loss": 0.7232,
|
2495 |
+
"step": 355
|
2496 |
+
},
|
2497 |
+
{
|
2498 |
+
"epoch": 1.9466484268125854,
|
2499 |
+
"grad_norm": 1.440329600791613,
|
2500 |
+
"learning_rate": 3.2915695268333726e-06,
|
2501 |
+
"loss": 0.6394,
|
2502 |
+
"step": 356
|
2503 |
+
},
|
2504 |
+
{
|
2505 |
+
"epoch": 1.9521203830369358,
|
2506 |
+
"grad_norm": 1.4813547878564215,
|
2507 |
+
"learning_rate": 3.2615383767812057e-06,
|
2508 |
+
"loss": 0.7413,
|
2509 |
+
"step": 357
|
2510 |
+
},
|
2511 |
+
{
|
2512 |
+
"epoch": 1.957592339261286,
|
2513 |
+
"grad_norm": 1.474791843511977,
|
2514 |
+
"learning_rate": 3.2315783972897717e-06,
|
2515 |
+
"loss": 0.6886,
|
2516 |
+
"step": 358
|
2517 |
+
},
|
2518 |
+
{
|
2519 |
+
"epoch": 1.963064295485636,
|
2520 |
+
"grad_norm": 1.446494422047024,
|
2521 |
+
"learning_rate": 3.2016908148851623e-06,
|
2522 |
+
"loss": 0.7419,
|
2523 |
+
"step": 359
|
2524 |
+
},
|
2525 |
+
{
|
2526 |
+
"epoch": 1.9685362517099865,
|
2527 |
+
"grad_norm": 1.6088902739669952,
|
2528 |
+
"learning_rate": 3.1718768531296196e-06,
|
2529 |
+
"loss": 0.7098,
|
2530 |
+
"step": 360
|
2531 |
+
},
|
2532 |
+
{
|
2533 |
+
"epoch": 1.9740082079343364,
|
2534 |
+
"grad_norm": 1.3453147029282566,
|
2535 |
+
"learning_rate": 3.1421377325714377e-06,
|
2536 |
+
"loss": 0.7467,
|
2537 |
+
"step": 361
|
2538 |
+
},
|
2539 |
+
{
|
2540 |
+
"epoch": 1.9794801641586868,
|
2541 |
+
"grad_norm": 1.5538504536249995,
|
2542 |
+
"learning_rate": 3.1124746706950084e-06,
|
2543 |
+
"loss": 0.6564,
|
2544 |
+
"step": 362
|
2545 |
+
},
|
2546 |
+
{
|
2547 |
+
"epoch": 1.984952120383037,
|
2548 |
+
"grad_norm": 1.3499847068940665,
|
2549 |
+
"learning_rate": 3.082888881870966e-06,
|
2550 |
+
"loss": 0.7629,
|
2551 |
+
"step": 363
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"epoch": 1.990424076607387,
|
2555 |
+
"grad_norm": 1.5075584552181835,
|
2556 |
+
"learning_rate": 3.0533815773064813e-06,
|
2557 |
+
"loss": 0.7057,
|
2558 |
+
"step": 364
|
2559 |
+
},
|
2560 |
+
{
|
2561 |
+
"epoch": 1.9958960328317374,
|
2562 |
+
"grad_norm": 1.3795911709840523,
|
2563 |
+
"learning_rate": 3.0239539649956665e-06,
|
2564 |
+
"loss": 0.7058,
|
2565 |
+
"step": 365
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 2.0,
|
2569 |
+
"grad_norm": 1.3703346650064345,
|
2570 |
+
"learning_rate": 2.9946072496701335e-06,
|
2571 |
+
"loss": 0.683,
|
2572 |
+
"step": 366
|
2573 |
+
},
|
2574 |
+
{
|
2575 |
+
"epoch": 2.0054719562243504,
|
2576 |
+
"grad_norm": 1.6357532362837628,
|
2577 |
+
"learning_rate": 2.9653426327496647e-06,
|
2578 |
+
"loss": 0.523,
|
2579 |
+
"step": 367
|
2580 |
+
},
|
2581 |
+
{
|
2582 |
+
"epoch": 2.0109439124487003,
|
2583 |
+
"grad_norm": 1.502628095771462,
|
2584 |
+
"learning_rate": 2.9361613122930304e-06,
|
2585 |
+
"loss": 0.5113,
|
2586 |
+
"step": 368
|
2587 |
+
},
|
2588 |
+
{
|
2589 |
+
"epoch": 2.0164158686730507,
|
2590 |
+
"grad_norm": 1.4538602295702325,
|
2591 |
+
"learning_rate": 2.9070644829489435e-06,
|
2592 |
+
"loss": 0.5389,
|
2593 |
+
"step": 369
|
2594 |
+
},
|
2595 |
+
{
|
2596 |
+
"epoch": 2.0218878248974006,
|
2597 |
+
"grad_norm": 1.3589285640466358,
|
2598 |
+
"learning_rate": 2.8780533359071505e-06,
|
2599 |
+
"loss": 0.5501,
|
2600 |
+
"step": 370
|
2601 |
+
},
|
2602 |
+
{
|
2603 |
+
"epoch": 2.027359781121751,
|
2604 |
+
"grad_norm": 1.4766865256778057,
|
2605 |
+
"learning_rate": 2.849129058849667e-06,
|
2606 |
+
"loss": 0.4893,
|
2607 |
+
"step": 371
|
2608 |
+
},
|
2609 |
+
{
|
2610 |
+
"epoch": 2.0328317373461013,
|
2611 |
+
"grad_norm": 1.3963959744059982,
|
2612 |
+
"learning_rate": 2.820292835902148e-06,
|
2613 |
+
"loss": 0.4969,
|
2614 |
+
"step": 372
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"epoch": 2.0383036935704513,
|
2618 |
+
"grad_norm": 1.2641835531351906,
|
2619 |
+
"learning_rate": 2.7915458475854284e-06,
|
2620 |
+
"loss": 0.4559,
|
2621 |
+
"step": 373
|
2622 |
+
},
|
2623 |
+
{
|
2624 |
+
"epoch": 2.0437756497948016,
|
2625 |
+
"grad_norm": 1.3368368693929367,
|
2626 |
+
"learning_rate": 2.762889270767175e-06,
|
2627 |
+
"loss": 0.5048,
|
2628 |
+
"step": 374
|
2629 |
+
},
|
2630 |
+
{
|
2631 |
+
"epoch": 2.049247606019152,
|
2632 |
+
"grad_norm": 1.3851893174817642,
|
2633 |
+
"learning_rate": 2.734324278613717e-06,
|
2634 |
+
"loss": 0.5823,
|
2635 |
+
"step": 375
|
2636 |
+
},
|
2637 |
+
{
|
2638 |
+
"epoch": 2.054719562243502,
|
2639 |
+
"grad_norm": 1.5747937453896854,
|
2640 |
+
"learning_rate": 2.7058520405420125e-06,
|
2641 |
+
"loss": 0.4977,
|
2642 |
+
"step": 376
|
2643 |
+
},
|
2644 |
+
{
|
2645 |
+
"epoch": 2.0601915184678523,
|
2646 |
+
"grad_norm": 1.4946196081751006,
|
2647 |
+
"learning_rate": 2.6774737221717863e-06,
|
2648 |
+
"loss": 0.4477,
|
2649 |
+
"step": 377
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 2.0656634746922027,
|
2653 |
+
"grad_norm": 1.4792712608077758,
|
2654 |
+
"learning_rate": 2.649190485277792e-06,
|
2655 |
+
"loss": 0.5448,
|
2656 |
+
"step": 378
|
2657 |
+
},
|
2658 |
+
{
|
2659 |
+
"epoch": 2.0711354309165526,
|
2660 |
+
"grad_norm": 1.4162952462270026,
|
2661 |
+
"learning_rate": 2.621003487742264e-06,
|
2662 |
+
"loss": 0.508,
|
2663 |
+
"step": 379
|
2664 |
+
},
|
2665 |
+
{
|
2666 |
+
"epoch": 2.076607387140903,
|
2667 |
+
"grad_norm": 1.3639753959051746,
|
2668 |
+
"learning_rate": 2.5929138835075153e-06,
|
2669 |
+
"loss": 0.4678,
|
2670 |
+
"step": 380
|
2671 |
+
},
|
2672 |
+
{
|
2673 |
+
"epoch": 2.082079343365253,
|
2674 |
+
"grad_norm": 1.4161912239857386,
|
2675 |
+
"learning_rate": 2.564922822528686e-06,
|
2676 |
+
"loss": 0.5316,
|
2677 |
+
"step": 381
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"epoch": 2.0875512995896033,
|
2681 |
+
"grad_norm": 1.481947355248839,
|
2682 |
+
"learning_rate": 2.537031450726676e-06,
|
2683 |
+
"loss": 0.5277,
|
2684 |
+
"step": 382
|
2685 |
+
},
|
2686 |
+
{
|
2687 |
+
"epoch": 2.0930232558139537,
|
2688 |
+
"grad_norm": 1.4771709370714692,
|
2689 |
+
"learning_rate": 2.509240909941223e-06,
|
2690 |
+
"loss": 0.5382,
|
2691 |
+
"step": 383
|
2692 |
+
},
|
2693 |
+
{
|
2694 |
+
"epoch": 2.0984952120383036,
|
2695 |
+
"grad_norm": 1.5329176679666325,
|
2696 |
+
"learning_rate": 2.4815523378841727e-06,
|
2697 |
+
"loss": 0.5502,
|
2698 |
+
"step": 384
|
2699 |
+
},
|
2700 |
+
{
|
2701 |
+
"epoch": 2.103967168262654,
|
2702 |
+
"grad_norm": 1.5150629673619693,
|
2703 |
+
"learning_rate": 2.4539668680928787e-06,
|
2704 |
+
"loss": 0.5108,
|
2705 |
+
"step": 385
|
2706 |
+
},
|
2707 |
+
{
|
2708 |
+
"epoch": 2.109439124487004,
|
2709 |
+
"grad_norm": 1.5208241763375903,
|
2710 |
+
"learning_rate": 2.426485629883821e-06,
|
2711 |
+
"loss": 0.4766,
|
2712 |
+
"step": 386
|
2713 |
+
},
|
2714 |
+
{
|
2715 |
+
"epoch": 2.1149110807113543,
|
2716 |
+
"grad_norm": 1.4310260812775966,
|
2717 |
+
"learning_rate": 2.399109748306355e-06,
|
2718 |
+
"loss": 0.4853,
|
2719 |
+
"step": 387
|
2720 |
+
},
|
2721 |
+
{
|
2722 |
+
"epoch": 2.1203830369357046,
|
2723 |
+
"grad_norm": 1.4787425876059686,
|
2724 |
+
"learning_rate": 2.3718403440966652e-06,
|
2725 |
+
"loss": 0.4601,
|
2726 |
+
"step": 388
|
2727 |
+
},
|
2728 |
+
{
|
2729 |
+
"epoch": 2.1258549931600546,
|
2730 |
+
"grad_norm": 1.3545092606077802,
|
2731 |
+
"learning_rate": 2.3446785336318757e-06,
|
2732 |
+
"loss": 0.5389,
|
2733 |
+
"step": 389
|
2734 |
+
},
|
2735 |
+
{
|
2736 |
+
"epoch": 2.131326949384405,
|
2737 |
+
"grad_norm": 1.5304444526317944,
|
2738 |
+
"learning_rate": 2.317625428884348e-06,
|
2739 |
+
"loss": 0.4517,
|
2740 |
+
"step": 390
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"epoch": 2.1367989056087553,
|
2744 |
+
"grad_norm": 1.3461980316273192,
|
2745 |
+
"learning_rate": 2.290682137376169e-06,
|
2746 |
+
"loss": 0.4634,
|
2747 |
+
"step": 391
|
2748 |
+
},
|
2749 |
+
{
|
2750 |
+
"epoch": 2.1422708618331052,
|
2751 |
+
"grad_norm": 1.4764411359427427,
|
2752 |
+
"learning_rate": 2.263849762133788e-06,
|
2753 |
+
"loss": 0.4511,
|
2754 |
+
"step": 392
|
2755 |
+
},
|
2756 |
+
{
|
2757 |
+
"epoch": 2.1477428180574556,
|
2758 |
+
"grad_norm": 1.444181275744001,
|
2759 |
+
"learning_rate": 2.237129401642887e-06,
|
2760 |
+
"loss": 0.5061,
|
2761 |
+
"step": 393
|
2762 |
+
},
|
2763 |
+
{
|
2764 |
+
"epoch": 2.153214774281806,
|
2765 |
+
"grad_norm": 1.371372879730552,
|
2766 |
+
"learning_rate": 2.2105221498033862e-06,
|
2767 |
+
"loss": 0.5527,
|
2768 |
+
"step": 394
|
2769 |
+
},
|
2770 |
+
{
|
2771 |
+
"epoch": 2.158686730506156,
|
2772 |
+
"grad_norm": 1.3868006606816583,
|
2773 |
+
"learning_rate": 2.1840290958846814e-06,
|
2774 |
+
"loss": 0.4913,
|
2775 |
+
"step": 395
|
2776 |
+
},
|
2777 |
+
{
|
2778 |
+
"epoch": 2.1641586867305063,
|
2779 |
+
"grad_norm": 1.4551977543316907,
|
2780 |
+
"learning_rate": 2.157651324481033e-06,
|
2781 |
+
"loss": 0.5053,
|
2782 |
+
"step": 396
|
2783 |
+
},
|
2784 |
+
{
|
2785 |
+
"epoch": 2.169630642954856,
|
2786 |
+
"grad_norm": 1.5398733854530477,
|
2787 |
+
"learning_rate": 2.131389915467173e-06,
|
2788 |
+
"loss": 0.5055,
|
2789 |
+
"step": 397
|
2790 |
+
},
|
2791 |
+
{
|
2792 |
+
"epoch": 2.1751025991792066,
|
2793 |
+
"grad_norm": 1.6200400551425358,
|
2794 |
+
"learning_rate": 2.1052459439541005e-06,
|
2795 |
+
"loss": 0.541,
|
2796 |
+
"step": 398
|
2797 |
+
},
|
2798 |
+
{
|
2799 |
+
"epoch": 2.180574555403557,
|
2800 |
+
"grad_norm": 1.4966929125707025,
|
2801 |
+
"learning_rate": 2.079220480245052e-06,
|
2802 |
+
"loss": 0.4877,
|
2803 |
+
"step": 399
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"epoch": 2.186046511627907,
|
2807 |
+
"grad_norm": 1.502524757017655,
|
2808 |
+
"learning_rate": 2.053314589791706e-06,
|
2809 |
+
"loss": 0.4802,
|
2810 |
+
"step": 400
|
2811 |
+
},
|
2812 |
+
{
|
2813 |
+
"epoch": 2.1915184678522572,
|
2814 |
+
"grad_norm": 1.5271740680821235,
|
2815 |
+
"learning_rate": 2.0275293331505437e-06,
|
2816 |
+
"loss": 0.489,
|
2817 |
+
"step": 401
|
2818 |
+
},
|
2819 |
+
{
|
2820 |
+
"epoch": 2.196990424076607,
|
2821 |
+
"grad_norm": 1.4558483682081753,
|
2822 |
+
"learning_rate": 2.00186576593945e-06,
|
2823 |
+
"loss": 0.5028,
|
2824 |
+
"step": 402
|
2825 |
+
},
|
2826 |
+
{
|
2827 |
+
"epoch": 2.2024623803009575,
|
2828 |
+
"grad_norm": 1.4554104689321878,
|
2829 |
+
"learning_rate": 1.9763249387944823e-06,
|
2830 |
+
"loss": 0.4967,
|
2831 |
+
"step": 403
|
2832 |
+
},
|
2833 |
+
{
|
2834 |
+
"epoch": 2.207934336525308,
|
2835 |
+
"grad_norm": 1.5790426238619297,
|
2836 |
+
"learning_rate": 1.9509078973268645e-06,
|
2837 |
+
"loss": 0.4922,
|
2838 |
+
"step": 404
|
2839 |
+
},
|
2840 |
+
{
|
2841 |
+
"epoch": 2.213406292749658,
|
2842 |
+
"grad_norm": 1.5008000283932275,
|
2843 |
+
"learning_rate": 1.9256156820801896e-06,
|
2844 |
+
"loss": 0.4575,
|
2845 |
+
"step": 405
|
2846 |
+
},
|
2847 |
+
{
|
2848 |
+
"epoch": 2.218878248974008,
|
2849 |
+
"grad_norm": 1.3473591034390053,
|
2850 |
+
"learning_rate": 1.9004493284877995e-06,
|
2851 |
+
"loss": 0.4652,
|
2852 |
+
"step": 406
|
2853 |
+
},
|
2854 |
+
{
|
2855 |
+
"epoch": 2.2243502051983586,
|
2856 |
+
"grad_norm": 1.3973493162237776,
|
2857 |
+
"learning_rate": 1.875409866830422e-06,
|
2858 |
+
"loss": 0.4066,
|
2859 |
+
"step": 407
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 2.2298221614227085,
|
2863 |
+
"grad_norm": 1.3301360078296918,
|
2864 |
+
"learning_rate": 1.850498322193972e-06,
|
2865 |
+
"loss": 0.4605,
|
2866 |
+
"step": 408
|
2867 |
+
},
|
2868 |
+
{
|
2869 |
+
"epoch": 2.235294117647059,
|
2870 |
+
"grad_norm": 1.4794345369785276,
|
2871 |
+
"learning_rate": 1.8257157144275939e-06,
|
2872 |
+
"loss": 0.4854,
|
2873 |
+
"step": 409
|
2874 |
+
},
|
2875 |
+
{
|
2876 |
+
"epoch": 2.2407660738714092,
|
2877 |
+
"grad_norm": 1.4936239899544665,
|
2878 |
+
"learning_rate": 1.8010630581019095e-06,
|
2879 |
+
"loss": 0.531,
|
2880 |
+
"step": 410
|
2881 |
+
},
|
2882 |
+
{
|
2883 |
+
"epoch": 2.246238030095759,
|
2884 |
+
"grad_norm": 1.4869150371752362,
|
2885 |
+
"learning_rate": 1.7765413624674866e-06,
|
2886 |
+
"loss": 0.5214,
|
2887 |
+
"step": 411
|
2888 |
+
},
|
2889 |
+
{
|
2890 |
+
"epoch": 2.2517099863201095,
|
2891 |
+
"grad_norm": 1.3739655876723418,
|
2892 |
+
"learning_rate": 1.7521516314135112e-06,
|
2893 |
+
"loss": 0.5112,
|
2894 |
+
"step": 412
|
2895 |
+
},
|
2896 |
+
{
|
2897 |
+
"epoch": 2.2571819425444595,
|
2898 |
+
"grad_norm": 1.5375846536669437,
|
2899 |
+
"learning_rate": 1.7278948634266968e-06,
|
2900 |
+
"loss": 0.5079,
|
2901 |
+
"step": 413
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 2.26265389876881,
|
2905 |
+
"grad_norm": 1.386235503083451,
|
2906 |
+
"learning_rate": 1.703772051550412e-06,
|
2907 |
+
"loss": 0.4258,
|
2908 |
+
"step": 414
|
2909 |
+
},
|
2910 |
+
{
|
2911 |
+
"epoch": 2.26812585499316,
|
2912 |
+
"grad_norm": 1.3545075405162734,
|
2913 |
+
"learning_rate": 1.6797841833440143e-06,
|
2914 |
+
"loss": 0.5307,
|
2915 |
+
"step": 415
|
2916 |
+
},
|
2917 |
+
{
|
2918 |
+
"epoch": 2.27359781121751,
|
2919 |
+
"grad_norm": 1.3483065041740272,
|
2920 |
+
"learning_rate": 1.6559322408424289e-06,
|
2921 |
+
"loss": 0.5173,
|
2922 |
+
"step": 416
|
2923 |
+
},
|
2924 |
+
{
|
2925 |
+
"epoch": 2.2790697674418605,
|
2926 |
+
"grad_norm": 1.521051247225777,
|
2927 |
+
"learning_rate": 1.6322172005159436e-06,
|
2928 |
+
"loss": 0.4889,
|
2929 |
+
"step": 417
|
2930 |
+
},
|
2931 |
+
{
|
2932 |
+
"epoch": 2.2845417236662104,
|
2933 |
+
"grad_norm": 1.3522836876430269,
|
2934 |
+
"learning_rate": 1.608640033230236e-06,
|
2935 |
+
"loss": 0.4621,
|
2936 |
+
"step": 418
|
2937 |
+
},
|
2938 |
+
{
|
2939 |
+
"epoch": 2.290013679890561,
|
2940 |
+
"grad_norm": 1.4484864881402804,
|
2941 |
+
"learning_rate": 1.5852017042066215e-06,
|
2942 |
+
"loss": 0.5532,
|
2943 |
+
"step": 419
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 2.295485636114911,
|
2947 |
+
"grad_norm": 1.4081058445468693,
|
2948 |
+
"learning_rate": 1.5619031729825401e-06,
|
2949 |
+
"loss": 0.707,
|
2950 |
+
"step": 420
|
2951 |
+
},
|
2952 |
+
{
|
2953 |
+
"epoch": 2.300957592339261,
|
2954 |
+
"grad_norm": 1.5080398469053937,
|
2955 |
+
"learning_rate": 1.5387453933722811e-06,
|
2956 |
+
"loss": 0.4628,
|
2957 |
+
"step": 421
|
2958 |
+
},
|
2959 |
+
{
|
2960 |
+
"epoch": 2.3064295485636115,
|
2961 |
+
"grad_norm": 1.4547079873600008,
|
2962 |
+
"learning_rate": 1.5157293134279243e-06,
|
2963 |
+
"loss": 0.5053,
|
2964 |
+
"step": 422
|
2965 |
+
},
|
2966 |
+
{
|
2967 |
+
"epoch": 2.311901504787962,
|
2968 |
+
"grad_norm": 1.45150277096092,
|
2969 |
+
"learning_rate": 1.492855875400534e-06,
|
2970 |
+
"loss": 0.4868,
|
2971 |
+
"step": 423
|
2972 |
+
},
|
2973 |
+
{
|
2974 |
+
"epoch": 2.317373461012312,
|
2975 |
+
"grad_norm": 1.4435098637476864,
|
2976 |
+
"learning_rate": 1.4701260157015823e-06,
|
2977 |
+
"loss": 0.5358,
|
2978 |
+
"step": 424
|
2979 |
+
},
|
2980 |
+
{
|
2981 |
+
"epoch": 2.322845417236662,
|
2982 |
+
"grad_norm": 1.5084808421509528,
|
2983 |
+
"learning_rate": 1.4475406648646195e-06,
|
2984 |
+
"loss": 0.4804,
|
2985 |
+
"step": 425
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 2.3283173734610125,
|
2989 |
+
"grad_norm": 1.442906701799641,
|
2990 |
+
"learning_rate": 1.4251007475071688e-06,
|
2991 |
+
"loss": 0.5275,
|
2992 |
+
"step": 426
|
2993 |
+
},
|
2994 |
+
{
|
2995 |
+
"epoch": 2.3337893296853625,
|
2996 |
+
"grad_norm": 1.339812450356771,
|
2997 |
+
"learning_rate": 1.402807182292879e-06,
|
2998 |
+
"loss": 0.4416,
|
2999 |
+
"step": 427
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 2.339261285909713,
|
3003 |
+
"grad_norm": 1.3780773179590375,
|
3004 |
+
"learning_rate": 1.3806608818939205e-06,
|
3005 |
+
"loss": 0.5521,
|
3006 |
+
"step": 428
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 2.3447332421340628,
|
3010 |
+
"grad_norm": 1.2755257471180248,
|
3011 |
+
"learning_rate": 1.3586627529536117e-06,
|
3012 |
+
"loss": 0.5347,
|
3013 |
+
"step": 429
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 2.350205198358413,
|
3017 |
+
"grad_norm": 1.303404246753018,
|
3018 |
+
"learning_rate": 1.3368136960493083e-06,
|
3019 |
+
"loss": 0.5075,
|
3020 |
+
"step": 430
|
3021 |
+
},
|
3022 |
+
{
|
3023 |
+
"epoch": 2.3556771545827635,
|
3024 |
+
"grad_norm": 1.4168515406143662,
|
3025 |
+
"learning_rate": 1.315114605655532e-06,
|
3026 |
+
"loss": 0.5016,
|
3027 |
+
"step": 431
|
3028 |
+
},
|
3029 |
+
{
|
3030 |
+
"epoch": 2.3611491108071134,
|
3031 |
+
"grad_norm": 1.3795049679286153,
|
3032 |
+
"learning_rate": 1.2935663701073586e-06,
|
3033 |
+
"loss": 0.4486,
|
3034 |
+
"step": 432
|
3035 |
+
},
|
3036 |
+
{
|
3037 |
+
"epoch": 2.366621067031464,
|
3038 |
+
"grad_norm": 1.5248587446004036,
|
3039 |
+
"learning_rate": 1.2721698715640352e-06,
|
3040 |
+
"loss": 0.5249,
|
3041 |
+
"step": 433
|
3042 |
+
},
|
3043 |
+
{
|
3044 |
+
"epoch": 2.3720930232558137,
|
3045 |
+
"grad_norm": 1.4838101121746883,
|
3046 |
+
"learning_rate": 1.2509259859728863e-06,
|
3047 |
+
"loss": 0.4806,
|
3048 |
+
"step": 434
|
3049 |
+
},
|
3050 |
+
{
|
3051 |
+
"epoch": 2.377564979480164,
|
3052 |
+
"grad_norm": 1.1655029771800829,
|
3053 |
+
"learning_rate": 1.2298355830334346e-06,
|
3054 |
+
"loss": 0.5134,
|
3055 |
+
"step": 435
|
3056 |
+
},
|
3057 |
+
{
|
3058 |
+
"epoch": 2.3830369357045145,
|
3059 |
+
"grad_norm": 1.4914215769081816,
|
3060 |
+
"learning_rate": 1.2088995261618108e-06,
|
3061 |
+
"loss": 0.5527,
|
3062 |
+
"step": 436
|
3063 |
+
},
|
3064 |
+
{
|
3065 |
+
"epoch": 2.3885088919288644,
|
3066 |
+
"grad_norm": 1.5112689169585833,
|
3067 |
+
"learning_rate": 1.1881186724553968e-06,
|
3068 |
+
"loss": 0.6196,
|
3069 |
+
"step": 437
|
3070 |
+
},
|
3071 |
+
{
|
3072 |
+
"epoch": 2.3939808481532148,
|
3073 |
+
"grad_norm": 1.342222210642936,
|
3074 |
+
"learning_rate": 1.167493872657739e-06,
|
3075 |
+
"loss": 0.5563,
|
3076 |
+
"step": 438
|
3077 |
+
},
|
3078 |
+
{
|
3079 |
+
"epoch": 2.399452804377565,
|
3080 |
+
"grad_norm": 1.4592320622635746,
|
3081 |
+
"learning_rate": 1.1470259711237285e-06,
|
3082 |
+
"loss": 0.5389,
|
3083 |
+
"step": 439
|
3084 |
+
},
|
3085 |
+
{
|
3086 |
+
"epoch": 2.404924760601915,
|
3087 |
+
"grad_norm": 1.4250388765811601,
|
3088 |
+
"learning_rate": 1.1267158057850174e-06,
|
3089 |
+
"loss": 0.56,
|
3090 |
+
"step": 440
|
3091 |
+
},
|
3092 |
+
{
|
3093 |
+
"epoch": 2.4103967168262654,
|
3094 |
+
"grad_norm": 1.368983756236314,
|
3095 |
+
"learning_rate": 1.106564208115734e-06,
|
3096 |
+
"loss": 0.5401,
|
3097 |
+
"step": 441
|
3098 |
+
},
|
3099 |
+
{
|
3100 |
+
"epoch": 2.415868673050616,
|
3101 |
+
"grad_norm": 1.3982141451343644,
|
3102 |
+
"learning_rate": 1.0865720030984268e-06,
|
3103 |
+
"loss": 0.4837,
|
3104 |
+
"step": 442
|
3105 |
+
},
|
3106 |
+
{
|
3107 |
+
"epoch": 2.4213406292749657,
|
3108 |
+
"grad_norm": 1.3682351521759626,
|
3109 |
+
"learning_rate": 1.066740009190304e-06,
|
3110 |
+
"loss": 0.5257,
|
3111 |
+
"step": 443
|
3112 |
+
},
|
3113 |
+
{
|
3114 |
+
"epoch": 2.426812585499316,
|
3115 |
+
"grad_norm": 1.4382522359380867,
|
3116 |
+
"learning_rate": 1.0470690382897164e-06,
|
3117 |
+
"loss": 0.4895,
|
3118 |
+
"step": 444
|
3119 |
+
},
|
3120 |
+
{
|
3121 |
+
"epoch": 2.432284541723666,
|
3122 |
+
"grad_norm": 1.416286032257865,
|
3123 |
+
"learning_rate": 1.0275598957029232e-06,
|
3124 |
+
"loss": 0.4759,
|
3125 |
+
"step": 445
|
3126 |
+
},
|
3127 |
+
{
|
3128 |
+
"epoch": 2.4377564979480164,
|
3129 |
+
"grad_norm": 1.4802031101711004,
|
3130 |
+
"learning_rate": 1.0082133801111294e-06,
|
3131 |
+
"loss": 0.5258,
|
3132 |
+
"step": 446
|
3133 |
+
},
|
3134 |
+
{
|
3135 |
+
"epoch": 2.443228454172367,
|
3136 |
+
"grad_norm": 1.4644345680300666,
|
3137 |
+
"learning_rate": 9.890302835377736e-07,
|
3138 |
+
"loss": 0.4596,
|
3139 |
+
"step": 447
|
3140 |
+
},
|
3141 |
+
{
|
3142 |
+
"epoch": 2.4487004103967167,
|
3143 |
+
"grad_norm": 1.3716111293233866,
|
3144 |
+
"learning_rate": 9.70011391316124e-07,
|
3145 |
+
"loss": 0.4904,
|
3146 |
+
"step": 448
|
3147 |
+
},
|
3148 |
+
{
|
3149 |
+
"epoch": 2.454172366621067,
|
3150 |
+
"grad_norm": 1.3982360411267778,
|
3151 |
+
"learning_rate": 9.511574820571095e-07,
|
3152 |
+
"loss": 0.5418,
|
3153 |
+
"step": 449
|
3154 |
+
},
|
3155 |
+
{
|
3156 |
+
"epoch": 2.459644322845417,
|
3157 |
+
"grad_norm": 1.4825747295752385,
|
3158 |
+
"learning_rate": 9.324693276174568e-07,
|
3159 |
+
"loss": 0.5106,
|
3160 |
+
"step": 450
|
3161 |
+
},
|
3162 |
+
{
|
3163 |
+
"epoch": 2.4651162790697674,
|
3164 |
+
"grad_norm": 1.5263566091863692,
|
3165 |
+
"learning_rate": 9.13947693068083e-07,
|
3166 |
+
"loss": 0.5182,
|
3167 |
+
"step": 451
|
3168 |
+
},
|
3169 |
+
{
|
3170 |
+
"epoch": 2.4705882352941178,
|
3171 |
+
"grad_norm": 1.3936617623110632,
|
3172 |
+
"learning_rate": 8.955933366627778e-07,
|
3173 |
+
"loss": 0.4735,
|
3174 |
+
"step": 452
|
3175 |
+
},
|
3176 |
+
{
|
3177 |
+
"epoch": 2.4760601915184677,
|
3178 |
+
"grad_norm": 1.5867568311822438,
|
3179 |
+
"learning_rate": 8.774070098071669e-07,
|
3180 |
+
"loss": 0.5394,
|
3181 |
+
"step": 453
|
3182 |
+
},
|
3183 |
+
{
|
3184 |
+
"epoch": 2.481532147742818,
|
3185 |
+
"grad_norm": 1.4741897369320691,
|
3186 |
+
"learning_rate": 8.593894570279365e-07,
|
3187 |
+
"loss": 0.4599,
|
3188 |
+
"step": 454
|
3189 |
+
},
|
3190 |
+
{
|
3191 |
+
"epoch": 2.4870041039671684,
|
3192 |
+
"grad_norm": 1.406751443338632,
|
3193 |
+
"learning_rate": 8.415414159423707e-07,
|
3194 |
+
"loss": 0.4669,
|
3195 |
+
"step": 455
|
3196 |
+
},
|
3197 |
+
{
|
3198 |
+
"epoch": 2.4924760601915183,
|
3199 |
+
"grad_norm": 1.3215191171391294,
|
3200 |
+
"learning_rate": 8.238636172281394e-07,
|
3201 |
+
"loss": 0.4434,
|
3202 |
+
"step": 456
|
3203 |
+
},
|
3204 |
+
{
|
3205 |
+
"epoch": 2.4979480164158687,
|
3206 |
+
"grad_norm": 1.4084167815838329,
|
3207 |
+
"learning_rate": 8.063567845933973e-07,
|
3208 |
+
"loss": 0.4307,
|
3209 |
+
"step": 457
|
3210 |
+
},
|
3211 |
+
{
|
3212 |
+
"epoch": 2.503419972640219,
|
3213 |
+
"grad_norm": 1.4853218010741858,
|
3214 |
+
"learning_rate": 7.890216347471408e-07,
|
3215 |
+
"loss": 0.5538,
|
3216 |
+
"step": 458
|
3217 |
+
},
|
3218 |
+
{
|
3219 |
+
"epoch": 2.508891928864569,
|
3220 |
+
"grad_norm": 1.4480270707420555,
|
3221 |
+
"learning_rate": 7.718588773698871e-07,
|
3222 |
+
"loss": 0.4864,
|
3223 |
+
"step": 459
|
3224 |
+
},
|
3225 |
+
{
|
3226 |
+
"epoch": 2.5143638850889194,
|
3227 |
+
"grad_norm": 1.6113443729005184,
|
3228 |
+
"learning_rate": 7.548692150846021e-07,
|
3229 |
+
"loss": 0.5177,
|
3230 |
+
"step": 460
|
3231 |
+
},
|
3232 |
+
{
|
3233 |
+
"epoch": 2.5198358413132693,
|
3234 |
+
"grad_norm": 1.4547360506787226,
|
3235 |
+
"learning_rate": 7.380533434279457e-07,
|
3236 |
+
"loss": 0.4806,
|
3237 |
+
"step": 461
|
3238 |
+
},
|
3239 |
+
{
|
3240 |
+
"epoch": 2.5253077975376197,
|
3241 |
+
"grad_norm": 1.3231121597455795,
|
3242 |
+
"learning_rate": 7.214119508217976e-07,
|
3243 |
+
"loss": 0.458,
|
3244 |
+
"step": 462
|
3245 |
+
},
|
3246 |
+
{
|
3247 |
+
"epoch": 2.53077975376197,
|
3248 |
+
"grad_norm": 1.3899574819257758,
|
3249 |
+
"learning_rate": 7.049457185450692e-07,
|
3250 |
+
"loss": 0.4749,
|
3251 |
+
"step": 463
|
3252 |
+
},
|
3253 |
+
{
|
3254 |
+
"epoch": 2.53625170998632,
|
3255 |
+
"grad_norm": 1.3331196362774294,
|
3256 |
+
"learning_rate": 6.88655320705815e-07,
|
3257 |
+
"loss": 0.717,
|
3258 |
+
"step": 464
|
3259 |
+
},
|
3260 |
+
{
|
3261 |
+
"epoch": 2.5417236662106704,
|
3262 |
+
"grad_norm": 1.2417699357747913,
|
3263 |
+
"learning_rate": 6.725414242136364e-07,
|
3264 |
+
"loss": 0.4695,
|
3265 |
+
"step": 465
|
3266 |
+
},
|
3267 |
+
{
|
3268 |
+
"epoch": 2.5471956224350203,
|
3269 |
+
"grad_norm": 1.3349259639480837,
|
3270 |
+
"learning_rate": 6.566046887523808e-07,
|
3271 |
+
"loss": 0.4882,
|
3272 |
+
"step": 466
|
3273 |
+
},
|
3274 |
+
{
|
3275 |
+
"epoch": 2.5526675786593707,
|
3276 |
+
"grad_norm": 1.482616477635212,
|
3277 |
+
"learning_rate": 6.408457667531298e-07,
|
3278 |
+
"loss": 0.4828,
|
3279 |
+
"step": 467
|
3280 |
+
},
|
3281 |
+
{
|
3282 |
+
"epoch": 2.558139534883721,
|
3283 |
+
"grad_norm": 1.5840846294068265,
|
3284 |
+
"learning_rate": 6.252653033674927e-07,
|
3285 |
+
"loss": 0.5306,
|
3286 |
+
"step": 468
|
3287 |
+
},
|
3288 |
+
{
|
3289 |
+
"epoch": 2.5636114911080714,
|
3290 |
+
"grad_norm": 1.502926751431085,
|
3291 |
+
"learning_rate": 6.098639364411974e-07,
|
3292 |
+
"loss": 0.519,
|
3293 |
+
"step": 469
|
3294 |
+
},
|
3295 |
+
{
|
3296 |
+
"epoch": 2.5690834473324213,
|
3297 |
+
"grad_norm": 1.4050801384243388,
|
3298 |
+
"learning_rate": 5.946422964879706e-07,
|
3299 |
+
"loss": 0.5186,
|
3300 |
+
"step": 470
|
3301 |
+
},
|
3302 |
+
{
|
3303 |
+
"epoch": 2.5745554035567717,
|
3304 |
+
"grad_norm": 1.448838027121391,
|
3305 |
+
"learning_rate": 5.79601006663732e-07,
|
3306 |
+
"loss": 0.4709,
|
3307 |
+
"step": 471
|
3308 |
+
},
|
3309 |
+
{
|
3310 |
+
"epoch": 2.5800273597811216,
|
3311 |
+
"grad_norm": 1.5456411352193866,
|
3312 |
+
"learning_rate": 5.647406827410796e-07,
|
3313 |
+
"loss": 0.4838,
|
3314 |
+
"step": 472
|
3315 |
+
},
|
3316 |
+
{
|
3317 |
+
"epoch": 2.585499316005472,
|
3318 |
+
"grad_norm": 1.4765565301084078,
|
3319 |
+
"learning_rate": 5.500619330840829e-07,
|
3320 |
+
"loss": 0.4406,
|
3321 |
+
"step": 473
|
3322 |
+
},
|
3323 |
+
{
|
3324 |
+
"epoch": 2.5909712722298224,
|
3325 |
+
"grad_norm": 1.4217166633141576,
|
3326 |
+
"learning_rate": 5.35565358623375e-07,
|
3327 |
+
"loss": 0.4995,
|
3328 |
+
"step": 474
|
3329 |
+
},
|
3330 |
+
{
|
3331 |
+
"epoch": 2.5964432284541723,
|
3332 |
+
"grad_norm": 1.4427834950699971,
|
3333 |
+
"learning_rate": 5.212515528315504e-07,
|
3334 |
+
"loss": 0.4685,
|
3335 |
+
"step": 475
|
3336 |
+
},
|
3337 |
+
{
|
3338 |
+
"epoch": 2.6019151846785227,
|
3339 |
+
"grad_norm": 1.4223288046270277,
|
3340 |
+
"learning_rate": 5.071211016988753e-07,
|
3341 |
+
"loss": 0.5057,
|
3342 |
+
"step": 476
|
3343 |
+
},
|
3344 |
+
{
|
3345 |
+
"epoch": 2.6073871409028726,
|
3346 |
+
"grad_norm": 1.445005124323909,
|
3347 |
+
"learning_rate": 4.931745837092888e-07,
|
3348 |
+
"loss": 0.4893,
|
3349 |
+
"step": 477
|
3350 |
+
},
|
3351 |
+
{
|
3352 |
+
"epoch": 2.612859097127223,
|
3353 |
+
"grad_norm": 1.4912981910569638,
|
3354 |
+
"learning_rate": 4.794125698167262e-07,
|
3355 |
+
"loss": 0.4931,
|
3356 |
+
"step": 478
|
3357 |
+
},
|
3358 |
+
{
|
3359 |
+
"epoch": 2.6183310533515733,
|
3360 |
+
"grad_norm": 1.5505485345430878,
|
3361 |
+
"learning_rate": 4.658356234217437e-07,
|
3362 |
+
"loss": 0.5162,
|
3363 |
+
"step": 479
|
3364 |
+
},
|
3365 |
+
{
|
3366 |
+
"epoch": 2.6238030095759233,
|
3367 |
+
"grad_norm": 1.582692545299596,
|
3368 |
+
"learning_rate": 4.5244430034845466e-07,
|
3369 |
+
"loss": 0.5098,
|
3370 |
+
"step": 480
|
3371 |
+
},
|
3372 |
+
{
|
3373 |
+
"epoch": 2.6292749658002736,
|
3374 |
+
"grad_norm": 1.3057040655424323,
|
3375 |
+
"learning_rate": 4.3923914882177124e-07,
|
3376 |
+
"loss": 0.5566,
|
3377 |
+
"step": 481
|
3378 |
+
},
|
3379 |
+
{
|
3380 |
+
"epoch": 2.6347469220246236,
|
3381 |
+
"grad_norm": 1.4402926968965952,
|
3382 |
+
"learning_rate": 4.262207094449633e-07,
|
3383 |
+
"loss": 0.4831,
|
3384 |
+
"step": 482
|
3385 |
+
},
|
3386 |
+
{
|
3387 |
+
"epoch": 2.640218878248974,
|
3388 |
+
"grad_norm": 1.3616226222730219,
|
3389 |
+
"learning_rate": 4.1338951517752637e-07,
|
3390 |
+
"loss": 0.4815,
|
3391 |
+
"step": 483
|
3392 |
+
},
|
3393 |
+
{
|
3394 |
+
"epoch": 2.6456908344733243,
|
3395 |
+
"grad_norm": 1.2604235585335275,
|
3396 |
+
"learning_rate": 4.007460913133643e-07,
|
3397 |
+
"loss": 0.4281,
|
3398 |
+
"step": 484
|
3399 |
+
},
|
3400 |
+
{
|
3401 |
+
"epoch": 2.6511627906976747,
|
3402 |
+
"grad_norm": 1.3141541799725363,
|
3403 |
+
"learning_rate": 3.8829095545928177e-07,
|
3404 |
+
"loss": 0.4698,
|
3405 |
+
"step": 485
|
3406 |
+
},
|
3407 |
+
{
|
3408 |
+
"epoch": 2.6566347469220246,
|
3409 |
+
"grad_norm": 1.4113085679556834,
|
3410 |
+
"learning_rate": 3.760246175137938e-07,
|
3411 |
+
"loss": 0.4481,
|
3412 |
+
"step": 486
|
3413 |
+
},
|
3414 |
+
{
|
3415 |
+
"epoch": 2.662106703146375,
|
3416 |
+
"grad_norm": 1.4221582445539107,
|
3417 |
+
"learning_rate": 3.639475796462577e-07,
|
3418 |
+
"loss": 0.4609,
|
3419 |
+
"step": 487
|
3420 |
+
},
|
3421 |
+
{
|
3422 |
+
"epoch": 2.667578659370725,
|
3423 |
+
"grad_norm": 1.2870484042595358,
|
3424 |
+
"learning_rate": 3.5206033627630145e-07,
|
3425 |
+
"loss": 0.4648,
|
3426 |
+
"step": 488
|
3427 |
+
},
|
3428 |
+
{
|
3429 |
+
"epoch": 2.6730506155950753,
|
3430 |
+
"grad_norm": 1.8538333641407392,
|
3431 |
+
"learning_rate": 3.403633740535983e-07,
|
3432 |
+
"loss": 0.4802,
|
3433 |
+
"step": 489
|
3434 |
+
},
|
3435 |
+
{
|
3436 |
+
"epoch": 2.6785225718194257,
|
3437 |
+
"grad_norm": 1.3905043619878708,
|
3438 |
+
"learning_rate": 3.2885717183793296e-07,
|
3439 |
+
"loss": 0.4876,
|
3440 |
+
"step": 490
|
3441 |
+
},
|
3442 |
+
{
|
3443 |
+
"epoch": 2.6839945280437756,
|
3444 |
+
"grad_norm": 1.3979605126017662,
|
3445 |
+
"learning_rate": 3.175422006796031e-07,
|
3446 |
+
"loss": 0.5916,
|
3447 |
+
"step": 491
|
3448 |
+
},
|
3449 |
+
{
|
3450 |
+
"epoch": 2.689466484268126,
|
3451 |
+
"grad_norm": 1.5179518019725438,
|
3452 |
+
"learning_rate": 3.064189238001325e-07,
|
3453 |
+
"loss": 0.5125,
|
3454 |
+
"step": 492
|
3455 |
+
},
|
3456 |
+
{
|
3457 |
+
"epoch": 2.694938440492476,
|
3458 |
+
"grad_norm": 1.487366400847455,
|
3459 |
+
"learning_rate": 2.9548779657330605e-07,
|
3460 |
+
"loss": 0.4823,
|
3461 |
+
"step": 493
|
3462 |
+
},
|
3463 |
+
{
|
3464 |
+
"epoch": 2.7004103967168263,
|
3465 |
+
"grad_norm": 1.3018965167310825,
|
3466 |
+
"learning_rate": 2.847492665065349e-07,
|
3467 |
+
"loss": 0.4857,
|
3468 |
+
"step": 494
|
3469 |
+
},
|
3470 |
+
{
|
3471 |
+
"epoch": 2.7058823529411766,
|
3472 |
+
"grad_norm": 1.5289210780477058,
|
3473 |
+
"learning_rate": 2.742037732225228e-07,
|
3474 |
+
"loss": 0.4717,
|
3475 |
+
"step": 495
|
3476 |
+
},
|
3477 |
+
{
|
3478 |
+
"epoch": 2.7113543091655266,
|
3479 |
+
"grad_norm": 1.449130021649974,
|
3480 |
+
"learning_rate": 2.6385174844128337e-07,
|
3481 |
+
"loss": 0.4976,
|
3482 |
+
"step": 496
|
3483 |
+
},
|
3484 |
+
{
|
3485 |
+
"epoch": 2.716826265389877,
|
3486 |
+
"grad_norm": 1.4189628801897225,
|
3487 |
+
"learning_rate": 2.5369361596245366e-07,
|
3488 |
+
"loss": 0.484,
|
3489 |
+
"step": 497
|
3490 |
+
},
|
3491 |
+
{
|
3492 |
+
"epoch": 2.722298221614227,
|
3493 |
+
"grad_norm": 1.374061457609359,
|
3494 |
+
"learning_rate": 2.4372979164795306e-07,
|
3495 |
+
"loss": 0.6527,
|
3496 |
+
"step": 498
|
3497 |
+
},
|
3498 |
+
{
|
3499 |
+
"epoch": 2.7277701778385772,
|
3500 |
+
"grad_norm": 1.4113236800472866,
|
3501 |
+
"learning_rate": 2.3396068340495182e-07,
|
3502 |
+
"loss": 0.4795,
|
3503 |
+
"step": 499
|
3504 |
+
},
|
3505 |
+
{
|
3506 |
+
"epoch": 2.7332421340629276,
|
3507 |
+
"grad_norm": 1.3747519076775883,
|
3508 |
+
"learning_rate": 2.2438669116917612e-07,
|
3509 |
+
"loss": 0.5174,
|
3510 |
+
"step": 500
|
3511 |
+
},
|
3512 |
+
{
|
3513 |
+
"epoch": 2.738714090287278,
|
3514 |
+
"grad_norm": 1.4071509491276206,
|
3515 |
+
"learning_rate": 2.1500820688853562e-07,
|
3516 |
+
"loss": 0.4473,
|
3517 |
+
"step": 501
|
3518 |
+
},
|
3519 |
+
{
|
3520 |
+
"epoch": 2.744186046511628,
|
3521 |
+
"grad_norm": 1.5010812602516383,
|
3522 |
+
"learning_rate": 2.0582561450707027e-07,
|
3523 |
+
"loss": 0.4801,
|
3524 |
+
"step": 502
|
3525 |
+
},
|
3526 |
+
{
|
3527 |
+
"epoch": 2.7496580027359783,
|
3528 |
+
"grad_norm": 1.4177840375890944,
|
3529 |
+
"learning_rate": 1.9683928994924383e-07,
|
3530 |
+
"loss": 0.5184,
|
3531 |
+
"step": 503
|
3532 |
+
},
|
3533 |
+
{
|
3534 |
+
"epoch": 2.755129958960328,
|
3535 |
+
"grad_norm": 1.3793947690023178,
|
3536 |
+
"learning_rate": 1.8804960110454406e-07,
|
3537 |
+
"loss": 0.4947,
|
3538 |
+
"step": 504
|
3539 |
+
},
|
3540 |
+
{
|
3541 |
+
"epoch": 2.7606019151846786,
|
3542 |
+
"grad_norm": 1.3354507535421465,
|
3543 |
+
"learning_rate": 1.7945690781242752e-07,
|
3544 |
+
"loss": 0.5058,
|
3545 |
+
"step": 505
|
3546 |
+
},
|
3547 |
+
{
|
3548 |
+
"epoch": 2.766073871409029,
|
3549 |
+
"grad_norm": 1.372975817116268,
|
3550 |
+
"learning_rate": 1.7106156184758248e-07,
|
3551 |
+
"loss": 0.4957,
|
3552 |
+
"step": 506
|
3553 |
+
},
|
3554 |
+
{
|
3555 |
+
"epoch": 2.771545827633379,
|
3556 |
+
"grad_norm": 1.3927564177847465,
|
3557 |
+
"learning_rate": 1.62863906905536e-07,
|
3558 |
+
"loss": 0.4932,
|
3559 |
+
"step": 507
|
3560 |
+
},
|
3561 |
+
{
|
3562 |
+
"epoch": 2.7770177838577292,
|
3563 |
+
"grad_norm": 1.2685868673247336,
|
3564 |
+
"learning_rate": 1.5486427858857567e-07,
|
3565 |
+
"loss": 0.5382,
|
3566 |
+
"step": 508
|
3567 |
+
},
|
3568 |
+
{
|
3569 |
+
"epoch": 2.782489740082079,
|
3570 |
+
"grad_norm": 1.3517266407465647,
|
3571 |
+
"learning_rate": 1.4706300439201348e-07,
|
3572 |
+
"loss": 0.532,
|
3573 |
+
"step": 509
|
3574 |
+
},
|
3575 |
+
{
|
3576 |
+
"epoch": 2.7879616963064295,
|
3577 |
+
"grad_norm": 1.4881697743456797,
|
3578 |
+
"learning_rate": 1.394604036907804e-07,
|
3579 |
+
"loss": 0.4758,
|
3580 |
+
"step": 510
|
3581 |
+
},
|
3582 |
+
{
|
3583 |
+
"epoch": 2.79343365253078,
|
3584 |
+
"grad_norm": 1.394221988348371,
|
3585 |
+
"learning_rate": 1.320567877263479e-07,
|
3586 |
+
"loss": 0.5132,
|
3587 |
+
"step": 511
|
3588 |
+
},
|
3589 |
+
{
|
3590 |
+
"epoch": 2.79890560875513,
|
3591 |
+
"grad_norm": 1.3720661616596983,
|
3592 |
+
"learning_rate": 1.2485245959398928e-07,
|
3593 |
+
"loss": 0.4615,
|
3594 |
+
"step": 512
|
3595 |
+
},
|
3596 |
+
{
|
3597 |
+
"epoch": 2.80437756497948,
|
3598 |
+
"grad_norm": 1.3415690826089453,
|
3599 |
+
"learning_rate": 1.17847714230368e-07,
|
3600 |
+
"loss": 0.4846,
|
3601 |
+
"step": 513
|
3602 |
+
},
|
3603 |
+
{
|
3604 |
+
"epoch": 2.80984952120383,
|
3605 |
+
"grad_norm": 1.450111290610619,
|
3606 |
+
"learning_rate": 1.1104283840146834e-07,
|
3607 |
+
"loss": 0.47,
|
3608 |
+
"step": 514
|
3609 |
+
},
|
3610 |
+
{
|
3611 |
+
"epoch": 2.8153214774281805,
|
3612 |
+
"grad_norm": 1.358999633729104,
|
3613 |
+
"learning_rate": 1.044381106908493e-07,
|
3614 |
+
"loss": 0.672,
|
3615 |
+
"step": 515
|
3616 |
+
},
|
3617 |
+
{
|
3618 |
+
"epoch": 2.820793433652531,
|
3619 |
+
"grad_norm": 1.3237496780591622,
|
3620 |
+
"learning_rate": 9.803380148824371e-08,
|
3621 |
+
"loss": 0.5008,
|
3622 |
+
"step": 516
|
3623 |
+
},
|
3624 |
+
{
|
3625 |
+
"epoch": 2.8262653898768813,
|
3626 |
+
"grad_norm": 1.3157858442276071,
|
3627 |
+
"learning_rate": 9.18301729784904e-08,
|
3628 |
+
"loss": 0.4817,
|
3629 |
+
"step": 517
|
3630 |
+
},
|
3631 |
+
{
|
3632 |
+
"epoch": 2.831737346101231,
|
3633 |
+
"grad_norm": 1.5340580922379239,
|
3634 |
+
"learning_rate": 8.582747913079448e-08,
|
3635 |
+
"loss": 0.5053,
|
3636 |
+
"step": 518
|
3637 |
+
},
|
3638 |
+
{
|
3639 |
+
"epoch": 2.8372093023255816,
|
3640 |
+
"grad_norm": 1.3677853840028062,
|
3641 |
+
"learning_rate": 8.00259656883362e-08,
|
3642 |
+
"loss": 0.4637,
|
3643 |
+
"step": 519
|
3644 |
+
},
|
3645 |
+
{
|
3646 |
+
"epoch": 2.8426812585499315,
|
3647 |
+
"grad_norm": 1.3598439809429441,
|
3648 |
+
"learning_rate": 7.442587015820734e-08,
|
3649 |
+
"loss": 0.4491,
|
3650 |
+
"step": 520
|
3651 |
+
},
|
3652 |
+
{
|
3653 |
+
"epoch": 2.848153214774282,
|
3654 |
+
"grad_norm": 1.50932549615933,
|
3655 |
+
"learning_rate": 6.902742180168953e-08,
|
3656 |
+
"loss": 0.5315,
|
3657 |
+
"step": 521
|
3658 |
+
},
|
3659 |
+
{
|
3660 |
+
"epoch": 2.853625170998632,
|
3661 |
+
"grad_norm": 1.4348534175076735,
|
3662 |
+
"learning_rate": 6.38308416248673e-08,
|
3663 |
+
"loss": 0.4774,
|
3664 |
+
"step": 522
|
3665 |
+
},
|
3666 |
+
{
|
3667 |
+
"epoch": 2.859097127222982,
|
3668 |
+
"grad_norm": 1.4398470821718545,
|
3669 |
+
"learning_rate": 5.883634236958091e-08,
|
3670 |
+
"loss": 0.479,
|
3671 |
+
"step": 523
|
3672 |
+
},
|
3673 |
+
{
|
3674 |
+
"epoch": 2.8645690834473325,
|
3675 |
+
"grad_norm": 1.3633451880172722,
|
3676 |
+
"learning_rate": 5.4044128504717186e-08,
|
3677 |
+
"loss": 0.441,
|
3678 |
+
"step": 524
|
3679 |
+
},
|
3680 |
+
{
|
3681 |
+
"epoch": 2.8700410396716824,
|
3682 |
+
"grad_norm": 1.5079206667748024,
|
3683 |
+
"learning_rate": 4.945439621783843e-08,
|
3684 |
+
"loss": 0.5298,
|
3685 |
+
"step": 525
|
3686 |
+
},
|
3687 |
+
{
|
3688 |
+
"epoch": 2.875512995896033,
|
3689 |
+
"grad_norm": 1.3887050230878182,
|
3690 |
+
"learning_rate": 4.506733340714997e-08,
|
3691 |
+
"loss": 0.5647,
|
3692 |
+
"step": 526
|
3693 |
+
},
|
3694 |
+
{
|
3695 |
+
"epoch": 2.880984952120383,
|
3696 |
+
"grad_norm": 1.5999769572204152,
|
3697 |
+
"learning_rate": 4.08831196738102e-08,
|
3698 |
+
"loss": 0.5675,
|
3699 |
+
"step": 527
|
3700 |
+
},
|
3701 |
+
{
|
3702 |
+
"epoch": 2.886456908344733,
|
3703 |
+
"grad_norm": 1.3968640089346598,
|
3704 |
+
"learning_rate": 3.69019263145759e-08,
|
3705 |
+
"loss": 0.5109,
|
3706 |
+
"step": 528
|
3707 |
+
},
|
3708 |
+
{
|
3709 |
+
"epoch": 2.8919288645690835,
|
3710 |
+
"grad_norm": 1.52921194939027,
|
3711 |
+
"learning_rate": 3.312391631479006e-08,
|
3712 |
+
"loss": 0.4531,
|
3713 |
+
"step": 529
|
3714 |
+
},
|
3715 |
+
{
|
3716 |
+
"epoch": 2.8974008207934334,
|
3717 |
+
"grad_norm": 1.399899851418665,
|
3718 |
+
"learning_rate": 2.9549244341708917e-08,
|
3719 |
+
"loss": 0.5175,
|
3720 |
+
"step": 530
|
3721 |
+
},
|
3722 |
+
{
|
3723 |
+
"epoch": 2.902872777017784,
|
3724 |
+
"grad_norm": 1.4335014611970036,
|
3725 |
+
"learning_rate": 2.6178056738170866e-08,
|
3726 |
+
"loss": 0.4984,
|
3727 |
+
"step": 531
|
3728 |
+
},
|
3729 |
+
{
|
3730 |
+
"epoch": 2.908344733242134,
|
3731 |
+
"grad_norm": 1.2787386075584488,
|
3732 |
+
"learning_rate": 2.301049151660628e-08,
|
3733 |
+
"loss": 0.5347,
|
3734 |
+
"step": 532
|
3735 |
+
},
|
3736 |
+
{
|
3737 |
+
"epoch": 2.9138166894664845,
|
3738 |
+
"grad_norm": 1.415937397596885,
|
3739 |
+
"learning_rate": 2.0046678353384254e-08,
|
3740 |
+
"loss": 0.5113,
|
3741 |
+
"step": 533
|
3742 |
+
},
|
3743 |
+
{
|
3744 |
+
"epoch": 2.9192886456908345,
|
3745 |
+
"grad_norm": 1.48137689826664,
|
3746 |
+
"learning_rate": 1.7286738583507935e-08,
|
3747 |
+
"loss": 0.4774,
|
3748 |
+
"step": 534
|
3749 |
+
},
|
3750 |
+
{
|
3751 |
+
"epoch": 2.924760601915185,
|
3752 |
+
"grad_norm": 1.492339871266796,
|
3753 |
+
"learning_rate": 1.4730785195643527e-08,
|
3754 |
+
"loss": 0.4891,
|
3755 |
+
"step": 535
|
3756 |
+
},
|
3757 |
+
{
|
3758 |
+
"epoch": 2.9302325581395348,
|
3759 |
+
"grad_norm": 1.3924981147788802,
|
3760 |
+
"learning_rate": 1.2378922827496198e-08,
|
3761 |
+
"loss": 0.4572,
|
3762 |
+
"step": 536
|
3763 |
+
},
|
3764 |
+
{
|
3765 |
+
"epoch": 2.935704514363885,
|
3766 |
+
"grad_norm": 1.3980536557644345,
|
3767 |
+
"learning_rate": 1.0231247761528507e-08,
|
3768 |
+
"loss": 0.4398,
|
3769 |
+
"step": 537
|
3770 |
+
},
|
3771 |
+
{
|
3772 |
+
"epoch": 2.9411764705882355,
|
3773 |
+
"grad_norm": 1.3251035750169489,
|
3774 |
+
"learning_rate": 8.287847921013558e-09,
|
3775 |
+
"loss": 0.4239,
|
3776 |
+
"step": 538
|
3777 |
+
},
|
3778 |
+
{
|
3779 |
+
"epoch": 2.9466484268125854,
|
3780 |
+
"grad_norm": 1.3490788515149803,
|
3781 |
+
"learning_rate": 6.548802866441218e-09,
|
3782 |
+
"loss": 0.5259,
|
3783 |
+
"step": 539
|
3784 |
+
},
|
3785 |
+
{
|
3786 |
+
"epoch": 2.952120383036936,
|
3787 |
+
"grad_norm": 1.3283269567366318,
|
3788 |
+
"learning_rate": 5.014183792256266e-09,
|
3789 |
+
"loss": 0.5069,
|
3790 |
+
"step": 540
|
3791 |
+
},
|
3792 |
+
{
|
3793 |
+
"epoch": 2.9575923392612857,
|
3794 |
+
"grad_norm": 1.3650538885339099,
|
3795 |
+
"learning_rate": 3.6840535239474063e-09,
|
3796 |
+
"loss": 0.4951,
|
3797 |
+
"step": 541
|
3798 |
+
},
|
3799 |
+
{
|
3800 |
+
"epoch": 2.963064295485636,
|
3801 |
+
"grad_norm": 1.4029820737971126,
|
3802 |
+
"learning_rate": 2.5584665154732015e-09,
|
3803 |
+
"loss": 0.4817,
|
3804 |
+
"step": 542
|
3805 |
+
},
|
3806 |
+
{
|
3807 |
+
"epoch": 2.9685362517099865,
|
3808 |
+
"grad_norm": 1.4777600539763867,
|
3809 |
+
"learning_rate": 1.6374688470327526e-09,
|
3810 |
+
"loss": 0.5063,
|
3811 |
+
"step": 543
|
3812 |
+
},
|
3813 |
+
{
|
3814 |
+
"epoch": 2.9740082079343364,
|
3815 |
+
"grad_norm": 1.4050217048219682,
|
3816 |
+
"learning_rate": 9.210982231805388e-10,
|
3817 |
+
"loss": 0.5241,
|
3818 |
+
"step": 544
|
3819 |
+
},
|
3820 |
+
{
|
3821 |
+
"epoch": 2.9794801641586868,
|
3822 |
+
"grad_norm": 1.399800255307597,
|
3823 |
+
"learning_rate": 4.093839712815406e-10,
|
3824 |
+
"loss": 0.5268,
|
3825 |
+
"step": 545
|
3826 |
+
},
|
3827 |
+
{
|
3828 |
+
"epoch": 2.9849521203830367,
|
3829 |
+
"grad_norm": 1.3917458241680738,
|
3830 |
+
"learning_rate": 1.0234704031220066e-10,
|
3831 |
+
"loss": 0.4545,
|
3832 |
+
"step": 546
|
3833 |
+
},
|
3834 |
+
{
|
3835 |
+
"epoch": 2.9849521203830367,
|
3836 |
+
"step": 546,
|
3837 |
+
"total_flos": 4.3229407025404314e+17,
|
3838 |
+
"train_loss": 0.7402517178774753,
|
3839 |
+
"train_runtime": 59397.3306,
|
3840 |
+
"train_samples_per_second": 0.443,
|
3841 |
+
"train_steps_per_second": 0.009
|
3842 |
+
}
|
3843 |
+
],
|
3844 |
+
"logging_steps": 1,
|
3845 |
+
"max_steps": 546,
|
3846 |
+
"num_input_tokens_seen": 0,
|
3847 |
+
"num_train_epochs": 3,
|
3848 |
+
"save_steps": 200,
|
3849 |
+
"stateful_callbacks": {
|
3850 |
+
"TrainerControl": {
|
3851 |
+
"args": {
|
3852 |
+
"should_epoch_stop": false,
|
3853 |
+
"should_evaluate": false,
|
3854 |
+
"should_log": false,
|
3855 |
+
"should_save": true,
|
3856 |
+
"should_training_stop": true
|
3857 |
+
},
|
3858 |
+
"attributes": {}
|
3859 |
+
}
|
3860 |
+
},
|
3861 |
+
"total_flos": 4.3229407025404314e+17,
|
3862 |
+
"train_batch_size": 1,
|
3863 |
+
"trial_name": null,
|
3864 |
+
"trial_params": null
|
3865 |
+
}
|
training_loss.png
ADDED
![]() |