End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +3676 -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_final_model_agent_train
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# best_n_no_rationale_poc_agent_final_model_agent_train
|
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_final_model_agent_train
|
|
|
16 |
|
17 |
# best_n_no_rationale_poc_agent_final_model_agent_train
|
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 dataset.
|
20 |
|
21 |
## Model description
|
22 |
|
all_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9841726618705033,
|
3 |
+
"total_flos": 4.226722994984059e+17,
|
4 |
+
"train_loss": 0.7501044547282203,
|
5 |
+
"train_runtime": 58208.7442,
|
6 |
+
"train_samples_per_second": 0.43,
|
7 |
+
"train_steps_per_second": 0.009
|
8 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9841726618705033,
|
3 |
+
"total_flos": 4.226722994984059e+17,
|
4 |
+
"train_loss": 0.7501044547282203,
|
5 |
+
"train_runtime": 58208.7442,
|
6 |
+
"train_samples_per_second": 0.43,
|
7 |
+
"train_steps_per_second": 0.009
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3676 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 2.9841726618705033,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 519,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.0057553956834532375,
|
14 |
+
"grad_norm": 8.975284118896141,
|
15 |
+
"learning_rate": 0.0,
|
16 |
+
"loss": 1.7935,
|
17 |
+
"step": 1
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.011510791366906475,
|
21 |
+
"grad_norm": 11.323986571043637,
|
22 |
+
"learning_rate": 1.9230769230769234e-07,
|
23 |
+
"loss": 2.0296,
|
24 |
+
"step": 2
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.017266187050359712,
|
28 |
+
"grad_norm": 10.863704937427778,
|
29 |
+
"learning_rate": 3.846153846153847e-07,
|
30 |
+
"loss": 2.0072,
|
31 |
+
"step": 3
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.02302158273381295,
|
35 |
+
"grad_norm": 9.658002878883053,
|
36 |
+
"learning_rate": 5.76923076923077e-07,
|
37 |
+
"loss": 1.9133,
|
38 |
+
"step": 4
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.02877697841726619,
|
42 |
+
"grad_norm": 10.47065617772297,
|
43 |
+
"learning_rate": 7.692307692307694e-07,
|
44 |
+
"loss": 1.9452,
|
45 |
+
"step": 5
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.034532374100719423,
|
49 |
+
"grad_norm": 9.591343689284908,
|
50 |
+
"learning_rate": 9.615384615384617e-07,
|
51 |
+
"loss": 1.8685,
|
52 |
+
"step": 6
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.04028776978417266,
|
56 |
+
"grad_norm": 9.36430038639622,
|
57 |
+
"learning_rate": 1.153846153846154e-06,
|
58 |
+
"loss": 1.8723,
|
59 |
+
"step": 7
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.0460431654676259,
|
63 |
+
"grad_norm": 9.793239712391749,
|
64 |
+
"learning_rate": 1.3461538461538462e-06,
|
65 |
+
"loss": 1.8741,
|
66 |
+
"step": 8
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.051798561151079135,
|
70 |
+
"grad_norm": 8.392569717828799,
|
71 |
+
"learning_rate": 1.5384615384615387e-06,
|
72 |
+
"loss": 1.7827,
|
73 |
+
"step": 9
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.05755395683453238,
|
77 |
+
"grad_norm": 7.599784904699743,
|
78 |
+
"learning_rate": 1.7307692307692308e-06,
|
79 |
+
"loss": 1.6319,
|
80 |
+
"step": 10
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.06330935251798561,
|
84 |
+
"grad_norm": 5.756305317893059,
|
85 |
+
"learning_rate": 1.9230769230769234e-06,
|
86 |
+
"loss": 1.4376,
|
87 |
+
"step": 11
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.06906474820143885,
|
91 |
+
"grad_norm": 5.888600887717791,
|
92 |
+
"learning_rate": 2.1153846153846155e-06,
|
93 |
+
"loss": 1.613,
|
94 |
+
"step": 12
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.07482014388489208,
|
98 |
+
"grad_norm": 6.445756277088529,
|
99 |
+
"learning_rate": 2.307692307692308e-06,
|
100 |
+
"loss": 1.7048,
|
101 |
+
"step": 13
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.08057553956834532,
|
105 |
+
"grad_norm": 5.397515284243656,
|
106 |
+
"learning_rate": 2.5e-06,
|
107 |
+
"loss": 1.5757,
|
108 |
+
"step": 14
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.08633093525179857,
|
112 |
+
"grad_norm": 4.2168819059926435,
|
113 |
+
"learning_rate": 2.6923076923076923e-06,
|
114 |
+
"loss": 1.6356,
|
115 |
+
"step": 15
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.0920863309352518,
|
119 |
+
"grad_norm": 3.5279506728061225,
|
120 |
+
"learning_rate": 2.8846153846153845e-06,
|
121 |
+
"loss": 1.4826,
|
122 |
+
"step": 16
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.09784172661870504,
|
126 |
+
"grad_norm": 3.3853850388767026,
|
127 |
+
"learning_rate": 3.0769230769230774e-06,
|
128 |
+
"loss": 1.4606,
|
129 |
+
"step": 17
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.10359712230215827,
|
133 |
+
"grad_norm": 3.2596920074222924,
|
134 |
+
"learning_rate": 3.2692307692307696e-06,
|
135 |
+
"loss": 1.3337,
|
136 |
+
"step": 18
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.1093525179856115,
|
140 |
+
"grad_norm": 3.235476203209979,
|
141 |
+
"learning_rate": 3.4615384615384617e-06,
|
142 |
+
"loss": 1.4853,
|
143 |
+
"step": 19
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.11510791366906475,
|
147 |
+
"grad_norm": 2.7469878182118546,
|
148 |
+
"learning_rate": 3.653846153846154e-06,
|
149 |
+
"loss": 1.4027,
|
150 |
+
"step": 20
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.12086330935251799,
|
154 |
+
"grad_norm": 2.761242571719062,
|
155 |
+
"learning_rate": 3.846153846153847e-06,
|
156 |
+
"loss": 1.3393,
|
157 |
+
"step": 21
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.12661870503597122,
|
161 |
+
"grad_norm": 3.1110754536028966,
|
162 |
+
"learning_rate": 4.0384615384615385e-06,
|
163 |
+
"loss": 1.2821,
|
164 |
+
"step": 22
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.13237410071942446,
|
168 |
+
"grad_norm": 2.790205260294228,
|
169 |
+
"learning_rate": 4.230769230769231e-06,
|
170 |
+
"loss": 1.3744,
|
171 |
+
"step": 23
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.1381294964028777,
|
175 |
+
"grad_norm": 2.7018374985754217,
|
176 |
+
"learning_rate": 4.423076923076924e-06,
|
177 |
+
"loss": 1.2871,
|
178 |
+
"step": 24
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.14388489208633093,
|
182 |
+
"grad_norm": 2.431270538491161,
|
183 |
+
"learning_rate": 4.615384615384616e-06,
|
184 |
+
"loss": 1.2311,
|
185 |
+
"step": 25
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.14964028776978416,
|
189 |
+
"grad_norm": 2.49634890151382,
|
190 |
+
"learning_rate": 4.807692307692308e-06,
|
191 |
+
"loss": 1.2826,
|
192 |
+
"step": 26
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.1553956834532374,
|
196 |
+
"grad_norm": 2.2382397507811826,
|
197 |
+
"learning_rate": 5e-06,
|
198 |
+
"loss": 1.3041,
|
199 |
+
"step": 27
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.16115107913669063,
|
203 |
+
"grad_norm": 1.9310878685613995,
|
204 |
+
"learning_rate": 5.192307692307693e-06,
|
205 |
+
"loss": 1.1603,
|
206 |
+
"step": 28
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.1669064748201439,
|
210 |
+
"grad_norm": 1.783352506029676,
|
211 |
+
"learning_rate": 5.384615384615385e-06,
|
212 |
+
"loss": 1.0935,
|
213 |
+
"step": 29
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.17266187050359713,
|
217 |
+
"grad_norm": 1.6849577741027812,
|
218 |
+
"learning_rate": 5.576923076923077e-06,
|
219 |
+
"loss": 1.2065,
|
220 |
+
"step": 30
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.17841726618705037,
|
224 |
+
"grad_norm": 1.8408328237557863,
|
225 |
+
"learning_rate": 5.769230769230769e-06,
|
226 |
+
"loss": 1.1324,
|
227 |
+
"step": 31
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.1841726618705036,
|
231 |
+
"grad_norm": 1.642858073647887,
|
232 |
+
"learning_rate": 5.961538461538462e-06,
|
233 |
+
"loss": 1.0275,
|
234 |
+
"step": 32
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.18992805755395684,
|
238 |
+
"grad_norm": 1.6097240208073438,
|
239 |
+
"learning_rate": 6.153846153846155e-06,
|
240 |
+
"loss": 1.128,
|
241 |
+
"step": 33
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.19568345323741007,
|
245 |
+
"grad_norm": 1.5639373130122167,
|
246 |
+
"learning_rate": 6.3461538461538466e-06,
|
247 |
+
"loss": 1.091,
|
248 |
+
"step": 34
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.2014388489208633,
|
252 |
+
"grad_norm": 1.6314793190946015,
|
253 |
+
"learning_rate": 6.538461538461539e-06,
|
254 |
+
"loss": 1.0903,
|
255 |
+
"step": 35
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.20719424460431654,
|
259 |
+
"grad_norm": 1.4473254024163007,
|
260 |
+
"learning_rate": 6.730769230769232e-06,
|
261 |
+
"loss": 1.0229,
|
262 |
+
"step": 36
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.21294964028776978,
|
266 |
+
"grad_norm": 1.5274681367692997,
|
267 |
+
"learning_rate": 6.923076923076923e-06,
|
268 |
+
"loss": 1.0516,
|
269 |
+
"step": 37
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.218705035971223,
|
273 |
+
"grad_norm": 1.466474218155718,
|
274 |
+
"learning_rate": 7.115384615384616e-06,
|
275 |
+
"loss": 1.0718,
|
276 |
+
"step": 38
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.22446043165467625,
|
280 |
+
"grad_norm": 1.6149235234842247,
|
281 |
+
"learning_rate": 7.307692307692308e-06,
|
282 |
+
"loss": 1.1291,
|
283 |
+
"step": 39
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.2302158273381295,
|
287 |
+
"grad_norm": 1.5045479971718942,
|
288 |
+
"learning_rate": 7.500000000000001e-06,
|
289 |
+
"loss": 1.1663,
|
290 |
+
"step": 40
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.23597122302158274,
|
294 |
+
"grad_norm": 1.4566148798426637,
|
295 |
+
"learning_rate": 7.692307692307694e-06,
|
296 |
+
"loss": 1.0187,
|
297 |
+
"step": 41
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.24172661870503598,
|
301 |
+
"grad_norm": 1.4469981425572818,
|
302 |
+
"learning_rate": 7.884615384615384e-06,
|
303 |
+
"loss": 1.0204,
|
304 |
+
"step": 42
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.2474820143884892,
|
308 |
+
"grad_norm": 1.443973622906549,
|
309 |
+
"learning_rate": 8.076923076923077e-06,
|
310 |
+
"loss": 1.04,
|
311 |
+
"step": 43
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.25323741007194245,
|
315 |
+
"grad_norm": 1.5422955586152927,
|
316 |
+
"learning_rate": 8.26923076923077e-06,
|
317 |
+
"loss": 1.0735,
|
318 |
+
"step": 44
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.2589928057553957,
|
322 |
+
"grad_norm": 1.4163604764062947,
|
323 |
+
"learning_rate": 8.461538461538462e-06,
|
324 |
+
"loss": 1.0329,
|
325 |
+
"step": 45
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.2647482014388489,
|
329 |
+
"grad_norm": 1.409596803987455,
|
330 |
+
"learning_rate": 8.653846153846155e-06,
|
331 |
+
"loss": 1.024,
|
332 |
+
"step": 46
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.27050359712230215,
|
336 |
+
"grad_norm": 1.4001334387747248,
|
337 |
+
"learning_rate": 8.846153846153847e-06,
|
338 |
+
"loss": 1.0477,
|
339 |
+
"step": 47
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.2762589928057554,
|
343 |
+
"grad_norm": 1.5130110902245912,
|
344 |
+
"learning_rate": 9.03846153846154e-06,
|
345 |
+
"loss": 1.0495,
|
346 |
+
"step": 48
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 0.2820143884892086,
|
350 |
+
"grad_norm": 1.366726557354173,
|
351 |
+
"learning_rate": 9.230769230769232e-06,
|
352 |
+
"loss": 1.0559,
|
353 |
+
"step": 49
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 0.28776978417266186,
|
357 |
+
"grad_norm": 1.3312503432521117,
|
358 |
+
"learning_rate": 9.423076923076923e-06,
|
359 |
+
"loss": 1.0597,
|
360 |
+
"step": 50
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.2935251798561151,
|
364 |
+
"grad_norm": 1.513669230441623,
|
365 |
+
"learning_rate": 9.615384615384616e-06,
|
366 |
+
"loss": 1.0003,
|
367 |
+
"step": 51
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 0.2992805755395683,
|
371 |
+
"grad_norm": 1.5652764193593758,
|
372 |
+
"learning_rate": 9.807692307692308e-06,
|
373 |
+
"loss": 1.0464,
|
374 |
+
"step": 52
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 0.30503597122302156,
|
378 |
+
"grad_norm": 1.451799189297248,
|
379 |
+
"learning_rate": 1e-05,
|
380 |
+
"loss": 1.0209,
|
381 |
+
"step": 53
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.3107913669064748,
|
385 |
+
"grad_norm": 1.2679629544309277,
|
386 |
+
"learning_rate": 9.99988686307906e-06,
|
387 |
+
"loss": 0.9504,
|
388 |
+
"step": 54
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 0.31654676258992803,
|
392 |
+
"grad_norm": 1.3318832072195237,
|
393 |
+
"learning_rate": 9.999547457436221e-06,
|
394 |
+
"loss": 0.9384,
|
395 |
+
"step": 55
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 0.32230215827338127,
|
399 |
+
"grad_norm": 1.4485064367033296,
|
400 |
+
"learning_rate": 9.99898179843121e-06,
|
401 |
+
"loss": 0.9907,
|
402 |
+
"step": 56
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.32805755395683456,
|
406 |
+
"grad_norm": 1.3260833117821373,
|
407 |
+
"learning_rate": 9.998189911662793e-06,
|
408 |
+
"loss": 0.948,
|
409 |
+
"step": 57
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 0.3338129496402878,
|
413 |
+
"grad_norm": 1.4113520649113298,
|
414 |
+
"learning_rate": 9.99717183296762e-06,
|
415 |
+
"loss": 0.9162,
|
416 |
+
"step": 58
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 0.339568345323741,
|
420 |
+
"grad_norm": 1.3667363457488535,
|
421 |
+
"learning_rate": 9.995927608418612e-06,
|
422 |
+
"loss": 1.0227,
|
423 |
+
"step": 59
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.34532374100719426,
|
427 |
+
"grad_norm": 1.310530168350235,
|
428 |
+
"learning_rate": 9.994457294322858e-06,
|
429 |
+
"loss": 0.9603,
|
430 |
+
"step": 60
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 0.3510791366906475,
|
434 |
+
"grad_norm": 1.3061008761447361,
|
435 |
+
"learning_rate": 9.992760957219083e-06,
|
436 |
+
"loss": 0.9107,
|
437 |
+
"step": 61
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 0.35683453237410073,
|
441 |
+
"grad_norm": 1.3286055187667132,
|
442 |
+
"learning_rate": 9.99083867387463e-06,
|
443 |
+
"loss": 0.9662,
|
444 |
+
"step": 62
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.36258992805755397,
|
448 |
+
"grad_norm": 1.6686899683313348,
|
449 |
+
"learning_rate": 9.988690531281988e-06,
|
450 |
+
"loss": 1.0106,
|
451 |
+
"step": 63
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 0.3683453237410072,
|
455 |
+
"grad_norm": 1.3550028478795078,
|
456 |
+
"learning_rate": 9.986316626654851e-06,
|
457 |
+
"loss": 0.916,
|
458 |
+
"step": 64
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 0.37410071942446044,
|
462 |
+
"grad_norm": 1.3544747166387667,
|
463 |
+
"learning_rate": 9.983717067423721e-06,
|
464 |
+
"loss": 0.9686,
|
465 |
+
"step": 65
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.37985611510791367,
|
469 |
+
"grad_norm": 1.3034815135009086,
|
470 |
+
"learning_rate": 9.980891971231052e-06,
|
471 |
+
"loss": 0.9517,
|
472 |
+
"step": 66
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 0.3856115107913669,
|
476 |
+
"grad_norm": 1.3440784557117191,
|
477 |
+
"learning_rate": 9.977841465925918e-06,
|
478 |
+
"loss": 0.8906,
|
479 |
+
"step": 67
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 0.39136690647482014,
|
483 |
+
"grad_norm": 1.366486614179357,
|
484 |
+
"learning_rate": 9.974565689558228e-06,
|
485 |
+
"loss": 0.9203,
|
486 |
+
"step": 68
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 0.3971223021582734,
|
490 |
+
"grad_norm": 1.310162085554399,
|
491 |
+
"learning_rate": 9.971064790372484e-06,
|
492 |
+
"loss": 0.9555,
|
493 |
+
"step": 69
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 0.4028776978417266,
|
497 |
+
"grad_norm": 1.318452423313763,
|
498 |
+
"learning_rate": 9.967338926801066e-06,
|
499 |
+
"loss": 0.8692,
|
500 |
+
"step": 70
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 0.40863309352517985,
|
504 |
+
"grad_norm": 1.352699221858637,
|
505 |
+
"learning_rate": 9.963388267457071e-06,
|
506 |
+
"loss": 0.936,
|
507 |
+
"step": 71
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.4143884892086331,
|
511 |
+
"grad_norm": 1.2955644205150691,
|
512 |
+
"learning_rate": 9.959212991126668e-06,
|
513 |
+
"loss": 0.9147,
|
514 |
+
"step": 72
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 0.4201438848920863,
|
518 |
+
"grad_norm": 1.4384729311084175,
|
519 |
+
"learning_rate": 9.954813286761021e-06,
|
520 |
+
"loss": 0.9762,
|
521 |
+
"step": 73
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 0.42589928057553955,
|
525 |
+
"grad_norm": 1.3586324126567526,
|
526 |
+
"learning_rate": 9.950189353467735e-06,
|
527 |
+
"loss": 0.9518,
|
528 |
+
"step": 74
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 0.4316546762589928,
|
532 |
+
"grad_norm": 1.4263389759466811,
|
533 |
+
"learning_rate": 9.945341400501838e-06,
|
534 |
+
"loss": 0.9229,
|
535 |
+
"step": 75
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 0.437410071942446,
|
539 |
+
"grad_norm": 1.451019983241051,
|
540 |
+
"learning_rate": 9.940269647256319e-06,
|
541 |
+
"loss": 1.0027,
|
542 |
+
"step": 76
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 0.44316546762589926,
|
546 |
+
"grad_norm": 1.3330745257562855,
|
547 |
+
"learning_rate": 9.934974323252195e-06,
|
548 |
+
"loss": 1.1001,
|
549 |
+
"step": 77
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.4489208633093525,
|
553 |
+
"grad_norm": 1.2666099369384314,
|
554 |
+
"learning_rate": 9.929455668128129e-06,
|
555 |
+
"loss": 0.9906,
|
556 |
+
"step": 78
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 0.4546762589928058,
|
560 |
+
"grad_norm": 1.3144724815098685,
|
561 |
+
"learning_rate": 9.923713931629582e-06,
|
562 |
+
"loss": 0.879,
|
563 |
+
"step": 79
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 0.460431654676259,
|
567 |
+
"grad_norm": 1.5320188737475722,
|
568 |
+
"learning_rate": 9.917749373597506e-06,
|
569 |
+
"loss": 0.8488,
|
570 |
+
"step": 80
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 0.46618705035971225,
|
574 |
+
"grad_norm": 1.4069061862394436,
|
575 |
+
"learning_rate": 9.911562263956593e-06,
|
576 |
+
"loss": 0.9084,
|
577 |
+
"step": 81
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 0.4719424460431655,
|
581 |
+
"grad_norm": 1.4225341617119585,
|
582 |
+
"learning_rate": 9.90515288270306e-06,
|
583 |
+
"loss": 0.9151,
|
584 |
+
"step": 82
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 0.4776978417266187,
|
588 |
+
"grad_norm": 1.3710772687902628,
|
589 |
+
"learning_rate": 9.898521519891968e-06,
|
590 |
+
"loss": 0.9084,
|
591 |
+
"step": 83
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.48345323741007196,
|
595 |
+
"grad_norm": 1.3515027431388174,
|
596 |
+
"learning_rate": 9.891668475624106e-06,
|
597 |
+
"loss": 0.9244,
|
598 |
+
"step": 84
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 0.4892086330935252,
|
602 |
+
"grad_norm": 1.3243983662420342,
|
603 |
+
"learning_rate": 9.884594060032407e-06,
|
604 |
+
"loss": 0.8863,
|
605 |
+
"step": 85
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 0.4949640287769784,
|
609 |
+
"grad_norm": 1.279270882697488,
|
610 |
+
"learning_rate": 9.877298593267906e-06,
|
611 |
+
"loss": 0.8856,
|
612 |
+
"step": 86
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 0.5007194244604316,
|
616 |
+
"grad_norm": 1.3730223887321789,
|
617 |
+
"learning_rate": 9.869782405485267e-06,
|
618 |
+
"loss": 0.9544,
|
619 |
+
"step": 87
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 0.5064748201438849,
|
623 |
+
"grad_norm": 1.4007311823422817,
|
624 |
+
"learning_rate": 9.862045836827821e-06,
|
625 |
+
"loss": 0.9075,
|
626 |
+
"step": 88
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 0.5122302158273381,
|
630 |
+
"grad_norm": 1.3467957992553403,
|
631 |
+
"learning_rate": 9.854089237412194e-06,
|
632 |
+
"loss": 0.8764,
|
633 |
+
"step": 89
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.5179856115107914,
|
637 |
+
"grad_norm": 1.3122669099718975,
|
638 |
+
"learning_rate": 9.84591296731245e-06,
|
639 |
+
"loss": 0.9502,
|
640 |
+
"step": 90
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 0.5237410071942447,
|
644 |
+
"grad_norm": 1.407633558435018,
|
645 |
+
"learning_rate": 9.837517396543799e-06,
|
646 |
+
"loss": 1.076,
|
647 |
+
"step": 91
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 0.5294964028776978,
|
651 |
+
"grad_norm": 1.4296985813904106,
|
652 |
+
"learning_rate": 9.82890290504585e-06,
|
653 |
+
"loss": 0.9747,
|
654 |
+
"step": 92
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 0.5352517985611511,
|
658 |
+
"grad_norm": 1.5439086494632601,
|
659 |
+
"learning_rate": 9.82006988266542e-06,
|
660 |
+
"loss": 0.9277,
|
661 |
+
"step": 93
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 0.5410071942446043,
|
665 |
+
"grad_norm": 1.3497518679583498,
|
666 |
+
"learning_rate": 9.811018729138892e-06,
|
667 |
+
"loss": 0.8989,
|
668 |
+
"step": 94
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 0.5467625899280576,
|
672 |
+
"grad_norm": 1.3000278338647449,
|
673 |
+
"learning_rate": 9.801749854074122e-06,
|
674 |
+
"loss": 0.879,
|
675 |
+
"step": 95
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 0.5525179856115108,
|
679 |
+
"grad_norm": 1.2854617503902783,
|
680 |
+
"learning_rate": 9.792263676931906e-06,
|
681 |
+
"loss": 0.8718,
|
682 |
+
"step": 96
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 0.5582733812949641,
|
686 |
+
"grad_norm": 1.4371538615704396,
|
687 |
+
"learning_rate": 9.78256062700699e-06,
|
688 |
+
"loss": 0.8544,
|
689 |
+
"step": 97
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 0.5640287769784172,
|
693 |
+
"grad_norm": 1.6678335573993524,
|
694 |
+
"learning_rate": 9.772641143408652e-06,
|
695 |
+
"loss": 0.8972,
|
696 |
+
"step": 98
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 0.5697841726618705,
|
700 |
+
"grad_norm": 1.4707162618941683,
|
701 |
+
"learning_rate": 9.762505675040826e-06,
|
702 |
+
"loss": 0.9948,
|
703 |
+
"step": 99
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 0.5755395683453237,
|
707 |
+
"grad_norm": 1.302643605876999,
|
708 |
+
"learning_rate": 9.752154680581783e-06,
|
709 |
+
"loss": 0.8942,
|
710 |
+
"step": 100
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 0.581294964028777,
|
714 |
+
"grad_norm": 1.470400565479666,
|
715 |
+
"learning_rate": 9.741588628463384e-06,
|
716 |
+
"loss": 0.8772,
|
717 |
+
"step": 101
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 0.5870503597122302,
|
721 |
+
"grad_norm": 1.3198260122393806,
|
722 |
+
"learning_rate": 9.730807996849864e-06,
|
723 |
+
"loss": 0.9314,
|
724 |
+
"step": 102
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 0.5928057553956835,
|
728 |
+
"grad_norm": 1.38466874094164,
|
729 |
+
"learning_rate": 9.719813273616216e-06,
|
730 |
+
"loss": 0.9802,
|
731 |
+
"step": 103
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 0.5985611510791367,
|
735 |
+
"grad_norm": 1.413732387374318,
|
736 |
+
"learning_rate": 9.70860495632609e-06,
|
737 |
+
"loss": 0.927,
|
738 |
+
"step": 104
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.60431654676259,
|
742 |
+
"grad_norm": 1.432067970140142,
|
743 |
+
"learning_rate": 9.697183552209289e-06,
|
744 |
+
"loss": 0.924,
|
745 |
+
"step": 105
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.6100719424460431,
|
749 |
+
"grad_norm": 1.4225526979714325,
|
750 |
+
"learning_rate": 9.68554957813881e-06,
|
751 |
+
"loss": 0.8945,
|
752 |
+
"step": 106
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.6158273381294964,
|
756 |
+
"grad_norm": 1.3994215483741608,
|
757 |
+
"learning_rate": 9.673703560607459e-06,
|
758 |
+
"loss": 0.8816,
|
759 |
+
"step": 107
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.6215827338129496,
|
763 |
+
"grad_norm": 1.4081464385021556,
|
764 |
+
"learning_rate": 9.661646035704009e-06,
|
765 |
+
"loss": 0.9575,
|
766 |
+
"step": 108
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.6273381294964029,
|
770 |
+
"grad_norm": 1.3831998335104672,
|
771 |
+
"learning_rate": 9.649377549088962e-06,
|
772 |
+
"loss": 0.8911,
|
773 |
+
"step": 109
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.6330935251798561,
|
777 |
+
"grad_norm": 1.25039905684941,
|
778 |
+
"learning_rate": 9.636898655969837e-06,
|
779 |
+
"loss": 0.8356,
|
780 |
+
"step": 110
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.6388489208633094,
|
784 |
+
"grad_norm": 1.4566434256857157,
|
785 |
+
"learning_rate": 9.62420992107605e-06,
|
786 |
+
"loss": 0.9424,
|
787 |
+
"step": 111
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 0.6446043165467625,
|
791 |
+
"grad_norm": 1.4012572424386722,
|
792 |
+
"learning_rate": 9.61131191863336e-06,
|
793 |
+
"loss": 0.8761,
|
794 |
+
"step": 112
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 0.6503597122302158,
|
798 |
+
"grad_norm": 1.480887681928585,
|
799 |
+
"learning_rate": 9.598205232337882e-06,
|
800 |
+
"loss": 1.0254,
|
801 |
+
"step": 113
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.6561151079136691,
|
805 |
+
"grad_norm": 1.4151874921849987,
|
806 |
+
"learning_rate": 9.584890455329667e-06,
|
807 |
+
"loss": 0.8775,
|
808 |
+
"step": 114
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 0.6618705035971223,
|
812 |
+
"grad_norm": 1.4089124443779928,
|
813 |
+
"learning_rate": 9.571368190165865e-06,
|
814 |
+
"loss": 0.9254,
|
815 |
+
"step": 115
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 0.6676258992805756,
|
819 |
+
"grad_norm": 1.4775768113442738,
|
820 |
+
"learning_rate": 9.557639048793453e-06,
|
821 |
+
"loss": 0.9581,
|
822 |
+
"step": 116
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 0.6733812949640288,
|
826 |
+
"grad_norm": 1.456787265761407,
|
827 |
+
"learning_rate": 9.543703652521543e-06,
|
828 |
+
"loss": 0.8931,
|
829 |
+
"step": 117
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 0.679136690647482,
|
833 |
+
"grad_norm": 1.291261636243845,
|
834 |
+
"learning_rate": 9.52956263199327e-06,
|
835 |
+
"loss": 0.856,
|
836 |
+
"step": 118
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 0.6848920863309352,
|
840 |
+
"grad_norm": 1.3258126652625302,
|
841 |
+
"learning_rate": 9.515216627157238e-06,
|
842 |
+
"loss": 0.9335,
|
843 |
+
"step": 119
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 0.6906474820143885,
|
847 |
+
"grad_norm": 1.140309211145621,
|
848 |
+
"learning_rate": 9.500666287238573e-06,
|
849 |
+
"loss": 0.7234,
|
850 |
+
"step": 120
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 0.6964028776978417,
|
854 |
+
"grad_norm": 1.3100134672897092,
|
855 |
+
"learning_rate": 9.485912270709542e-06,
|
856 |
+
"loss": 0.8393,
|
857 |
+
"step": 121
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 0.702158273381295,
|
861 |
+
"grad_norm": 1.3927054037897126,
|
862 |
+
"learning_rate": 9.470955245259742e-06,
|
863 |
+
"loss": 0.9068,
|
864 |
+
"step": 122
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 0.7079136690647482,
|
868 |
+
"grad_norm": 1.4262606386994323,
|
869 |
+
"learning_rate": 9.455795887765896e-06,
|
870 |
+
"loss": 0.9699,
|
871 |
+
"step": 123
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 0.7136690647482015,
|
875 |
+
"grad_norm": 1.5371232043316245,
|
876 |
+
"learning_rate": 9.440434884261216e-06,
|
877 |
+
"loss": 0.9426,
|
878 |
+
"step": 124
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 0.7194244604316546,
|
882 |
+
"grad_norm": 1.4219147327655326,
|
883 |
+
"learning_rate": 9.424872929904359e-06,
|
884 |
+
"loss": 0.8158,
|
885 |
+
"step": 125
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 0.7251798561151079,
|
889 |
+
"grad_norm": 1.510311130458067,
|
890 |
+
"learning_rate": 9.409110728947964e-06,
|
891 |
+
"loss": 0.8735,
|
892 |
+
"step": 126
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 0.7309352517985611,
|
896 |
+
"grad_norm": 1.3416836724257437,
|
897 |
+
"learning_rate": 9.393148994706785e-06,
|
898 |
+
"loss": 0.8947,
|
899 |
+
"step": 127
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 0.7366906474820144,
|
903 |
+
"grad_norm": 1.3400081462001328,
|
904 |
+
"learning_rate": 9.376988449525405e-06,
|
905 |
+
"loss": 0.9162,
|
906 |
+
"step": 128
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 0.7424460431654676,
|
910 |
+
"grad_norm": 1.3322305008849582,
|
911 |
+
"learning_rate": 9.360629824745558e-06,
|
912 |
+
"loss": 0.8933,
|
913 |
+
"step": 129
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 0.7482014388489209,
|
917 |
+
"grad_norm": 1.2395774040774388,
|
918 |
+
"learning_rate": 9.344073860673016e-06,
|
919 |
+
"loss": 0.8792,
|
920 |
+
"step": 130
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 0.753956834532374,
|
924 |
+
"grad_norm": 1.2239837364251502,
|
925 |
+
"learning_rate": 9.327321306544097e-06,
|
926 |
+
"loss": 0.8591,
|
927 |
+
"step": 131
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 0.7597122302158273,
|
931 |
+
"grad_norm": 1.2867499707058148,
|
932 |
+
"learning_rate": 9.310372920491761e-06,
|
933 |
+
"loss": 0.9339,
|
934 |
+
"step": 132
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 0.7654676258992805,
|
938 |
+
"grad_norm": 1.2807449670165525,
|
939 |
+
"learning_rate": 9.293229469511293e-06,
|
940 |
+
"loss": 0.8689,
|
941 |
+
"step": 133
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 0.7712230215827338,
|
945 |
+
"grad_norm": 1.3771333806838677,
|
946 |
+
"learning_rate": 9.275891729425595e-06,
|
947 |
+
"loss": 0.8855,
|
948 |
+
"step": 134
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 0.7769784172661871,
|
952 |
+
"grad_norm": 1.4233446694683969,
|
953 |
+
"learning_rate": 9.25836048485008e-06,
|
954 |
+
"loss": 0.8467,
|
955 |
+
"step": 135
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 0.7827338129496403,
|
959 |
+
"grad_norm": 1.3675382510228975,
|
960 |
+
"learning_rate": 9.240636529157158e-06,
|
961 |
+
"loss": 0.9453,
|
962 |
+
"step": 136
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 0.7884892086330936,
|
966 |
+
"grad_norm": 1.3715839104877778,
|
967 |
+
"learning_rate": 9.22272066444034e-06,
|
968 |
+
"loss": 0.8845,
|
969 |
+
"step": 137
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 0.7942446043165468,
|
973 |
+
"grad_norm": 1.4095011875547687,
|
974 |
+
"learning_rate": 9.204613701477935e-06,
|
975 |
+
"loss": 0.9174,
|
976 |
+
"step": 138
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 0.8,
|
980 |
+
"grad_norm": 1.45550777383388,
|
981 |
+
"learning_rate": 9.186316459696359e-06,
|
982 |
+
"loss": 0.94,
|
983 |
+
"step": 139
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 0.8057553956834532,
|
987 |
+
"grad_norm": 1.3735083568361708,
|
988 |
+
"learning_rate": 9.167829767133047e-06,
|
989 |
+
"loss": 0.8102,
|
990 |
+
"step": 140
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 0.8115107913669065,
|
994 |
+
"grad_norm": 1.434021896165783,
|
995 |
+
"learning_rate": 9.149154460398993e-06,
|
996 |
+
"loss": 0.9431,
|
997 |
+
"step": 141
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 0.8172661870503597,
|
1001 |
+
"grad_norm": 1.404548232045249,
|
1002 |
+
"learning_rate": 9.130291384640873e-06,
|
1003 |
+
"loss": 0.9047,
|
1004 |
+
"step": 142
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 0.823021582733813,
|
1008 |
+
"grad_norm": 1.4285500048110893,
|
1009 |
+
"learning_rate": 9.111241393502814e-06,
|
1010 |
+
"loss": 0.8925,
|
1011 |
+
"step": 143
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 0.8287769784172662,
|
1015 |
+
"grad_norm": 1.358374683380466,
|
1016 |
+
"learning_rate": 9.092005349087754e-06,
|
1017 |
+
"loss": 0.8559,
|
1018 |
+
"step": 144
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 0.8345323741007195,
|
1022 |
+
"grad_norm": 1.2989466599781907,
|
1023 |
+
"learning_rate": 9.072584121918426e-06,
|
1024 |
+
"loss": 1.0086,
|
1025 |
+
"step": 145
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 0.8402877697841726,
|
1029 |
+
"grad_norm": 1.333962178961646,
|
1030 |
+
"learning_rate": 9.052978590897964e-06,
|
1031 |
+
"loss": 0.8812,
|
1032 |
+
"step": 146
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 0.8460431654676259,
|
1036 |
+
"grad_norm": 1.344601708602515,
|
1037 |
+
"learning_rate": 9.033189643270139e-06,
|
1038 |
+
"loss": 0.8282,
|
1039 |
+
"step": 147
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 0.8517985611510791,
|
1043 |
+
"grad_norm": 1.449391869268265,
|
1044 |
+
"learning_rate": 9.013218174579189e-06,
|
1045 |
+
"loss": 0.8924,
|
1046 |
+
"step": 148
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 0.8575539568345324,
|
1050 |
+
"grad_norm": 1.3678846029396583,
|
1051 |
+
"learning_rate": 8.993065088629304e-06,
|
1052 |
+
"loss": 0.8948,
|
1053 |
+
"step": 149
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 0.8633093525179856,
|
1057 |
+
"grad_norm": 1.448619911340056,
|
1058 |
+
"learning_rate": 8.972731297443722e-06,
|
1059 |
+
"loss": 0.9222,
|
1060 |
+
"step": 150
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 0.8690647482014389,
|
1064 |
+
"grad_norm": 1.4606172647273845,
|
1065 |
+
"learning_rate": 8.95221772122345e-06,
|
1066 |
+
"loss": 0.9079,
|
1067 |
+
"step": 151
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 0.874820143884892,
|
1071 |
+
"grad_norm": 1.2950313725539828,
|
1072 |
+
"learning_rate": 8.931525288305633e-06,
|
1073 |
+
"loss": 0.8233,
|
1074 |
+
"step": 152
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 0.8805755395683453,
|
1078 |
+
"grad_norm": 1.4135364456515347,
|
1079 |
+
"learning_rate": 8.910654935121528e-06,
|
1080 |
+
"loss": 0.9171,
|
1081 |
+
"step": 153
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 0.8863309352517985,
|
1085 |
+
"grad_norm": 1.3655216977120033,
|
1086 |
+
"learning_rate": 8.889607606154132e-06,
|
1087 |
+
"loss": 0.9143,
|
1088 |
+
"step": 154
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 0.8920863309352518,
|
1092 |
+
"grad_norm": 1.4635948034980835,
|
1093 |
+
"learning_rate": 8.868384253895445e-06,
|
1094 |
+
"loss": 0.9422,
|
1095 |
+
"step": 155
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 0.897841726618705,
|
1099 |
+
"grad_norm": 1.4270203579696321,
|
1100 |
+
"learning_rate": 8.846985838803357e-06,
|
1101 |
+
"loss": 0.9429,
|
1102 |
+
"step": 156
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 0.9035971223021583,
|
1106 |
+
"grad_norm": 1.3695179132446353,
|
1107 |
+
"learning_rate": 8.825413329258187e-06,
|
1108 |
+
"loss": 0.8393,
|
1109 |
+
"step": 157
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 0.9093525179856116,
|
1113 |
+
"grad_norm": 1.3710142713567868,
|
1114 |
+
"learning_rate": 8.803667701518857e-06,
|
1115 |
+
"loss": 0.8822,
|
1116 |
+
"step": 158
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 0.9151079136690647,
|
1120 |
+
"grad_norm": 1.337204406181296,
|
1121 |
+
"learning_rate": 8.781749939678712e-06,
|
1122 |
+
"loss": 0.8766,
|
1123 |
+
"step": 159
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 0.920863309352518,
|
1127 |
+
"grad_norm": 1.402362172978113,
|
1128 |
+
"learning_rate": 8.759661035620992e-06,
|
1129 |
+
"loss": 0.8679,
|
1130 |
+
"step": 160
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 0.9266187050359712,
|
1134 |
+
"grad_norm": 1.3931005034739852,
|
1135 |
+
"learning_rate": 8.73740198897393e-06,
|
1136 |
+
"loss": 0.9487,
|
1137 |
+
"step": 161
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 0.9323741007194245,
|
1141 |
+
"grad_norm": 1.3309735517216448,
|
1142 |
+
"learning_rate": 8.714973807065525e-06,
|
1143 |
+
"loss": 0.9371,
|
1144 |
+
"step": 162
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 0.9381294964028777,
|
1148 |
+
"grad_norm": 1.3551081328314547,
|
1149 |
+
"learning_rate": 8.69237750487796e-06,
|
1150 |
+
"loss": 0.8064,
|
1151 |
+
"step": 163
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 0.943884892086331,
|
1155 |
+
"grad_norm": 1.315828137071389,
|
1156 |
+
"learning_rate": 8.669614105001652e-06,
|
1157 |
+
"loss": 0.9083,
|
1158 |
+
"step": 164
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 0.9496402877697842,
|
1162 |
+
"grad_norm": 1.4259776336357026,
|
1163 |
+
"learning_rate": 8.646684637588992e-06,
|
1164 |
+
"loss": 0.9082,
|
1165 |
+
"step": 165
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 0.9553956834532374,
|
1169 |
+
"grad_norm": 1.3159838519083396,
|
1170 |
+
"learning_rate": 8.623590140307715e-06,
|
1171 |
+
"loss": 0.8174,
|
1172 |
+
"step": 166
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 0.9611510791366906,
|
1176 |
+
"grad_norm": 1.4252318231148138,
|
1177 |
+
"learning_rate": 8.600331658293948e-06,
|
1178 |
+
"loss": 0.8157,
|
1179 |
+
"step": 167
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 0.9669064748201439,
|
1183 |
+
"grad_norm": 1.2649634886770398,
|
1184 |
+
"learning_rate": 8.576910244104905e-06,
|
1185 |
+
"loss": 0.7305,
|
1186 |
+
"step": 168
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 0.9726618705035971,
|
1190 |
+
"grad_norm": 1.3307857336639481,
|
1191 |
+
"learning_rate": 8.553326957671264e-06,
|
1192 |
+
"loss": 0.8627,
|
1193 |
+
"step": 169
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 0.9784172661870504,
|
1197 |
+
"grad_norm": 1.3746350298770584,
|
1198 |
+
"learning_rate": 8.529582866249187e-06,
|
1199 |
+
"loss": 0.9088,
|
1200 |
+
"step": 170
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 0.9841726618705036,
|
1204 |
+
"grad_norm": 1.186119729031261,
|
1205 |
+
"learning_rate": 8.50567904437203e-06,
|
1206 |
+
"loss": 0.8252,
|
1207 |
+
"step": 171
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 0.9899280575539569,
|
1211 |
+
"grad_norm": 1.3942347828808497,
|
1212 |
+
"learning_rate": 8.48161657380172e-06,
|
1213 |
+
"loss": 1.0713,
|
1214 |
+
"step": 172
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 0.99568345323741,
|
1218 |
+
"grad_norm": 1.4889353800007157,
|
1219 |
+
"learning_rate": 8.457396543479787e-06,
|
1220 |
+
"loss": 0.9502,
|
1221 |
+
"step": 173
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 1.0,
|
1225 |
+
"grad_norm": 1.4889353800007157,
|
1226 |
+
"learning_rate": 8.433020049478093e-06,
|
1227 |
+
"loss": 0.9039,
|
1228 |
+
"step": 174
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 1.0057553956834533,
|
1232 |
+
"grad_norm": 1.6798612830096225,
|
1233 |
+
"learning_rate": 8.408488194949229e-06,
|
1234 |
+
"loss": 0.6914,
|
1235 |
+
"step": 175
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 1.0115107913669066,
|
1239 |
+
"grad_norm": 1.31253375768345,
|
1240 |
+
"learning_rate": 8.383802090076589e-06,
|
1241 |
+
"loss": 0.6407,
|
1242 |
+
"step": 176
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 1.0172661870503596,
|
1246 |
+
"grad_norm": 1.335156989683522,
|
1247 |
+
"learning_rate": 8.358962852024128e-06,
|
1248 |
+
"loss": 0.7788,
|
1249 |
+
"step": 177
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 1.023021582733813,
|
1253 |
+
"grad_norm": 1.316138582453117,
|
1254 |
+
"learning_rate": 8.333971604885817e-06,
|
1255 |
+
"loss": 0.7455,
|
1256 |
+
"step": 178
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 1.0287769784172662,
|
1260 |
+
"grad_norm": 1.3453499160735478,
|
1261 |
+
"learning_rate": 8.308829479634753e-06,
|
1262 |
+
"loss": 0.6916,
|
1263 |
+
"step": 179
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 1.0345323741007195,
|
1267 |
+
"grad_norm": 1.4175949421290355,
|
1268 |
+
"learning_rate": 8.283537614071987e-06,
|
1269 |
+
"loss": 0.7177,
|
1270 |
+
"step": 180
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 1.0402877697841726,
|
1274 |
+
"grad_norm": 1.3997457844106422,
|
1275 |
+
"learning_rate": 8.258097152775045e-06,
|
1276 |
+
"loss": 0.7206,
|
1277 |
+
"step": 181
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 1.0460431654676259,
|
1281 |
+
"grad_norm": 1.4597001846269448,
|
1282 |
+
"learning_rate": 8.232509247046106e-06,
|
1283 |
+
"loss": 0.7317,
|
1284 |
+
"step": 182
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 1.0517985611510792,
|
1288 |
+
"grad_norm": 1.4220254162855657,
|
1289 |
+
"learning_rate": 8.206775054859914e-06,
|
1290 |
+
"loss": 0.7968,
|
1291 |
+
"step": 183
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 1.0575539568345325,
|
1295 |
+
"grad_norm": 1.468378142441146,
|
1296 |
+
"learning_rate": 8.180895740811381e-06,
|
1297 |
+
"loss": 0.7372,
|
1298 |
+
"step": 184
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 1.0633093525179855,
|
1302 |
+
"grad_norm": 1.6116462753847913,
|
1303 |
+
"learning_rate": 8.154872476062868e-06,
|
1304 |
+
"loss": 0.7368,
|
1305 |
+
"step": 185
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 1.0690647482014388,
|
1309 |
+
"grad_norm": 1.5966503617319083,
|
1310 |
+
"learning_rate": 8.128706438291193e-06,
|
1311 |
+
"loss": 0.6986,
|
1312 |
+
"step": 186
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 1.074820143884892,
|
1316 |
+
"grad_norm": 1.449809087396639,
|
1317 |
+
"learning_rate": 8.102398811634338e-06,
|
1318 |
+
"loss": 0.6884,
|
1319 |
+
"step": 187
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 1.0805755395683454,
|
1323 |
+
"grad_norm": 1.3422183280174744,
|
1324 |
+
"learning_rate": 8.075950786637847e-06,
|
1325 |
+
"loss": 0.7078,
|
1326 |
+
"step": 188
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 1.0863309352517985,
|
1330 |
+
"grad_norm": 1.4785903233591355,
|
1331 |
+
"learning_rate": 8.049363560200972e-06,
|
1332 |
+
"loss": 0.6763,
|
1333 |
+
"step": 189
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 1.0920863309352518,
|
1337 |
+
"grad_norm": 1.3646811426232823,
|
1338 |
+
"learning_rate": 8.022638335522484e-06,
|
1339 |
+
"loss": 0.6993,
|
1340 |
+
"step": 190
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 1.097841726618705,
|
1344 |
+
"grad_norm": 1.4668294964957904,
|
1345 |
+
"learning_rate": 7.995776322046236e-06,
|
1346 |
+
"loss": 0.7091,
|
1347 |
+
"step": 191
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 1.1035971223021583,
|
1351 |
+
"grad_norm": 1.4282743394633324,
|
1352 |
+
"learning_rate": 7.968778735406426e-06,
|
1353 |
+
"loss": 0.7383,
|
1354 |
+
"step": 192
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 1.1093525179856114,
|
1358 |
+
"grad_norm": 1.376947997336659,
|
1359 |
+
"learning_rate": 7.941646797372584e-06,
|
1360 |
+
"loss": 0.6835,
|
1361 |
+
"step": 193
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 1.1151079136690647,
|
1365 |
+
"grad_norm": 1.4597078032376365,
|
1366 |
+
"learning_rate": 7.914381735794282e-06,
|
1367 |
+
"loss": 0.724,
|
1368 |
+
"step": 194
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 1.120863309352518,
|
1372 |
+
"grad_norm": 1.4554795555999986,
|
1373 |
+
"learning_rate": 7.886984784545565e-06,
|
1374 |
+
"loss": 0.7099,
|
1375 |
+
"step": 195
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 1.1266187050359713,
|
1379 |
+
"grad_norm": 1.4872751454433646,
|
1380 |
+
"learning_rate": 7.859457183469119e-06,
|
1381 |
+
"loss": 0.7086,
|
1382 |
+
"step": 196
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 1.1323741007194243,
|
1386 |
+
"grad_norm": 1.4920046904201234,
|
1387 |
+
"learning_rate": 7.831800178320153e-06,
|
1388 |
+
"loss": 0.7393,
|
1389 |
+
"step": 197
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 1.1381294964028776,
|
1393 |
+
"grad_norm": 1.3775281158227461,
|
1394 |
+
"learning_rate": 7.804015020710028e-06,
|
1395 |
+
"loss": 0.6799,
|
1396 |
+
"step": 198
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 1.143884892086331,
|
1400 |
+
"grad_norm": 1.4702918520493105,
|
1401 |
+
"learning_rate": 7.776102968049616e-06,
|
1402 |
+
"loss": 0.6896,
|
1403 |
+
"step": 199
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 1.1496402877697842,
|
1407 |
+
"grad_norm": 1.4499705425925438,
|
1408 |
+
"learning_rate": 7.748065283492397e-06,
|
1409 |
+
"loss": 0.698,
|
1410 |
+
"step": 200
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 1.1553956834532375,
|
1414 |
+
"grad_norm": 1.4183829329519066,
|
1415 |
+
"learning_rate": 7.719903235877289e-06,
|
1416 |
+
"loss": 0.6811,
|
1417 |
+
"step": 201
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 1.1611510791366906,
|
1421 |
+
"grad_norm": 1.5378274738136353,
|
1422 |
+
"learning_rate": 7.691618099671235e-06,
|
1423 |
+
"loss": 0.7107,
|
1424 |
+
"step": 202
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 1.1669064748201439,
|
1428 |
+
"grad_norm": 1.4705075983348321,
|
1429 |
+
"learning_rate": 7.663211154911523e-06,
|
1430 |
+
"loss": 0.6863,
|
1431 |
+
"step": 203
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 1.1726618705035972,
|
1435 |
+
"grad_norm": 1.535077368304991,
|
1436 |
+
"learning_rate": 7.634683687147857e-06,
|
1437 |
+
"loss": 0.6931,
|
1438 |
+
"step": 204
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 1.1784172661870504,
|
1442 |
+
"grad_norm": 1.3710808620962809,
|
1443 |
+
"learning_rate": 7.606036987384185e-06,
|
1444 |
+
"loss": 0.7119,
|
1445 |
+
"step": 205
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 1.1841726618705035,
|
1449 |
+
"grad_norm": 1.4645795424935797,
|
1450 |
+
"learning_rate": 7.577272352020269e-06,
|
1451 |
+
"loss": 0.7179,
|
1452 |
+
"step": 206
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 1.1899280575539568,
|
1456 |
+
"grad_norm": 1.4825292428823498,
|
1457 |
+
"learning_rate": 7.5483910827930186e-06,
|
1458 |
+
"loss": 0.7106,
|
1459 |
+
"step": 207
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 1.19568345323741,
|
1463 |
+
"grad_norm": 1.4413831174677387,
|
1464 |
+
"learning_rate": 7.519394486717583e-06,
|
1465 |
+
"loss": 0.719,
|
1466 |
+
"step": 208
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 1.2014388489208634,
|
1470 |
+
"grad_norm": 1.5294766891619676,
|
1471 |
+
"learning_rate": 7.4902838760282024e-06,
|
1472 |
+
"loss": 0.7704,
|
1473 |
+
"step": 209
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 1.2071942446043165,
|
1477 |
+
"grad_norm": 1.4456158639990568,
|
1478 |
+
"learning_rate": 7.461060568118822e-06,
|
1479 |
+
"loss": 0.8145,
|
1480 |
+
"step": 210
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 1.2129496402877697,
|
1484 |
+
"grad_norm": 1.5445683750570578,
|
1485 |
+
"learning_rate": 7.43172588548347e-06,
|
1486 |
+
"loss": 0.695,
|
1487 |
+
"step": 211
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 1.218705035971223,
|
1491 |
+
"grad_norm": 1.3925019584821468,
|
1492 |
+
"learning_rate": 7.402281155656415e-06,
|
1493 |
+
"loss": 0.6792,
|
1494 |
+
"step": 212
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 1.2244604316546763,
|
1498 |
+
"grad_norm": 1.3478822235613934,
|
1499 |
+
"learning_rate": 7.3727277111520864e-06,
|
1500 |
+
"loss": 0.6892,
|
1501 |
+
"step": 213
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 1.2302158273381294,
|
1505 |
+
"grad_norm": 1.3735278891419456,
|
1506 |
+
"learning_rate": 7.343066889404769e-06,
|
1507 |
+
"loss": 0.678,
|
1508 |
+
"step": 214
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 1.2359712230215827,
|
1512 |
+
"grad_norm": 1.4926513388871656,
|
1513 |
+
"learning_rate": 7.313300032708081e-06,
|
1514 |
+
"loss": 0.679,
|
1515 |
+
"step": 215
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 1.241726618705036,
|
1519 |
+
"grad_norm": 1.4731363022498112,
|
1520 |
+
"learning_rate": 7.283428488154227e-06,
|
1521 |
+
"loss": 0.695,
|
1522 |
+
"step": 216
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 1.2474820143884893,
|
1526 |
+
"grad_norm": 1.500614742977872,
|
1527 |
+
"learning_rate": 7.253453607573037e-06,
|
1528 |
+
"loss": 0.7413,
|
1529 |
+
"step": 217
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 1.2532374100719426,
|
1533 |
+
"grad_norm": 1.4162285947577158,
|
1534 |
+
"learning_rate": 7.223376747470792e-06,
|
1535 |
+
"loss": 0.664,
|
1536 |
+
"step": 218
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 1.2589928057553956,
|
1540 |
+
"grad_norm": 1.434606140992374,
|
1541 |
+
"learning_rate": 7.193199268968825e-06,
|
1542 |
+
"loss": 0.7102,
|
1543 |
+
"step": 219
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 1.264748201438849,
|
1547 |
+
"grad_norm": 1.4792752496267212,
|
1548 |
+
"learning_rate": 7.162922537741937e-06,
|
1549 |
+
"loss": 0.698,
|
1550 |
+
"step": 220
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 1.2705035971223022,
|
1554 |
+
"grad_norm": 1.4436615539172177,
|
1555 |
+
"learning_rate": 7.1325479239565875e-06,
|
1556 |
+
"loss": 0.7079,
|
1557 |
+
"step": 221
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 1.2762589928057553,
|
1561 |
+
"grad_norm": 1.2742897323885565,
|
1562 |
+
"learning_rate": 7.102076802208887e-06,
|
1563 |
+
"loss": 0.6929,
|
1564 |
+
"step": 222
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 1.2820143884892086,
|
1568 |
+
"grad_norm": 1.5060994956111649,
|
1569 |
+
"learning_rate": 7.071510551462395e-06,
|
1570 |
+
"loss": 0.7202,
|
1571 |
+
"step": 223
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 1.2877697841726619,
|
1575 |
+
"grad_norm": 1.3926138995024935,
|
1576 |
+
"learning_rate": 7.040850554985706e-06,
|
1577 |
+
"loss": 0.679,
|
1578 |
+
"step": 224
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 1.2935251798561151,
|
1582 |
+
"grad_norm": 1.4649859852076426,
|
1583 |
+
"learning_rate": 7.01009820028986e-06,
|
1584 |
+
"loss": 0.7292,
|
1585 |
+
"step": 225
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 1.2992805755395684,
|
1589 |
+
"grad_norm": 1.3567332620214738,
|
1590 |
+
"learning_rate": 6.9792548790655465e-06,
|
1591 |
+
"loss": 0.7093,
|
1592 |
+
"step": 226
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 1.3050359712230215,
|
1596 |
+
"grad_norm": 1.3997904231729403,
|
1597 |
+
"learning_rate": 6.948321987120122e-06,
|
1598 |
+
"loss": 0.6863,
|
1599 |
+
"step": 227
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 1.3107913669064748,
|
1603 |
+
"grad_norm": 1.4300378576908204,
|
1604 |
+
"learning_rate": 6.9173009243144485e-06,
|
1605 |
+
"loss": 0.6594,
|
1606 |
+
"step": 228
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 1.316546762589928,
|
1610 |
+
"grad_norm": 1.2875029006810836,
|
1611 |
+
"learning_rate": 6.886193094499537e-06,
|
1612 |
+
"loss": 0.8031,
|
1613 |
+
"step": 229
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 1.3223021582733812,
|
1617 |
+
"grad_norm": 1.5529040164830814,
|
1618 |
+
"learning_rate": 6.854999905453022e-06,
|
1619 |
+
"loss": 0.6892,
|
1620 |
+
"step": 230
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 1.3280575539568344,
|
1624 |
+
"grad_norm": 1.381833179233217,
|
1625 |
+
"learning_rate": 6.823722768815446e-06,
|
1626 |
+
"loss": 0.6961,
|
1627 |
+
"step": 231
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 1.3338129496402877,
|
1631 |
+
"grad_norm": 1.5873895447266677,
|
1632 |
+
"learning_rate": 6.792363100026383e-06,
|
1633 |
+
"loss": 0.7407,
|
1634 |
+
"step": 232
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 1.339568345323741,
|
1638 |
+
"grad_norm": 1.3407233041811095,
|
1639 |
+
"learning_rate": 6.760922318260384e-06,
|
1640 |
+
"loss": 0.6666,
|
1641 |
+
"step": 233
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 1.3453237410071943,
|
1645 |
+
"grad_norm": 1.4246874549295347,
|
1646 |
+
"learning_rate": 6.729401846362743e-06,
|
1647 |
+
"loss": 0.6779,
|
1648 |
+
"step": 234
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 1.3510791366906476,
|
1652 |
+
"grad_norm": 1.3088143655452398,
|
1653 |
+
"learning_rate": 6.697803110785115e-06,
|
1654 |
+
"loss": 0.6549,
|
1655 |
+
"step": 235
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 1.3568345323741007,
|
1659 |
+
"grad_norm": 1.3311189193840116,
|
1660 |
+
"learning_rate": 6.666127541520958e-06,
|
1661 |
+
"loss": 0.667,
|
1662 |
+
"step": 236
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 1.362589928057554,
|
1666 |
+
"grad_norm": 1.4754438818533273,
|
1667 |
+
"learning_rate": 6.634376572040826e-06,
|
1668 |
+
"loss": 0.694,
|
1669 |
+
"step": 237
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 1.3683453237410073,
|
1673 |
+
"grad_norm": 1.3946729568189524,
|
1674 |
+
"learning_rate": 6.602551639227486e-06,
|
1675 |
+
"loss": 0.6864,
|
1676 |
+
"step": 238
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 1.3741007194244603,
|
1680 |
+
"grad_norm": 1.430554189217815,
|
1681 |
+
"learning_rate": 6.570654183310901e-06,
|
1682 |
+
"loss": 0.6287,
|
1683 |
+
"step": 239
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 1.3798561151079136,
|
1687 |
+
"grad_norm": 1.4713282299603654,
|
1688 |
+
"learning_rate": 6.538685647803049e-06,
|
1689 |
+
"loss": 0.6384,
|
1690 |
+
"step": 240
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 1.385611510791367,
|
1694 |
+
"grad_norm": 1.3330699059988473,
|
1695 |
+
"learning_rate": 6.506647479432604e-06,
|
1696 |
+
"loss": 0.6997,
|
1697 |
+
"step": 241
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 1.3913669064748202,
|
1701 |
+
"grad_norm": 1.5722592344357025,
|
1702 |
+
"learning_rate": 6.474541128079452e-06,
|
1703 |
+
"loss": 0.6802,
|
1704 |
+
"step": 242
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 1.3971223021582735,
|
1708 |
+
"grad_norm": 1.382327839685819,
|
1709 |
+
"learning_rate": 6.442368046709087e-06,
|
1710 |
+
"loss": 0.7196,
|
1711 |
+
"step": 243
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 1.4028776978417266,
|
1715 |
+
"grad_norm": 1.424617522098925,
|
1716 |
+
"learning_rate": 6.410129691306855e-06,
|
1717 |
+
"loss": 0.6745,
|
1718 |
+
"step": 244
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 1.4086330935251798,
|
1722 |
+
"grad_norm": 1.4218642041517453,
|
1723 |
+
"learning_rate": 6.377827520812061e-06,
|
1724 |
+
"loss": 0.7232,
|
1725 |
+
"step": 245
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 1.4143884892086331,
|
1729 |
+
"grad_norm": 1.4497090959471781,
|
1730 |
+
"learning_rate": 6.34546299705195e-06,
|
1731 |
+
"loss": 0.667,
|
1732 |
+
"step": 246
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 1.4201438848920862,
|
1736 |
+
"grad_norm": 1.2830497516354644,
|
1737 |
+
"learning_rate": 6.31303758467555e-06,
|
1738 |
+
"loss": 0.7218,
|
1739 |
+
"step": 247
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 1.4258992805755395,
|
1743 |
+
"grad_norm": 1.5470928845192435,
|
1744 |
+
"learning_rate": 6.280552751087384e-06,
|
1745 |
+
"loss": 0.6958,
|
1746 |
+
"step": 248
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 1.4316546762589928,
|
1750 |
+
"grad_norm": 1.4446182883235026,
|
1751 |
+
"learning_rate": 6.248009966381074e-06,
|
1752 |
+
"loss": 0.6534,
|
1753 |
+
"step": 249
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 1.437410071942446,
|
1757 |
+
"grad_norm": 1.4452599392623058,
|
1758 |
+
"learning_rate": 6.215410703272805e-06,
|
1759 |
+
"loss": 0.7373,
|
1760 |
+
"step": 250
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 1.4431654676258994,
|
1764 |
+
"grad_norm": 1.5231983401485016,
|
1765 |
+
"learning_rate": 6.182756437034677e-06,
|
1766 |
+
"loss": 0.6619,
|
1767 |
+
"step": 251
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 1.4489208633093524,
|
1771 |
+
"grad_norm": 1.2778452998895347,
|
1772 |
+
"learning_rate": 6.150048645427944e-06,
|
1773 |
+
"loss": 0.6201,
|
1774 |
+
"step": 252
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 1.4546762589928057,
|
1778 |
+
"grad_norm": 1.4910613220644982,
|
1779 |
+
"learning_rate": 6.117288808636141e-06,
|
1780 |
+
"loss": 0.713,
|
1781 |
+
"step": 253
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 1.460431654676259,
|
1785 |
+
"grad_norm": 1.5479763621054092,
|
1786 |
+
"learning_rate": 6.0844784091980955e-06,
|
1787 |
+
"loss": 0.6734,
|
1788 |
+
"step": 254
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 1.4661870503597123,
|
1792 |
+
"grad_norm": 1.3078124819533825,
|
1793 |
+
"learning_rate": 6.05161893194083e-06,
|
1794 |
+
"loss": 0.6237,
|
1795 |
+
"step": 255
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 1.4719424460431654,
|
1799 |
+
"grad_norm": 1.3252656208229996,
|
1800 |
+
"learning_rate": 6.018711863912381e-06,
|
1801 |
+
"loss": 0.7034,
|
1802 |
+
"step": 256
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 1.4776978417266187,
|
1806 |
+
"grad_norm": 1.5886772866342525,
|
1807 |
+
"learning_rate": 5.985758694314486e-06,
|
1808 |
+
"loss": 0.6963,
|
1809 |
+
"step": 257
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 1.483453237410072,
|
1813 |
+
"grad_norm": 1.404885072529941,
|
1814 |
+
"learning_rate": 5.9527609144352075e-06,
|
1815 |
+
"loss": 0.6765,
|
1816 |
+
"step": 258
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 1.4892086330935252,
|
1820 |
+
"grad_norm": 1.5069873524127197,
|
1821 |
+
"learning_rate": 5.9197200175814275e-06,
|
1822 |
+
"loss": 0.7481,
|
1823 |
+
"step": 259
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 1.4949640287769785,
|
1827 |
+
"grad_norm": 1.5219996452651918,
|
1828 |
+
"learning_rate": 5.8866374990112785e-06,
|
1829 |
+
"loss": 0.7551,
|
1830 |
+
"step": 260
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 1.5007194244604316,
|
1834 |
+
"grad_norm": 1.4635117723937998,
|
1835 |
+
"learning_rate": 5.853514855866481e-06,
|
1836 |
+
"loss": 0.6504,
|
1837 |
+
"step": 261
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 1.506474820143885,
|
1841 |
+
"grad_norm": 1.2395981636603912,
|
1842 |
+
"learning_rate": 5.8203535871045735e-06,
|
1843 |
+
"loss": 0.7282,
|
1844 |
+
"step": 262
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 1.512230215827338,
|
1848 |
+
"grad_norm": 1.3537651975406249,
|
1849 |
+
"learning_rate": 5.787155193431095e-06,
|
1850 |
+
"loss": 0.6668,
|
1851 |
+
"step": 263
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 1.5179856115107913,
|
1855 |
+
"grad_norm": 1.3421598813855533,
|
1856 |
+
"learning_rate": 5.75392117723166e-06,
|
1857 |
+
"loss": 0.6983,
|
1858 |
+
"step": 264
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 1.5237410071942445,
|
1862 |
+
"grad_norm": 1.4346516234573874,
|
1863 |
+
"learning_rate": 5.7206530425039785e-06,
|
1864 |
+
"loss": 0.6951,
|
1865 |
+
"step": 265
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 1.5294964028776978,
|
1869 |
+
"grad_norm": 1.4494478051996553,
|
1870 |
+
"learning_rate": 5.687352294789776e-06,
|
1871 |
+
"loss": 0.6918,
|
1872 |
+
"step": 266
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 1.5352517985611511,
|
1876 |
+
"grad_norm": 1.3504174744187654,
|
1877 |
+
"learning_rate": 5.654020441106682e-06,
|
1878 |
+
"loss": 0.6611,
|
1879 |
+
"step": 267
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 1.5410071942446044,
|
1883 |
+
"grad_norm": 1.3538781587325135,
|
1884 |
+
"learning_rate": 5.620658989880011e-06,
|
1885 |
+
"loss": 0.6628,
|
1886 |
+
"step": 268
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 1.5467625899280577,
|
1890 |
+
"grad_norm": 1.411512244174525,
|
1891 |
+
"learning_rate": 5.587269450874513e-06,
|
1892 |
+
"loss": 0.6971,
|
1893 |
+
"step": 269
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 1.5525179856115108,
|
1897 |
+
"grad_norm": 1.349008078986391,
|
1898 |
+
"learning_rate": 5.5538533351260395e-06,
|
1899 |
+
"loss": 0.6436,
|
1900 |
+
"step": 270
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 1.558273381294964,
|
1904 |
+
"grad_norm": 1.3279310489321288,
|
1905 |
+
"learning_rate": 5.52041215487317e-06,
|
1906 |
+
"loss": 0.6373,
|
1907 |
+
"step": 271
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 1.5640287769784171,
|
1911 |
+
"grad_norm": 1.3665207548642804,
|
1912 |
+
"learning_rate": 5.486947423488774e-06,
|
1913 |
+
"loss": 0.6753,
|
1914 |
+
"step": 272
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 1.5697841726618704,
|
1918 |
+
"grad_norm": 1.458248740752457,
|
1919 |
+
"learning_rate": 5.453460655411515e-06,
|
1920 |
+
"loss": 0.7233,
|
1921 |
+
"step": 273
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 1.5755395683453237,
|
1925 |
+
"grad_norm": 1.4369995192876641,
|
1926 |
+
"learning_rate": 5.4199533660773276e-06,
|
1927 |
+
"loss": 0.7088,
|
1928 |
+
"step": 274
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 1.581294964028777,
|
1932 |
+
"grad_norm": 1.438567850892845,
|
1933 |
+
"learning_rate": 5.3864270718508305e-06,
|
1934 |
+
"loss": 0.6952,
|
1935 |
+
"step": 275
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 1.5870503597122303,
|
1939 |
+
"grad_norm": 1.3615776629356982,
|
1940 |
+
"learning_rate": 5.352883289956701e-06,
|
1941 |
+
"loss": 0.7004,
|
1942 |
+
"step": 276
|
1943 |
+
},
|
1944 |
+
{
|
1945 |
+
"epoch": 1.5928057553956836,
|
1946 |
+
"grad_norm": 1.5371553152731863,
|
1947 |
+
"learning_rate": 5.319323538411021e-06,
|
1948 |
+
"loss": 0.7215,
|
1949 |
+
"step": 277
|
1950 |
+
},
|
1951 |
+
{
|
1952 |
+
"epoch": 1.5985611510791367,
|
1953 |
+
"grad_norm": 1.4341816589262721,
|
1954 |
+
"learning_rate": 5.285749335952573e-06,
|
1955 |
+
"loss": 0.6228,
|
1956 |
+
"step": 278
|
1957 |
+
},
|
1958 |
+
{
|
1959 |
+
"epoch": 1.60431654676259,
|
1960 |
+
"grad_norm": 1.3936039107974154,
|
1961 |
+
"learning_rate": 5.252162201974112e-06,
|
1962 |
+
"loss": 0.7415,
|
1963 |
+
"step": 279
|
1964 |
+
},
|
1965 |
+
{
|
1966 |
+
"epoch": 1.610071942446043,
|
1967 |
+
"grad_norm": 1.5084811058835383,
|
1968 |
+
"learning_rate": 5.218563656453609e-06,
|
1969 |
+
"loss": 0.6251,
|
1970 |
+
"step": 280
|
1971 |
+
},
|
1972 |
+
{
|
1973 |
+
"epoch": 1.6158273381294963,
|
1974 |
+
"grad_norm": 1.2955259282307172,
|
1975 |
+
"learning_rate": 5.184955219885457e-06,
|
1976 |
+
"loss": 0.7588,
|
1977 |
+
"step": 281
|
1978 |
+
},
|
1979 |
+
{
|
1980 |
+
"epoch": 1.6215827338129496,
|
1981 |
+
"grad_norm": 1.5916282471378098,
|
1982 |
+
"learning_rate": 5.15133841321167e-06,
|
1983 |
+
"loss": 0.6935,
|
1984 |
+
"step": 282
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"epoch": 1.6273381294964029,
|
1988 |
+
"grad_norm": 1.4343361146249547,
|
1989 |
+
"learning_rate": 5.117714757753045e-06,
|
1990 |
+
"loss": 0.6358,
|
1991 |
+
"step": 283
|
1992 |
+
},
|
1993 |
+
{
|
1994 |
+
"epoch": 1.6330935251798562,
|
1995 |
+
"grad_norm": 1.3547405908559451,
|
1996 |
+
"learning_rate": 5.084085775140324e-06,
|
1997 |
+
"loss": 0.8555,
|
1998 |
+
"step": 284
|
1999 |
+
},
|
2000 |
+
{
|
2001 |
+
"epoch": 1.6388489208633095,
|
2002 |
+
"grad_norm": 1.458111017809787,
|
2003 |
+
"learning_rate": 5.050452987245325e-06,
|
2004 |
+
"loss": 0.7104,
|
2005 |
+
"step": 285
|
2006 |
+
},
|
2007 |
+
{
|
2008 |
+
"epoch": 1.6446043165467625,
|
2009 |
+
"grad_norm": 1.4623381005825575,
|
2010 |
+
"learning_rate": 5.016817916112075e-06,
|
2011 |
+
"loss": 0.8006,
|
2012 |
+
"step": 286
|
2013 |
+
},
|
2014 |
+
{
|
2015 |
+
"epoch": 1.6503597122302158,
|
2016 |
+
"grad_norm": 1.4456492353105892,
|
2017 |
+
"learning_rate": 4.9831820838879255e-06,
|
2018 |
+
"loss": 0.6964,
|
2019 |
+
"step": 287
|
2020 |
+
},
|
2021 |
+
{
|
2022 |
+
"epoch": 1.6561151079136691,
|
2023 |
+
"grad_norm": 1.4564555721924566,
|
2024 |
+
"learning_rate": 4.949547012754676e-06,
|
2025 |
+
"loss": 0.6571,
|
2026 |
+
"step": 288
|
2027 |
+
},
|
2028 |
+
{
|
2029 |
+
"epoch": 1.6618705035971222,
|
2030 |
+
"grad_norm": 1.5370289389801688,
|
2031 |
+
"learning_rate": 4.915914224859677e-06,
|
2032 |
+
"loss": 0.618,
|
2033 |
+
"step": 289
|
2034 |
+
},
|
2035 |
+
{
|
2036 |
+
"epoch": 1.6676258992805755,
|
2037 |
+
"grad_norm": 1.3657320775014004,
|
2038 |
+
"learning_rate": 4.882285242246958e-06,
|
2039 |
+
"loss": 0.6338,
|
2040 |
+
"step": 290
|
2041 |
+
},
|
2042 |
+
{
|
2043 |
+
"epoch": 1.6733812949640288,
|
2044 |
+
"grad_norm": 1.2766562303548328,
|
2045 |
+
"learning_rate": 4.848661586788334e-06,
|
2046 |
+
"loss": 0.8558,
|
2047 |
+
"step": 291
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"epoch": 1.679136690647482,
|
2051 |
+
"grad_norm": 1.4069208062576049,
|
2052 |
+
"learning_rate": 4.815044780114544e-06,
|
2053 |
+
"loss": 0.6766,
|
2054 |
+
"step": 292
|
2055 |
+
},
|
2056 |
+
{
|
2057 |
+
"epoch": 1.6848920863309353,
|
2058 |
+
"grad_norm": 1.4989329281948764,
|
2059 |
+
"learning_rate": 4.781436343546392e-06,
|
2060 |
+
"loss": 0.6638,
|
2061 |
+
"step": 293
|
2062 |
+
},
|
2063 |
+
{
|
2064 |
+
"epoch": 1.6906474820143886,
|
2065 |
+
"grad_norm": 1.494183220295251,
|
2066 |
+
"learning_rate": 4.7478377980258885e-06,
|
2067 |
+
"loss": 0.6481,
|
2068 |
+
"step": 294
|
2069 |
+
},
|
2070 |
+
{
|
2071 |
+
"epoch": 1.6964028776978417,
|
2072 |
+
"grad_norm": 1.423984952332584,
|
2073 |
+
"learning_rate": 4.714250664047428e-06,
|
2074 |
+
"loss": 0.6775,
|
2075 |
+
"step": 295
|
2076 |
+
},
|
2077 |
+
{
|
2078 |
+
"epoch": 1.702158273381295,
|
2079 |
+
"grad_norm": 1.4289428900724996,
|
2080 |
+
"learning_rate": 4.68067646158898e-06,
|
2081 |
+
"loss": 0.653,
|
2082 |
+
"step": 296
|
2083 |
+
},
|
2084 |
+
{
|
2085 |
+
"epoch": 1.707913669064748,
|
2086 |
+
"grad_norm": 1.3188329912324324,
|
2087 |
+
"learning_rate": 4.647116710043302e-06,
|
2088 |
+
"loss": 0.6251,
|
2089 |
+
"step": 297
|
2090 |
+
},
|
2091 |
+
{
|
2092 |
+
"epoch": 1.7136690647482014,
|
2093 |
+
"grad_norm": 1.3295724167564706,
|
2094 |
+
"learning_rate": 4.613572928149172e-06,
|
2095 |
+
"loss": 0.6956,
|
2096 |
+
"step": 298
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"epoch": 1.7194244604316546,
|
2100 |
+
"grad_norm": 1.3978339556557298,
|
2101 |
+
"learning_rate": 4.580046633922675e-06,
|
2102 |
+
"loss": 0.677,
|
2103 |
+
"step": 299
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 1.725179856115108,
|
2107 |
+
"grad_norm": 1.432656781239549,
|
2108 |
+
"learning_rate": 4.546539344588486e-06,
|
2109 |
+
"loss": 0.6711,
|
2110 |
+
"step": 300
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"epoch": 1.7309352517985612,
|
2114 |
+
"grad_norm": 1.4875448171153507,
|
2115 |
+
"learning_rate": 4.513052576511227e-06,
|
2116 |
+
"loss": 0.6968,
|
2117 |
+
"step": 301
|
2118 |
+
},
|
2119 |
+
{
|
2120 |
+
"epoch": 1.7366906474820145,
|
2121 |
+
"grad_norm": 1.3278447730989198,
|
2122 |
+
"learning_rate": 4.47958784512683e-06,
|
2123 |
+
"loss": 0.8098,
|
2124 |
+
"step": 302
|
2125 |
+
},
|
2126 |
+
{
|
2127 |
+
"epoch": 1.7424460431654676,
|
2128 |
+
"grad_norm": 1.5829332497081456,
|
2129 |
+
"learning_rate": 4.446146664873961e-06,
|
2130 |
+
"loss": 0.6577,
|
2131 |
+
"step": 303
|
2132 |
+
},
|
2133 |
+
{
|
2134 |
+
"epoch": 1.7482014388489209,
|
2135 |
+
"grad_norm": 1.39877901494755,
|
2136 |
+
"learning_rate": 4.41273054912549e-06,
|
2137 |
+
"loss": 0.6844,
|
2138 |
+
"step": 304
|
2139 |
+
},
|
2140 |
+
{
|
2141 |
+
"epoch": 1.753956834532374,
|
2142 |
+
"grad_norm": 1.369649981682954,
|
2143 |
+
"learning_rate": 4.379341010119992e-06,
|
2144 |
+
"loss": 0.6633,
|
2145 |
+
"step": 305
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 1.7597122302158272,
|
2149 |
+
"grad_norm": 1.3427524637685617,
|
2150 |
+
"learning_rate": 4.34597955889332e-06,
|
2151 |
+
"loss": 0.6536,
|
2152 |
+
"step": 306
|
2153 |
+
},
|
2154 |
+
{
|
2155 |
+
"epoch": 1.7654676258992805,
|
2156 |
+
"grad_norm": 1.3972329435269606,
|
2157 |
+
"learning_rate": 4.312647705210226e-06,
|
2158 |
+
"loss": 0.6902,
|
2159 |
+
"step": 307
|
2160 |
+
},
|
2161 |
+
{
|
2162 |
+
"epoch": 1.7712230215827338,
|
2163 |
+
"grad_norm": 1.400402688634292,
|
2164 |
+
"learning_rate": 4.279346957496023e-06,
|
2165 |
+
"loss": 0.8367,
|
2166 |
+
"step": 308
|
2167 |
+
},
|
2168 |
+
{
|
2169 |
+
"epoch": 1.776978417266187,
|
2170 |
+
"grad_norm": 1.5175223994658775,
|
2171 |
+
"learning_rate": 4.246078822768339e-06,
|
2172 |
+
"loss": 0.7799,
|
2173 |
+
"step": 309
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"epoch": 1.7827338129496404,
|
2177 |
+
"grad_norm": 1.3001941274257236,
|
2178 |
+
"learning_rate": 4.212844806568906e-06,
|
2179 |
+
"loss": 0.6799,
|
2180 |
+
"step": 310
|
2181 |
+
},
|
2182 |
+
{
|
2183 |
+
"epoch": 1.7884892086330937,
|
2184 |
+
"grad_norm": 1.4613629528040961,
|
2185 |
+
"learning_rate": 4.17964641289543e-06,
|
2186 |
+
"loss": 0.6607,
|
2187 |
+
"step": 311
|
2188 |
+
},
|
2189 |
+
{
|
2190 |
+
"epoch": 1.7942446043165468,
|
2191 |
+
"grad_norm": 1.480968313844926,
|
2192 |
+
"learning_rate": 4.1464851441335215e-06,
|
2193 |
+
"loss": 0.6865,
|
2194 |
+
"step": 312
|
2195 |
+
},
|
2196 |
+
{
|
2197 |
+
"epoch": 1.8,
|
2198 |
+
"grad_norm": 1.5094729089699839,
|
2199 |
+
"learning_rate": 4.113362500988722e-06,
|
2200 |
+
"loss": 0.8913,
|
2201 |
+
"step": 313
|
2202 |
+
},
|
2203 |
+
{
|
2204 |
+
"epoch": 1.8057553956834531,
|
2205 |
+
"grad_norm": 1.3840926954449198,
|
2206 |
+
"learning_rate": 4.080279982418574e-06,
|
2207 |
+
"loss": 0.6947,
|
2208 |
+
"step": 314
|
2209 |
+
},
|
2210 |
+
{
|
2211 |
+
"epoch": 1.8115107913669064,
|
2212 |
+
"grad_norm": 1.3104389246100092,
|
2213 |
+
"learning_rate": 4.047239085564794e-06,
|
2214 |
+
"loss": 0.6774,
|
2215 |
+
"step": 315
|
2216 |
+
},
|
2217 |
+
{
|
2218 |
+
"epoch": 1.8172661870503597,
|
2219 |
+
"grad_norm": 1.3946909612681881,
|
2220 |
+
"learning_rate": 4.014241305685514e-06,
|
2221 |
+
"loss": 0.6544,
|
2222 |
+
"step": 316
|
2223 |
+
},
|
2224 |
+
{
|
2225 |
+
"epoch": 1.823021582733813,
|
2226 |
+
"grad_norm": 1.4976088880952643,
|
2227 |
+
"learning_rate": 3.98128813608762e-06,
|
2228 |
+
"loss": 0.6861,
|
2229 |
+
"step": 317
|
2230 |
+
},
|
2231 |
+
{
|
2232 |
+
"epoch": 1.8287769784172663,
|
2233 |
+
"grad_norm": 1.4371234110900912,
|
2234 |
+
"learning_rate": 3.948381068059171e-06,
|
2235 |
+
"loss": 0.7021,
|
2236 |
+
"step": 318
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"epoch": 1.8345323741007196,
|
2240 |
+
"grad_norm": 1.467751157947971,
|
2241 |
+
"learning_rate": 3.915521590801907e-06,
|
2242 |
+
"loss": 0.6811,
|
2243 |
+
"step": 319
|
2244 |
+
},
|
2245 |
+
{
|
2246 |
+
"epoch": 1.8402877697841726,
|
2247 |
+
"grad_norm": 1.3710926689346201,
|
2248 |
+
"learning_rate": 3.88271119136386e-06,
|
2249 |
+
"loss": 0.7366,
|
2250 |
+
"step": 320
|
2251 |
+
},
|
2252 |
+
{
|
2253 |
+
"epoch": 1.846043165467626,
|
2254 |
+
"grad_norm": 1.3704335746432914,
|
2255 |
+
"learning_rate": 3.849951354572057e-06,
|
2256 |
+
"loss": 0.7691,
|
2257 |
+
"step": 321
|
2258 |
+
},
|
2259 |
+
{
|
2260 |
+
"epoch": 1.851798561151079,
|
2261 |
+
"grad_norm": 1.4280263023299677,
|
2262 |
+
"learning_rate": 3.817243562965324e-06,
|
2263 |
+
"loss": 0.6939,
|
2264 |
+
"step": 322
|
2265 |
+
},
|
2266 |
+
{
|
2267 |
+
"epoch": 1.8575539568345323,
|
2268 |
+
"grad_norm": 1.4008887769661036,
|
2269 |
+
"learning_rate": 3.7845892967271963e-06,
|
2270 |
+
"loss": 0.6895,
|
2271 |
+
"step": 323
|
2272 |
+
},
|
2273 |
+
{
|
2274 |
+
"epoch": 1.8633093525179856,
|
2275 |
+
"grad_norm": 1.4022305790500016,
|
2276 |
+
"learning_rate": 3.7519900336189265e-06,
|
2277 |
+
"loss": 0.6852,
|
2278 |
+
"step": 324
|
2279 |
+
},
|
2280 |
+
{
|
2281 |
+
"epoch": 1.8690647482014389,
|
2282 |
+
"grad_norm": 1.5457474277625742,
|
2283 |
+
"learning_rate": 3.7194472489126176e-06,
|
2284 |
+
"loss": 0.6993,
|
2285 |
+
"step": 325
|
2286 |
+
},
|
2287 |
+
{
|
2288 |
+
"epoch": 1.8748201438848922,
|
2289 |
+
"grad_norm": 1.3849147882634172,
|
2290 |
+
"learning_rate": 3.686962415324452e-06,
|
2291 |
+
"loss": 0.7417,
|
2292 |
+
"step": 326
|
2293 |
+
},
|
2294 |
+
{
|
2295 |
+
"epoch": 1.8805755395683454,
|
2296 |
+
"grad_norm": 1.2529362927618384,
|
2297 |
+
"learning_rate": 3.6545370029480515e-06,
|
2298 |
+
"loss": 0.5912,
|
2299 |
+
"step": 327
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"epoch": 1.8863309352517985,
|
2303 |
+
"grad_norm": 1.4044031839898943,
|
2304 |
+
"learning_rate": 3.6221724791879406e-06,
|
2305 |
+
"loss": 0.7581,
|
2306 |
+
"step": 328
|
2307 |
+
},
|
2308 |
+
{
|
2309 |
+
"epoch": 1.8920863309352518,
|
2310 |
+
"grad_norm": 1.573159256410031,
|
2311 |
+
"learning_rate": 3.5898703086931474e-06,
|
2312 |
+
"loss": 0.7051,
|
2313 |
+
"step": 329
|
2314 |
+
},
|
2315 |
+
{
|
2316 |
+
"epoch": 1.8978417266187049,
|
2317 |
+
"grad_norm": 1.3910007490173437,
|
2318 |
+
"learning_rate": 3.557631953290914e-06,
|
2319 |
+
"loss": 0.7636,
|
2320 |
+
"step": 330
|
2321 |
+
},
|
2322 |
+
{
|
2323 |
+
"epoch": 1.9035971223021582,
|
2324 |
+
"grad_norm": 1.4522801133071914,
|
2325 |
+
"learning_rate": 3.5254588719205494e-06,
|
2326 |
+
"loss": 0.6732,
|
2327 |
+
"step": 331
|
2328 |
+
},
|
2329 |
+
{
|
2330 |
+
"epoch": 1.9093525179856115,
|
2331 |
+
"grad_norm": 1.4462460568757962,
|
2332 |
+
"learning_rate": 3.4933525205673977e-06,
|
2333 |
+
"loss": 0.6288,
|
2334 |
+
"step": 332
|
2335 |
+
},
|
2336 |
+
{
|
2337 |
+
"epoch": 1.9151079136690647,
|
2338 |
+
"grad_norm": 1.541230257971578,
|
2339 |
+
"learning_rate": 3.461314352196952e-06,
|
2340 |
+
"loss": 0.6892,
|
2341 |
+
"step": 333
|
2342 |
+
},
|
2343 |
+
{
|
2344 |
+
"epoch": 1.920863309352518,
|
2345 |
+
"grad_norm": 1.8431712034226,
|
2346 |
+
"learning_rate": 3.429345816689101e-06,
|
2347 |
+
"loss": 0.7457,
|
2348 |
+
"step": 334
|
2349 |
+
},
|
2350 |
+
{
|
2351 |
+
"epoch": 1.9266187050359713,
|
2352 |
+
"grad_norm": 1.4553623292641784,
|
2353 |
+
"learning_rate": 3.397448360772516e-06,
|
2354 |
+
"loss": 0.6886,
|
2355 |
+
"step": 335
|
2356 |
+
},
|
2357 |
+
{
|
2358 |
+
"epoch": 1.9323741007194246,
|
2359 |
+
"grad_norm": 1.4348609039293514,
|
2360 |
+
"learning_rate": 3.365623427959175e-06,
|
2361 |
+
"loss": 0.6978,
|
2362 |
+
"step": 336
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"epoch": 1.9381294964028777,
|
2366 |
+
"grad_norm": 1.415153427148624,
|
2367 |
+
"learning_rate": 3.3338724584790427e-06,
|
2368 |
+
"loss": 0.6482,
|
2369 |
+
"step": 337
|
2370 |
+
},
|
2371 |
+
{
|
2372 |
+
"epoch": 1.943884892086331,
|
2373 |
+
"grad_norm": 1.4083736160826446,
|
2374 |
+
"learning_rate": 3.302196889214886e-06,
|
2375 |
+
"loss": 0.6818,
|
2376 |
+
"step": 338
|
2377 |
+
},
|
2378 |
+
{
|
2379 |
+
"epoch": 1.949640287769784,
|
2380 |
+
"grad_norm": 1.3929257947767029,
|
2381 |
+
"learning_rate": 3.270598153637259e-06,
|
2382 |
+
"loss": 0.659,
|
2383 |
+
"step": 339
|
2384 |
+
},
|
2385 |
+
{
|
2386 |
+
"epoch": 1.9553956834532373,
|
2387 |
+
"grad_norm": 1.375977818983056,
|
2388 |
+
"learning_rate": 3.239077681739618e-06,
|
2389 |
+
"loss": 0.6659,
|
2390 |
+
"step": 340
|
2391 |
+
},
|
2392 |
+
{
|
2393 |
+
"epoch": 1.9611510791366906,
|
2394 |
+
"grad_norm": 1.458008636421117,
|
2395 |
+
"learning_rate": 3.2076368999736175e-06,
|
2396 |
+
"loss": 0.5983,
|
2397 |
+
"step": 341
|
2398 |
+
},
|
2399 |
+
{
|
2400 |
+
"epoch": 1.966906474820144,
|
2401 |
+
"grad_norm": 1.2172356556264106,
|
2402 |
+
"learning_rate": 3.176277231184556e-06,
|
2403 |
+
"loss": 0.6545,
|
2404 |
+
"step": 342
|
2405 |
+
},
|
2406 |
+
{
|
2407 |
+
"epoch": 1.9726618705035972,
|
2408 |
+
"grad_norm": 1.456654804522519,
|
2409 |
+
"learning_rate": 3.14500009454698e-06,
|
2410 |
+
"loss": 0.6875,
|
2411 |
+
"step": 343
|
2412 |
+
},
|
2413 |
+
{
|
2414 |
+
"epoch": 1.9784172661870505,
|
2415 |
+
"grad_norm": 1.3545712295454608,
|
2416 |
+
"learning_rate": 3.1138069055004628e-06,
|
2417 |
+
"loss": 0.6699,
|
2418 |
+
"step": 344
|
2419 |
+
},
|
2420 |
+
{
|
2421 |
+
"epoch": 1.9841726618705036,
|
2422 |
+
"grad_norm": 1.4922047149305653,
|
2423 |
+
"learning_rate": 3.0826990756855528e-06,
|
2424 |
+
"loss": 0.622,
|
2425 |
+
"step": 345
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"epoch": 1.9899280575539569,
|
2429 |
+
"grad_norm": 1.3097489620471516,
|
2430 |
+
"learning_rate": 3.0516780128798794e-06,
|
2431 |
+
"loss": 0.655,
|
2432 |
+
"step": 346
|
2433 |
+
},
|
2434 |
+
{
|
2435 |
+
"epoch": 1.99568345323741,
|
2436 |
+
"grad_norm": 1.5030309234716739,
|
2437 |
+
"learning_rate": 3.020745120934455e-06,
|
2438 |
+
"loss": 0.6965,
|
2439 |
+
"step": 347
|
2440 |
+
},
|
2441 |
+
{
|
2442 |
+
"epoch": 2.0,
|
2443 |
+
"grad_norm": 1.5799097018498731,
|
2444 |
+
"learning_rate": 2.989901799710142e-06,
|
2445 |
+
"loss": 0.6721,
|
2446 |
+
"step": 348
|
2447 |
+
},
|
2448 |
+
{
|
2449 |
+
"epoch": 2.0057553956834533,
|
2450 |
+
"grad_norm": 1.5384543552447154,
|
2451 |
+
"learning_rate": 2.9591494450142957e-06,
|
2452 |
+
"loss": 0.5199,
|
2453 |
+
"step": 349
|
2454 |
+
},
|
2455 |
+
{
|
2456 |
+
"epoch": 2.0115107913669066,
|
2457 |
+
"grad_norm": 1.4011961151039625,
|
2458 |
+
"learning_rate": 2.9284894485376057e-06,
|
2459 |
+
"loss": 0.5154,
|
2460 |
+
"step": 350
|
2461 |
+
},
|
2462 |
+
{
|
2463 |
+
"epoch": 2.01726618705036,
|
2464 |
+
"grad_norm": 1.444716011331813,
|
2465 |
+
"learning_rate": 2.8979231977911127e-06,
|
2466 |
+
"loss": 0.5543,
|
2467 |
+
"step": 351
|
2468 |
+
},
|
2469 |
+
{
|
2470 |
+
"epoch": 2.023021582733813,
|
2471 |
+
"grad_norm": 1.414264346010769,
|
2472 |
+
"learning_rate": 2.8674520760434145e-06,
|
2473 |
+
"loss": 0.4366,
|
2474 |
+
"step": 352
|
2475 |
+
},
|
2476 |
+
{
|
2477 |
+
"epoch": 2.028776978417266,
|
2478 |
+
"grad_norm": 1.3161032263490933,
|
2479 |
+
"learning_rate": 2.8370774622580644e-06,
|
2480 |
+
"loss": 0.5227,
|
2481 |
+
"step": 353
|
2482 |
+
},
|
2483 |
+
{
|
2484 |
+
"epoch": 2.0345323741007193,
|
2485 |
+
"grad_norm": 1.32708738719698,
|
2486 |
+
"learning_rate": 2.806800731031176e-06,
|
2487 |
+
"loss": 0.502,
|
2488 |
+
"step": 354
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"epoch": 2.0402877697841726,
|
2492 |
+
"grad_norm": 1.3858313142314658,
|
2493 |
+
"learning_rate": 2.7766232525292104e-06,
|
2494 |
+
"loss": 0.5378,
|
2495 |
+
"step": 355
|
2496 |
+
},
|
2497 |
+
{
|
2498 |
+
"epoch": 2.046043165467626,
|
2499 |
+
"grad_norm": 1.3504357752921585,
|
2500 |
+
"learning_rate": 2.746546392426963e-06,
|
2501 |
+
"loss": 0.5947,
|
2502 |
+
"step": 356
|
2503 |
+
},
|
2504 |
+
{
|
2505 |
+
"epoch": 2.051798561151079,
|
2506 |
+
"grad_norm": 1.456982482394388,
|
2507 |
+
"learning_rate": 2.7165715118457735e-06,
|
2508 |
+
"loss": 0.5277,
|
2509 |
+
"step": 357
|
2510 |
+
},
|
2511 |
+
{
|
2512 |
+
"epoch": 2.0575539568345325,
|
2513 |
+
"grad_norm": 1.4183682403259945,
|
2514 |
+
"learning_rate": 2.6866999672919198e-06,
|
2515 |
+
"loss": 0.464,
|
2516 |
+
"step": 358
|
2517 |
+
},
|
2518 |
+
{
|
2519 |
+
"epoch": 2.0633093525179858,
|
2520 |
+
"grad_norm": 1.453357981571888,
|
2521 |
+
"learning_rate": 2.656933110595233e-06,
|
2522 |
+
"loss": 0.4613,
|
2523 |
+
"step": 359
|
2524 |
+
},
|
2525 |
+
{
|
2526 |
+
"epoch": 2.069064748201439,
|
2527 |
+
"grad_norm": 1.4318321758549342,
|
2528 |
+
"learning_rate": 2.6272722888479152e-06,
|
2529 |
+
"loss": 0.5465,
|
2530 |
+
"step": 360
|
2531 |
+
},
|
2532 |
+
{
|
2533 |
+
"epoch": 2.074820143884892,
|
2534 |
+
"grad_norm": 1.5705214998819301,
|
2535 |
+
"learning_rate": 2.5977188443435874e-06,
|
2536 |
+
"loss": 0.5154,
|
2537 |
+
"step": 361
|
2538 |
+
},
|
2539 |
+
{
|
2540 |
+
"epoch": 2.080575539568345,
|
2541 |
+
"grad_norm": 1.5524107276796024,
|
2542 |
+
"learning_rate": 2.5682741145165325e-06,
|
2543 |
+
"loss": 0.4975,
|
2544 |
+
"step": 362
|
2545 |
+
},
|
2546 |
+
{
|
2547 |
+
"epoch": 2.0863309352517985,
|
2548 |
+
"grad_norm": 1.4860485267619432,
|
2549 |
+
"learning_rate": 2.53893943188118e-06,
|
2550 |
+
"loss": 0.4933,
|
2551 |
+
"step": 363
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"epoch": 2.0920863309352518,
|
2555 |
+
"grad_norm": 1.441231632693477,
|
2556 |
+
"learning_rate": 2.5097161239717975e-06,
|
2557 |
+
"loss": 0.5313,
|
2558 |
+
"step": 364
|
2559 |
+
},
|
2560 |
+
{
|
2561 |
+
"epoch": 2.097841726618705,
|
2562 |
+
"grad_norm": 1.566503611865734,
|
2563 |
+
"learning_rate": 2.4806055132824186e-06,
|
2564 |
+
"loss": 0.58,
|
2565 |
+
"step": 365
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 2.1035971223021583,
|
2569 |
+
"grad_norm": 1.4535179790407338,
|
2570 |
+
"learning_rate": 2.4516089172069852e-06,
|
2571 |
+
"loss": 0.4474,
|
2572 |
+
"step": 366
|
2573 |
+
},
|
2574 |
+
{
|
2575 |
+
"epoch": 2.1093525179856116,
|
2576 |
+
"grad_norm": 1.4065944445349983,
|
2577 |
+
"learning_rate": 2.422727647979734e-06,
|
2578 |
+
"loss": 0.4819,
|
2579 |
+
"step": 367
|
2580 |
+
},
|
2581 |
+
{
|
2582 |
+
"epoch": 2.115107913669065,
|
2583 |
+
"grad_norm": 1.4524409945045922,
|
2584 |
+
"learning_rate": 2.393963012615817e-06,
|
2585 |
+
"loss": 0.5624,
|
2586 |
+
"step": 368
|
2587 |
+
},
|
2588 |
+
{
|
2589 |
+
"epoch": 2.1208633093525178,
|
2590 |
+
"grad_norm": 1.5459335717499265,
|
2591 |
+
"learning_rate": 2.365316312852144e-06,
|
2592 |
+
"loss": 0.4784,
|
2593 |
+
"step": 369
|
2594 |
+
},
|
2595 |
+
{
|
2596 |
+
"epoch": 2.126618705035971,
|
2597 |
+
"grad_norm": 1.4014141194810679,
|
2598 |
+
"learning_rate": 2.336788845088478e-06,
|
2599 |
+
"loss": 0.5462,
|
2600 |
+
"step": 370
|
2601 |
+
},
|
2602 |
+
{
|
2603 |
+
"epoch": 2.1323741007194243,
|
2604 |
+
"grad_norm": 1.4442032553726138,
|
2605 |
+
"learning_rate": 2.308381900328767e-06,
|
2606 |
+
"loss": 0.4866,
|
2607 |
+
"step": 371
|
2608 |
+
},
|
2609 |
+
{
|
2610 |
+
"epoch": 2.1381294964028776,
|
2611 |
+
"grad_norm": 1.4004501234195819,
|
2612 |
+
"learning_rate": 2.2800967641227127e-06,
|
2613 |
+
"loss": 0.5739,
|
2614 |
+
"step": 372
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"epoch": 2.143884892086331,
|
2618 |
+
"grad_norm": 1.3604420713760152,
|
2619 |
+
"learning_rate": 2.2519347165076067e-06,
|
2620 |
+
"loss": 0.4577,
|
2621 |
+
"step": 373
|
2622 |
+
},
|
2623 |
+
{
|
2624 |
+
"epoch": 2.149640287769784,
|
2625 |
+
"grad_norm": 1.39364908972241,
|
2626 |
+
"learning_rate": 2.223897031950386e-06,
|
2627 |
+
"loss": 0.4845,
|
2628 |
+
"step": 374
|
2629 |
+
},
|
2630 |
+
{
|
2631 |
+
"epoch": 2.1553956834532375,
|
2632 |
+
"grad_norm": 1.4226690517141671,
|
2633 |
+
"learning_rate": 2.195984979289974e-06,
|
2634 |
+
"loss": 0.5107,
|
2635 |
+
"step": 375
|
2636 |
+
},
|
2637 |
+
{
|
2638 |
+
"epoch": 2.161151079136691,
|
2639 |
+
"grad_norm": 1.415708692108885,
|
2640 |
+
"learning_rate": 2.1681998216798476e-06,
|
2641 |
+
"loss": 0.5247,
|
2642 |
+
"step": 376
|
2643 |
+
},
|
2644 |
+
{
|
2645 |
+
"epoch": 2.166906474820144,
|
2646 |
+
"grad_norm": 1.479676456879866,
|
2647 |
+
"learning_rate": 2.140542816530882e-06,
|
2648 |
+
"loss": 0.4759,
|
2649 |
+
"step": 377
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 2.172661870503597,
|
2653 |
+
"grad_norm": 1.3414595432361576,
|
2654 |
+
"learning_rate": 2.1130152154544346e-06,
|
2655 |
+
"loss": 0.4769,
|
2656 |
+
"step": 378
|
2657 |
+
},
|
2658 |
+
{
|
2659 |
+
"epoch": 2.1784172661870502,
|
2660 |
+
"grad_norm": 1.4580316458169462,
|
2661 |
+
"learning_rate": 2.0856182642057182e-06,
|
2662 |
+
"loss": 0.5102,
|
2663 |
+
"step": 379
|
2664 |
+
},
|
2665 |
+
{
|
2666 |
+
"epoch": 2.1841726618705035,
|
2667 |
+
"grad_norm": 1.373119247134508,
|
2668 |
+
"learning_rate": 2.058353202627417e-06,
|
2669 |
+
"loss": 0.4789,
|
2670 |
+
"step": 380
|
2671 |
+
},
|
2672 |
+
{
|
2673 |
+
"epoch": 2.189928057553957,
|
2674 |
+
"grad_norm": 1.4636158497300926,
|
2675 |
+
"learning_rate": 2.0312212645935755e-06,
|
2676 |
+
"loss": 0.478,
|
2677 |
+
"step": 381
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"epoch": 2.19568345323741,
|
2681 |
+
"grad_norm": 1.4373357551520838,
|
2682 |
+
"learning_rate": 2.0042236779537668e-06,
|
2683 |
+
"loss": 0.5163,
|
2684 |
+
"step": 382
|
2685 |
+
},
|
2686 |
+
{
|
2687 |
+
"epoch": 2.2014388489208634,
|
2688 |
+
"grad_norm": 1.4927332947970118,
|
2689 |
+
"learning_rate": 1.977361664477518e-06,
|
2690 |
+
"loss": 0.4759,
|
2691 |
+
"step": 383
|
2692 |
+
},
|
2693 |
+
{
|
2694 |
+
"epoch": 2.2071942446043167,
|
2695 |
+
"grad_norm": 1.5499545436337445,
|
2696 |
+
"learning_rate": 1.950636439799029e-06,
|
2697 |
+
"loss": 0.5309,
|
2698 |
+
"step": 384
|
2699 |
+
},
|
2700 |
+
{
|
2701 |
+
"epoch": 2.21294964028777,
|
2702 |
+
"grad_norm": 1.3978268111661183,
|
2703 |
+
"learning_rate": 1.924049213362153e-06,
|
2704 |
+
"loss": 0.6753,
|
2705 |
+
"step": 385
|
2706 |
+
},
|
2707 |
+
{
|
2708 |
+
"epoch": 2.218705035971223,
|
2709 |
+
"grad_norm": 1.4362350584554568,
|
2710 |
+
"learning_rate": 1.8976011883656632e-06,
|
2711 |
+
"loss": 0.4753,
|
2712 |
+
"step": 386
|
2713 |
+
},
|
2714 |
+
{
|
2715 |
+
"epoch": 2.224460431654676,
|
2716 |
+
"grad_norm": 1.3557989613139372,
|
2717 |
+
"learning_rate": 1.8712935617088067e-06,
|
2718 |
+
"loss": 0.4556,
|
2719 |
+
"step": 387
|
2720 |
+
},
|
2721 |
+
{
|
2722 |
+
"epoch": 2.2302158273381294,
|
2723 |
+
"grad_norm": 1.4623450607022541,
|
2724 |
+
"learning_rate": 1.8451275239371337e-06,
|
2725 |
+
"loss": 0.5515,
|
2726 |
+
"step": 388
|
2727 |
+
},
|
2728 |
+
{
|
2729 |
+
"epoch": 2.2359712230215827,
|
2730 |
+
"grad_norm": 1.4460952727747438,
|
2731 |
+
"learning_rate": 1.8191042591886198e-06,
|
2732 |
+
"loss": 0.5036,
|
2733 |
+
"step": 389
|
2734 |
+
},
|
2735 |
+
{
|
2736 |
+
"epoch": 2.241726618705036,
|
2737 |
+
"grad_norm": 1.415595236688259,
|
2738 |
+
"learning_rate": 1.7932249451400863e-06,
|
2739 |
+
"loss": 0.436,
|
2740 |
+
"step": 390
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"epoch": 2.2474820143884893,
|
2744 |
+
"grad_norm": 1.409991217908193,
|
2745 |
+
"learning_rate": 1.767490752953896e-06,
|
2746 |
+
"loss": 0.5105,
|
2747 |
+
"step": 391
|
2748 |
+
},
|
2749 |
+
{
|
2750 |
+
"epoch": 2.2532374100719426,
|
2751 |
+
"grad_norm": 1.5073919735744494,
|
2752 |
+
"learning_rate": 1.7419028472249566e-06,
|
2753 |
+
"loss": 0.5342,
|
2754 |
+
"step": 392
|
2755 |
+
},
|
2756 |
+
{
|
2757 |
+
"epoch": 2.258992805755396,
|
2758 |
+
"grad_norm": 1.3929836227407455,
|
2759 |
+
"learning_rate": 1.7164623859280144e-06,
|
2760 |
+
"loss": 0.5287,
|
2761 |
+
"step": 393
|
2762 |
+
},
|
2763 |
+
{
|
2764 |
+
"epoch": 2.2647482014388487,
|
2765 |
+
"grad_norm": 1.3202392645680905,
|
2766 |
+
"learning_rate": 1.6911705203652506e-06,
|
2767 |
+
"loss": 0.5356,
|
2768 |
+
"step": 394
|
2769 |
+
},
|
2770 |
+
{
|
2771 |
+
"epoch": 2.270503597122302,
|
2772 |
+
"grad_norm": 1.4129051228121443,
|
2773 |
+
"learning_rate": 1.6660283951141847e-06,
|
2774 |
+
"loss": 0.5275,
|
2775 |
+
"step": 395
|
2776 |
+
},
|
2777 |
+
{
|
2778 |
+
"epoch": 2.2762589928057553,
|
2779 |
+
"grad_norm": 1.4449125587756706,
|
2780 |
+
"learning_rate": 1.641037147975872e-06,
|
2781 |
+
"loss": 0.4799,
|
2782 |
+
"step": 396
|
2783 |
+
},
|
2784 |
+
{
|
2785 |
+
"epoch": 2.2820143884892086,
|
2786 |
+
"grad_norm": 1.3373010588883805,
|
2787 |
+
"learning_rate": 1.616197909923412e-06,
|
2788 |
+
"loss": 0.4607,
|
2789 |
+
"step": 397
|
2790 |
+
},
|
2791 |
+
{
|
2792 |
+
"epoch": 2.287769784172662,
|
2793 |
+
"grad_norm": 1.3841115301970095,
|
2794 |
+
"learning_rate": 1.591511805050772e-06,
|
2795 |
+
"loss": 0.4804,
|
2796 |
+
"step": 398
|
2797 |
+
},
|
2798 |
+
{
|
2799 |
+
"epoch": 2.293525179856115,
|
2800 |
+
"grad_norm": 1.5658556526287004,
|
2801 |
+
"learning_rate": 1.5669799505219069e-06,
|
2802 |
+
"loss": 0.5657,
|
2803 |
+
"step": 399
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"epoch": 2.2992805755395684,
|
2807 |
+
"grad_norm": 1.4410476254144169,
|
2808 |
+
"learning_rate": 1.542603456520214e-06,
|
2809 |
+
"loss": 0.5256,
|
2810 |
+
"step": 400
|
2811 |
+
},
|
2812 |
+
{
|
2813 |
+
"epoch": 2.3050359712230217,
|
2814 |
+
"grad_norm": 1.396226499263132,
|
2815 |
+
"learning_rate": 1.5183834261982804e-06,
|
2816 |
+
"loss": 0.5076,
|
2817 |
+
"step": 401
|
2818 |
+
},
|
2819 |
+
{
|
2820 |
+
"epoch": 2.310791366906475,
|
2821 |
+
"grad_norm": 1.4741013879844251,
|
2822 |
+
"learning_rate": 1.4943209556279698e-06,
|
2823 |
+
"loss": 0.4741,
|
2824 |
+
"step": 402
|
2825 |
+
},
|
2826 |
+
{
|
2827 |
+
"epoch": 2.316546762589928,
|
2828 |
+
"grad_norm": 1.3887345574987577,
|
2829 |
+
"learning_rate": 1.4704171337508144e-06,
|
2830 |
+
"loss": 0.4733,
|
2831 |
+
"step": 403
|
2832 |
+
},
|
2833 |
+
{
|
2834 |
+
"epoch": 2.322302158273381,
|
2835 |
+
"grad_norm": 1.4960547967615687,
|
2836 |
+
"learning_rate": 1.4466730423287385e-06,
|
2837 |
+
"loss": 0.5249,
|
2838 |
+
"step": 404
|
2839 |
+
},
|
2840 |
+
{
|
2841 |
+
"epoch": 2.3280575539568344,
|
2842 |
+
"grad_norm": 1.39072943170089,
|
2843 |
+
"learning_rate": 1.423089755895095e-06,
|
2844 |
+
"loss": 0.5019,
|
2845 |
+
"step": 405
|
2846 |
+
},
|
2847 |
+
{
|
2848 |
+
"epoch": 2.3338129496402877,
|
2849 |
+
"grad_norm": 1.3482954314170201,
|
2850 |
+
"learning_rate": 1.399668341706053e-06,
|
2851 |
+
"loss": 0.4565,
|
2852 |
+
"step": 406
|
2853 |
+
},
|
2854 |
+
{
|
2855 |
+
"epoch": 2.339568345323741,
|
2856 |
+
"grad_norm": 1.380967235956737,
|
2857 |
+
"learning_rate": 1.3764098596922865e-06,
|
2858 |
+
"loss": 0.4556,
|
2859 |
+
"step": 407
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 2.3453237410071943,
|
2863 |
+
"grad_norm": 1.6744734227791809,
|
2864 |
+
"learning_rate": 1.3533153624110097e-06,
|
2865 |
+
"loss": 0.5122,
|
2866 |
+
"step": 408
|
2867 |
+
},
|
2868 |
+
{
|
2869 |
+
"epoch": 2.3510791366906476,
|
2870 |
+
"grad_norm": 1.308747786329407,
|
2871 |
+
"learning_rate": 1.3303858949983495e-06,
|
2872 |
+
"loss": 0.4796,
|
2873 |
+
"step": 409
|
2874 |
+
},
|
2875 |
+
{
|
2876 |
+
"epoch": 2.356834532374101,
|
2877 |
+
"grad_norm": 1.3262402008626963,
|
2878 |
+
"learning_rate": 1.3076224951220413e-06,
|
2879 |
+
"loss": 0.5239,
|
2880 |
+
"step": 410
|
2881 |
+
},
|
2882 |
+
{
|
2883 |
+
"epoch": 2.362589928057554,
|
2884 |
+
"grad_norm": 1.4771851566884782,
|
2885 |
+
"learning_rate": 1.2850261929344748e-06,
|
2886 |
+
"loss": 0.5123,
|
2887 |
+
"step": 411
|
2888 |
+
},
|
2889 |
+
{
|
2890 |
+
"epoch": 2.368345323741007,
|
2891 |
+
"grad_norm": 1.5006417349129988,
|
2892 |
+
"learning_rate": 1.2625980110260711e-06,
|
2893 |
+
"loss": 0.458,
|
2894 |
+
"step": 412
|
2895 |
+
},
|
2896 |
+
{
|
2897 |
+
"epoch": 2.3741007194244603,
|
2898 |
+
"grad_norm": 1.3926430193253936,
|
2899 |
+
"learning_rate": 1.2403389643790086e-06,
|
2900 |
+
"loss": 0.4927,
|
2901 |
+
"step": 413
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 2.3798561151079136,
|
2905 |
+
"grad_norm": 1.4510762046369095,
|
2906 |
+
"learning_rate": 1.2182500603212882e-06,
|
2907 |
+
"loss": 0.451,
|
2908 |
+
"step": 414
|
2909 |
+
},
|
2910 |
+
{
|
2911 |
+
"epoch": 2.385611510791367,
|
2912 |
+
"grad_norm": 1.4381927072283884,
|
2913 |
+
"learning_rate": 1.1963322984811454e-06,
|
2914 |
+
"loss": 0.4933,
|
2915 |
+
"step": 415
|
2916 |
+
},
|
2917 |
+
{
|
2918 |
+
"epoch": 2.39136690647482,
|
2919 |
+
"grad_norm": 1.497304538711396,
|
2920 |
+
"learning_rate": 1.1745866707418146e-06,
|
2921 |
+
"loss": 0.4908,
|
2922 |
+
"step": 416
|
2923 |
+
},
|
2924 |
+
{
|
2925 |
+
"epoch": 2.3971223021582735,
|
2926 |
+
"grad_norm": 1.4476753575546282,
|
2927 |
+
"learning_rate": 1.1530141611966438e-06,
|
2928 |
+
"loss": 0.4922,
|
2929 |
+
"step": 417
|
2930 |
+
},
|
2931 |
+
{
|
2932 |
+
"epoch": 2.402877697841727,
|
2933 |
+
"grad_norm": 1.4591571350741641,
|
2934 |
+
"learning_rate": 1.1316157461045553e-06,
|
2935 |
+
"loss": 0.701,
|
2936 |
+
"step": 418
|
2937 |
+
},
|
2938 |
+
{
|
2939 |
+
"epoch": 2.4086330935251796,
|
2940 |
+
"grad_norm": 1.3455302080893126,
|
2941 |
+
"learning_rate": 1.1103923938458677e-06,
|
2942 |
+
"loss": 0.522,
|
2943 |
+
"step": 419
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 2.414388489208633,
|
2947 |
+
"grad_norm": 1.4071620638813152,
|
2948 |
+
"learning_rate": 1.0893450648784736e-06,
|
2949 |
+
"loss": 0.4477,
|
2950 |
+
"step": 420
|
2951 |
+
},
|
2952 |
+
{
|
2953 |
+
"epoch": 2.420143884892086,
|
2954 |
+
"grad_norm": 1.4459614537857528,
|
2955 |
+
"learning_rate": 1.0684747116943683e-06,
|
2956 |
+
"loss": 0.5156,
|
2957 |
+
"step": 421
|
2958 |
+
},
|
2959 |
+
{
|
2960 |
+
"epoch": 2.4258992805755395,
|
2961 |
+
"grad_norm": 1.5441740748483679,
|
2962 |
+
"learning_rate": 1.04778227877655e-06,
|
2963 |
+
"loss": 0.4778,
|
2964 |
+
"step": 422
|
2965 |
+
},
|
2966 |
+
{
|
2967 |
+
"epoch": 2.431654676258993,
|
2968 |
+
"grad_norm": 1.4798748382068003,
|
2969 |
+
"learning_rate": 1.0272687025562794e-06,
|
2970 |
+
"loss": 0.4488,
|
2971 |
+
"step": 423
|
2972 |
+
},
|
2973 |
+
{
|
2974 |
+
"epoch": 2.437410071942446,
|
2975 |
+
"grad_norm": 1.3561667072800911,
|
2976 |
+
"learning_rate": 1.006934911370696e-06,
|
2977 |
+
"loss": 0.495,
|
2978 |
+
"step": 424
|
2979 |
+
},
|
2980 |
+
{
|
2981 |
+
"epoch": 2.4431654676258994,
|
2982 |
+
"grad_norm": 1.527860492725263,
|
2983 |
+
"learning_rate": 9.867818254208122e-07,
|
2984 |
+
"loss": 0.4817,
|
2985 |
+
"step": 425
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 2.4489208633093527,
|
2989 |
+
"grad_norm": 1.4394782345640176,
|
2990 |
+
"learning_rate": 9.668103567298615e-07,
|
2991 |
+
"loss": 0.5245,
|
2992 |
+
"step": 426
|
2993 |
+
},
|
2994 |
+
{
|
2995 |
+
"epoch": 2.454676258992806,
|
2996 |
+
"grad_norm": 1.2994965760054968,
|
2997 |
+
"learning_rate": 9.470214091020358e-07,
|
2998 |
+
"loss": 0.4637,
|
2999 |
+
"step": 427
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 2.460431654676259,
|
3003 |
+
"grad_norm": 1.2874005330680733,
|
3004 |
+
"learning_rate": 9.274158780815767e-07,
|
3005 |
+
"loss": 0.5404,
|
3006 |
+
"step": 428
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 2.466187050359712,
|
3010 |
+
"grad_norm": 1.5269622339070275,
|
3011 |
+
"learning_rate": 9.079946509122473e-07,
|
3012 |
+
"loss": 0.5136,
|
3013 |
+
"step": 429
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 2.4719424460431654,
|
3017 |
+
"grad_norm": 1.4550501719811364,
|
3018 |
+
"learning_rate": 8.887586064971859e-07,
|
3019 |
+
"loss": 0.5591,
|
3020 |
+
"step": 430
|
3021 |
+
},
|
3022 |
+
{
|
3023 |
+
"epoch": 2.4776978417266187,
|
3024 |
+
"grad_norm": 1.39766846078072,
|
3025 |
+
"learning_rate": 8.697086153591289e-07,
|
3026 |
+
"loss": 0.5136,
|
3027 |
+
"step": 431
|
3028 |
+
},
|
3029 |
+
{
|
3030 |
+
"epoch": 2.483453237410072,
|
3031 |
+
"grad_norm": 1.370797215046593,
|
3032 |
+
"learning_rate": 8.508455396010096e-07,
|
3033 |
+
"loss": 0.4559,
|
3034 |
+
"step": 432
|
3035 |
+
},
|
3036 |
+
{
|
3037 |
+
"epoch": 2.4892086330935252,
|
3038 |
+
"grad_norm": 1.4244740354584018,
|
3039 |
+
"learning_rate": 8.321702328669534e-07,
|
3040 |
+
"loss": 0.442,
|
3041 |
+
"step": 433
|
3042 |
+
},
|
3043 |
+
{
|
3044 |
+
"epoch": 2.4949640287769785,
|
3045 |
+
"grad_norm": 1.3428260355210209,
|
3046 |
+
"learning_rate": 8.136835403036413e-07,
|
3047 |
+
"loss": 0.5349,
|
3048 |
+
"step": 434
|
3049 |
+
},
|
3050 |
+
{
|
3051 |
+
"epoch": 2.5007194244604314,
|
3052 |
+
"grad_norm": 1.4772346312485674,
|
3053 |
+
"learning_rate": 7.95386298522065e-07,
|
3054 |
+
"loss": 0.5127,
|
3055 |
+
"step": 435
|
3056 |
+
},
|
3057 |
+
{
|
3058 |
+
"epoch": 2.506474820143885,
|
3059 |
+
"grad_norm": 1.4410455867849274,
|
3060 |
+
"learning_rate": 7.772793355596597e-07,
|
3061 |
+
"loss": 0.4862,
|
3062 |
+
"step": 436
|
3063 |
+
},
|
3064 |
+
{
|
3065 |
+
"epoch": 2.512230215827338,
|
3066 |
+
"grad_norm": 1.5076709055926627,
|
3067 |
+
"learning_rate": 7.593634708428438e-07,
|
3068 |
+
"loss": 0.4961,
|
3069 |
+
"step": 437
|
3070 |
+
},
|
3071 |
+
{
|
3072 |
+
"epoch": 2.5179856115107913,
|
3073 |
+
"grad_norm": 1.4236619141715625,
|
3074 |
+
"learning_rate": 7.416395151499223e-07,
|
3075 |
+
"loss": 0.4983,
|
3076 |
+
"step": 438
|
3077 |
+
},
|
3078 |
+
{
|
3079 |
+
"epoch": 2.5237410071942445,
|
3080 |
+
"grad_norm": 1.374108165442237,
|
3081 |
+
"learning_rate": 7.241082705744057e-07,
|
3082 |
+
"loss": 0.463,
|
3083 |
+
"step": 439
|
3084 |
+
},
|
3085 |
+
{
|
3086 |
+
"epoch": 2.529496402877698,
|
3087 |
+
"grad_norm": 1.351950367512,
|
3088 |
+
"learning_rate": 7.067705304887074e-07,
|
3089 |
+
"loss": 0.5056,
|
3090 |
+
"step": 440
|
3091 |
+
},
|
3092 |
+
{
|
3093 |
+
"epoch": 2.535251798561151,
|
3094 |
+
"grad_norm": 1.4377685351017533,
|
3095 |
+
"learning_rate": 6.896270795082394e-07,
|
3096 |
+
"loss": 0.5157,
|
3097 |
+
"step": 441
|
3098 |
+
},
|
3099 |
+
{
|
3100 |
+
"epoch": 2.5410071942446044,
|
3101 |
+
"grad_norm": 1.3991942447953665,
|
3102 |
+
"learning_rate": 6.726786934559048e-07,
|
3103 |
+
"loss": 0.4595,
|
3104 |
+
"step": 442
|
3105 |
+
},
|
3106 |
+
{
|
3107 |
+
"epoch": 2.5467625899280577,
|
3108 |
+
"grad_norm": 1.424674987885115,
|
3109 |
+
"learning_rate": 6.559261393269872e-07,
|
3110 |
+
"loss": 0.5126,
|
3111 |
+
"step": 443
|
3112 |
+
},
|
3113 |
+
{
|
3114 |
+
"epoch": 2.5525179856115106,
|
3115 |
+
"grad_norm": 1.4167917674267438,
|
3116 |
+
"learning_rate": 6.39370175254444e-07,
|
3117 |
+
"loss": 0.4932,
|
3118 |
+
"step": 444
|
3119 |
+
},
|
3120 |
+
{
|
3121 |
+
"epoch": 2.5582733812949643,
|
3122 |
+
"grad_norm": 1.5312464974087947,
|
3123 |
+
"learning_rate": 6.230115504745954e-07,
|
3124 |
+
"loss": 0.4962,
|
3125 |
+
"step": 445
|
3126 |
+
},
|
3127 |
+
{
|
3128 |
+
"epoch": 2.564028776978417,
|
3129 |
+
"grad_norm": 1.4990800401772026,
|
3130 |
+
"learning_rate": 6.06851005293217e-07,
|
3131 |
+
"loss": 0.5231,
|
3132 |
+
"step": 446
|
3133 |
+
},
|
3134 |
+
{
|
3135 |
+
"epoch": 2.5697841726618704,
|
3136 |
+
"grad_norm": 1.4268095211478857,
|
3137 |
+
"learning_rate": 5.908892710520375e-07,
|
3138 |
+
"loss": 0.5426,
|
3139 |
+
"step": 447
|
3140 |
+
},
|
3141 |
+
{
|
3142 |
+
"epoch": 2.5755395683453237,
|
3143 |
+
"grad_norm": 1.4261657348161196,
|
3144 |
+
"learning_rate": 5.75127070095643e-07,
|
3145 |
+
"loss": 0.4934,
|
3146 |
+
"step": 448
|
3147 |
+
},
|
3148 |
+
{
|
3149 |
+
"epoch": 2.581294964028777,
|
3150 |
+
"grad_norm": 1.5401518150175277,
|
3151 |
+
"learning_rate": 5.595651157387855e-07,
|
3152 |
+
"loss": 0.5111,
|
3153 |
+
"step": 449
|
3154 |
+
},
|
3155 |
+
{
|
3156 |
+
"epoch": 2.5870503597122303,
|
3157 |
+
"grad_norm": 1.4908443652360226,
|
3158 |
+
"learning_rate": 5.442041122341057e-07,
|
3159 |
+
"loss": 0.5216,
|
3160 |
+
"step": 450
|
3161 |
+
},
|
3162 |
+
{
|
3163 |
+
"epoch": 2.5928057553956836,
|
3164 |
+
"grad_norm": 1.423810583953011,
|
3165 |
+
"learning_rate": 5.290447547402594e-07,
|
3166 |
+
"loss": 0.5339,
|
3167 |
+
"step": 451
|
3168 |
+
},
|
3169 |
+
{
|
3170 |
+
"epoch": 2.598561151079137,
|
3171 |
+
"grad_norm": 1.4243627689716587,
|
3172 |
+
"learning_rate": 5.14087729290459e-07,
|
3173 |
+
"loss": 0.5093,
|
3174 |
+
"step": 452
|
3175 |
+
},
|
3176 |
+
{
|
3177 |
+
"epoch": 2.6043165467625897,
|
3178 |
+
"grad_norm": 1.4534187291002245,
|
3179 |
+
"learning_rate": 4.993337127614273e-07,
|
3180 |
+
"loss": 0.4698,
|
3181 |
+
"step": 453
|
3182 |
+
},
|
3183 |
+
{
|
3184 |
+
"epoch": 2.610071942446043,
|
3185 |
+
"grad_norm": 1.3476942698266083,
|
3186 |
+
"learning_rate": 4.847833728427636e-07,
|
3187 |
+
"loss": 0.4943,
|
3188 |
+
"step": 454
|
3189 |
+
},
|
3190 |
+
{
|
3191 |
+
"epoch": 2.6158273381294963,
|
3192 |
+
"grad_norm": 1.5930441971010927,
|
3193 |
+
"learning_rate": 4.7043736800673254e-07,
|
3194 |
+
"loss": 0.5932,
|
3195 |
+
"step": 455
|
3196 |
+
},
|
3197 |
+
{
|
3198 |
+
"epoch": 2.6215827338129496,
|
3199 |
+
"grad_norm": 1.48344645609647,
|
3200 |
+
"learning_rate": 4.5629634747845764e-07,
|
3201 |
+
"loss": 0.5455,
|
3202 |
+
"step": 456
|
3203 |
+
},
|
3204 |
+
{
|
3205 |
+
"epoch": 2.627338129496403,
|
3206 |
+
"grad_norm": 1.3457429750649241,
|
3207 |
+
"learning_rate": 4.423609512065485e-07,
|
3208 |
+
"loss": 0.4991,
|
3209 |
+
"step": 457
|
3210 |
+
},
|
3211 |
+
{
|
3212 |
+
"epoch": 2.633093525179856,
|
3213 |
+
"grad_norm": 1.5107543307770037,
|
3214 |
+
"learning_rate": 4.2863180983413744e-07,
|
3215 |
+
"loss": 0.4706,
|
3216 |
+
"step": 458
|
3217 |
+
},
|
3218 |
+
{
|
3219 |
+
"epoch": 2.6388489208633095,
|
3220 |
+
"grad_norm": 1.5552826637073596,
|
3221 |
+
"learning_rate": 4.1510954467033457e-07,
|
3222 |
+
"loss": 0.5151,
|
3223 |
+
"step": 459
|
3224 |
+
},
|
3225 |
+
{
|
3226 |
+
"epoch": 2.6446043165467623,
|
3227 |
+
"grad_norm": 1.511602864357795,
|
3228 |
+
"learning_rate": 4.0179476766211865e-07,
|
3229 |
+
"loss": 0.4439,
|
3230 |
+
"step": 460
|
3231 |
+
},
|
3232 |
+
{
|
3233 |
+
"epoch": 2.650359712230216,
|
3234 |
+
"grad_norm": 1.4133323220879965,
|
3235 |
+
"learning_rate": 3.8868808136663995e-07,
|
3236 |
+
"loss": 0.493,
|
3237 |
+
"step": 461
|
3238 |
+
},
|
3239 |
+
{
|
3240 |
+
"epoch": 2.656115107913669,
|
3241 |
+
"grad_norm": 1.3640394323334066,
|
3242 |
+
"learning_rate": 3.757900789239516e-07,
|
3243 |
+
"loss": 0.539,
|
3244 |
+
"step": 462
|
3245 |
+
},
|
3246 |
+
{
|
3247 |
+
"epoch": 2.661870503597122,
|
3248 |
+
"grad_norm": 1.4531907210640926,
|
3249 |
+
"learning_rate": 3.631013440301645e-07,
|
3250 |
+
"loss": 0.5056,
|
3251 |
+
"step": 463
|
3252 |
+
},
|
3253 |
+
{
|
3254 |
+
"epoch": 2.6676258992805755,
|
3255 |
+
"grad_norm": 1.2341263739518196,
|
3256 |
+
"learning_rate": 3.50622450911039e-07,
|
3257 |
+
"loss": 0.4809,
|
3258 |
+
"step": 464
|
3259 |
+
},
|
3260 |
+
{
|
3261 |
+
"epoch": 2.6733812949640288,
|
3262 |
+
"grad_norm": 1.3535259462137235,
|
3263 |
+
"learning_rate": 3.383539642959915e-07,
|
3264 |
+
"loss": 0.5173,
|
3265 |
+
"step": 465
|
3266 |
+
},
|
3267 |
+
{
|
3268 |
+
"epoch": 2.679136690647482,
|
3269 |
+
"grad_norm": 1.2717925450759886,
|
3270 |
+
"learning_rate": 3.262964393925433e-07,
|
3271 |
+
"loss": 0.498,
|
3272 |
+
"step": 466
|
3273 |
+
},
|
3274 |
+
{
|
3275 |
+
"epoch": 2.6848920863309353,
|
3276 |
+
"grad_norm": 1.534833028283325,
|
3277 |
+
"learning_rate": 3.144504218611899e-07,
|
3278 |
+
"loss": 0.51,
|
3279 |
+
"step": 467
|
3280 |
+
},
|
3281 |
+
{
|
3282 |
+
"epoch": 2.6906474820143886,
|
3283 |
+
"grad_norm": 1.434016469477903,
|
3284 |
+
"learning_rate": 3.028164477907125e-07,
|
3285 |
+
"loss": 0.4468,
|
3286 |
+
"step": 468
|
3287 |
+
},
|
3288 |
+
{
|
3289 |
+
"epoch": 2.6964028776978415,
|
3290 |
+
"grad_norm": 1.4466502715231508,
|
3291 |
+
"learning_rate": 2.913950436739116e-07,
|
3292 |
+
"loss": 0.451,
|
3293 |
+
"step": 469
|
3294 |
+
},
|
3295 |
+
{
|
3296 |
+
"epoch": 2.702158273381295,
|
3297 |
+
"grad_norm": 1.3985545137247108,
|
3298 |
+
"learning_rate": 2.8018672638378486e-07,
|
3299 |
+
"loss": 0.7395,
|
3300 |
+
"step": 470
|
3301 |
+
},
|
3302 |
+
{
|
3303 |
+
"epoch": 2.707913669064748,
|
3304 |
+
"grad_norm": 1.4226087235155982,
|
3305 |
+
"learning_rate": 2.6919200315013606e-07,
|
3306 |
+
"loss": 0.5293,
|
3307 |
+
"step": 471
|
3308 |
+
},
|
3309 |
+
{
|
3310 |
+
"epoch": 2.7136690647482014,
|
3311 |
+
"grad_norm": 1.5543293605728232,
|
3312 |
+
"learning_rate": 2.5841137153661765e-07,
|
3313 |
+
"loss": 0.5105,
|
3314 |
+
"step": 472
|
3315 |
+
},
|
3316 |
+
{
|
3317 |
+
"epoch": 2.7194244604316546,
|
3318 |
+
"grad_norm": 1.3787651810116257,
|
3319 |
+
"learning_rate": 2.4784531941821675e-07,
|
3320 |
+
"loss": 0.4938,
|
3321 |
+
"step": 473
|
3322 |
+
},
|
3323 |
+
{
|
3324 |
+
"epoch": 2.725179856115108,
|
3325 |
+
"grad_norm": 1.2755000464637145,
|
3326 |
+
"learning_rate": 2.3749432495917546e-07,
|
3327 |
+
"loss": 0.4779,
|
3328 |
+
"step": 474
|
3329 |
+
},
|
3330 |
+
{
|
3331 |
+
"epoch": 2.7309352517985612,
|
3332 |
+
"grad_norm": 1.4247225161211758,
|
3333 |
+
"learning_rate": 2.2735885659134927e-07,
|
3334 |
+
"loss": 0.5872,
|
3335 |
+
"step": 475
|
3336 |
+
},
|
3337 |
+
{
|
3338 |
+
"epoch": 2.7366906474820145,
|
3339 |
+
"grad_norm": 1.3774551457458113,
|
3340 |
+
"learning_rate": 2.1743937299301242e-07,
|
3341 |
+
"loss": 0.5173,
|
3342 |
+
"step": 476
|
3343 |
+
},
|
3344 |
+
{
|
3345 |
+
"epoch": 2.742446043165468,
|
3346 |
+
"grad_norm": 1.32333951318816,
|
3347 |
+
"learning_rate": 2.0773632306809622e-07,
|
3348 |
+
"loss": 0.6346,
|
3349 |
+
"step": 477
|
3350 |
+
},
|
3351 |
+
{
|
3352 |
+
"epoch": 2.7482014388489207,
|
3353 |
+
"grad_norm": 1.428335557104341,
|
3354 |
+
"learning_rate": 1.9825014592587844e-07,
|
3355 |
+
"loss": 0.5229,
|
3356 |
+
"step": 478
|
3357 |
+
},
|
3358 |
+
{
|
3359 |
+
"epoch": 2.753956834532374,
|
3360 |
+
"grad_norm": 1.5253782080180924,
|
3361 |
+
"learning_rate": 1.889812708611083e-07,
|
3362 |
+
"loss": 0.5246,
|
3363 |
+
"step": 479
|
3364 |
+
},
|
3365 |
+
{
|
3366 |
+
"epoch": 2.7597122302158272,
|
3367 |
+
"grad_norm": 1.4084048391223996,
|
3368 |
+
"learning_rate": 1.7993011733458077e-07,
|
3369 |
+
"loss": 0.504,
|
3370 |
+
"step": 480
|
3371 |
+
},
|
3372 |
+
{
|
3373 |
+
"epoch": 2.7654676258992805,
|
3374 |
+
"grad_norm": 1.5303085706562192,
|
3375 |
+
"learning_rate": 1.7109709495415073e-07,
|
3376 |
+
"loss": 0.5248,
|
3377 |
+
"step": 481
|
3378 |
+
},
|
3379 |
+
{
|
3380 |
+
"epoch": 2.771223021582734,
|
3381 |
+
"grad_norm": 1.4650647549311855,
|
3382 |
+
"learning_rate": 1.624826034562016e-07,
|
3383 |
+
"loss": 0.5073,
|
3384 |
+
"step": 482
|
3385 |
+
},
|
3386 |
+
{
|
3387 |
+
"epoch": 2.776978417266187,
|
3388 |
+
"grad_norm": 1.5227538988370155,
|
3389 |
+
"learning_rate": 1.5408703268754988e-07,
|
3390 |
+
"loss": 0.5848,
|
3391 |
+
"step": 483
|
3392 |
+
},
|
3393 |
+
{
|
3394 |
+
"epoch": 2.7827338129496404,
|
3395 |
+
"grad_norm": 1.5441631260871607,
|
3396 |
+
"learning_rate": 1.459107625878059e-07,
|
3397 |
+
"loss": 0.5471,
|
3398 |
+
"step": 484
|
3399 |
+
},
|
3400 |
+
{
|
3401 |
+
"epoch": 2.7884892086330937,
|
3402 |
+
"grad_norm": 1.3168339616382214,
|
3403 |
+
"learning_rate": 1.3795416317218036e-07,
|
3404 |
+
"loss": 0.4482,
|
3405 |
+
"step": 485
|
3406 |
+
},
|
3407 |
+
{
|
3408 |
+
"epoch": 2.794244604316547,
|
3409 |
+
"grad_norm": 1.367273457686962,
|
3410 |
+
"learning_rate": 1.3021759451473548e-07,
|
3411 |
+
"loss": 0.575,
|
3412 |
+
"step": 486
|
3413 |
+
},
|
3414 |
+
{
|
3415 |
+
"epoch": 2.8,
|
3416 |
+
"grad_norm": 1.3719280025477179,
|
3417 |
+
"learning_rate": 1.2270140673209473e-07,
|
3418 |
+
"loss": 0.5192,
|
3419 |
+
"step": 487
|
3420 |
+
},
|
3421 |
+
{
|
3422 |
+
"epoch": 2.805755395683453,
|
3423 |
+
"grad_norm": 1.4869115987076804,
|
3424 |
+
"learning_rate": 1.1540593996759441e-07,
|
3425 |
+
"loss": 0.4412,
|
3426 |
+
"step": 488
|
3427 |
+
},
|
3428 |
+
{
|
3429 |
+
"epoch": 2.8115107913669064,
|
3430 |
+
"grad_norm": 1.367254851197092,
|
3431 |
+
"learning_rate": 1.0833152437589423e-07,
|
3432 |
+
"loss": 0.5257,
|
3433 |
+
"step": 489
|
3434 |
+
},
|
3435 |
+
{
|
3436 |
+
"epoch": 2.8172661870503597,
|
3437 |
+
"grad_norm": 1.3370242128553043,
|
3438 |
+
"learning_rate": 1.0147848010803319e-07,
|
3439 |
+
"loss": 0.5181,
|
3440 |
+
"step": 490
|
3441 |
+
},
|
3442 |
+
{
|
3443 |
+
"epoch": 2.823021582733813,
|
3444 |
+
"grad_norm": 1.5318265381712992,
|
3445 |
+
"learning_rate": 9.484711729694229e-08,
|
3446 |
+
"loss": 0.4949,
|
3447 |
+
"step": 491
|
3448 |
+
},
|
3449 |
+
{
|
3450 |
+
"epoch": 2.8287769784172663,
|
3451 |
+
"grad_norm": 1.4329136201613037,
|
3452 |
+
"learning_rate": 8.8437736043408e-08,
|
3453 |
+
"loss": 0.5203,
|
3454 |
+
"step": 492
|
3455 |
+
},
|
3456 |
+
{
|
3457 |
+
"epoch": 2.8345323741007196,
|
3458 |
+
"grad_norm": 1.46138637291591,
|
3459 |
+
"learning_rate": 8.225062640249636e-08,
|
3460 |
+
"loss": 0.4804,
|
3461 |
+
"step": 493
|
3462 |
+
},
|
3463 |
+
{
|
3464 |
+
"epoch": 2.8402877697841724,
|
3465 |
+
"grad_norm": 1.4142415323843383,
|
3466 |
+
"learning_rate": 7.628606837041974e-08,
|
3467 |
+
"loss": 0.4957,
|
3468 |
+
"step": 494
|
3469 |
+
},
|
3470 |
+
{
|
3471 |
+
"epoch": 2.846043165467626,
|
3472 |
+
"grad_norm": 1.5297571505651404,
|
3473 |
+
"learning_rate": 7.05443318718707e-08,
|
3474 |
+
"loss": 0.5022,
|
3475 |
+
"step": 495
|
3476 |
+
},
|
3477 |
+
{
|
3478 |
+
"epoch": 2.851798561151079,
|
3479 |
+
"grad_norm": 1.3488484858194676,
|
3480 |
+
"learning_rate": 6.502567674780524e-08,
|
3481 |
+
"loss": 0.5056,
|
3482 |
+
"step": 496
|
3483 |
+
},
|
3484 |
+
{
|
3485 |
+
"epoch": 2.8575539568345323,
|
3486 |
+
"grad_norm": 1.443312999262605,
|
3487 |
+
"learning_rate": 5.973035274368266e-08,
|
3488 |
+
"loss": 0.5337,
|
3489 |
+
"step": 497
|
3490 |
+
},
|
3491 |
+
{
|
3492 |
+
"epoch": 2.8633093525179856,
|
3493 |
+
"grad_norm": 1.5511488557005244,
|
3494 |
+
"learning_rate": 5.465859949816299e-08,
|
3495 |
+
"loss": 0.5996,
|
3496 |
+
"step": 498
|
3497 |
+
},
|
3498 |
+
{
|
3499 |
+
"epoch": 2.869064748201439,
|
3500 |
+
"grad_norm": 1.3328832159109907,
|
3501 |
+
"learning_rate": 4.981064653226564e-08,
|
3502 |
+
"loss": 0.4837,
|
3503 |
+
"step": 499
|
3504 |
+
},
|
3505 |
+
{
|
3506 |
+
"epoch": 2.874820143884892,
|
3507 |
+
"grad_norm": 1.478429965603256,
|
3508 |
+
"learning_rate": 4.5186713238979385e-08,
|
3509 |
+
"loss": 0.4907,
|
3510 |
+
"step": 500
|
3511 |
+
},
|
3512 |
+
{
|
3513 |
+
"epoch": 2.8805755395683454,
|
3514 |
+
"grad_norm": 1.322358938042936,
|
3515 |
+
"learning_rate": 4.078700887333365e-08,
|
3516 |
+
"loss": 0.5007,
|
3517 |
+
"step": 501
|
3518 |
+
},
|
3519 |
+
{
|
3520 |
+
"epoch": 2.8863309352517987,
|
3521 |
+
"grad_norm": 1.4155961679670235,
|
3522 |
+
"learning_rate": 3.6611732542931044e-08,
|
3523 |
+
"loss": 0.5438,
|
3524 |
+
"step": 502
|
3525 |
+
},
|
3526 |
+
{
|
3527 |
+
"epoch": 2.8920863309352516,
|
3528 |
+
"grad_norm": 1.4978442625141069,
|
3529 |
+
"learning_rate": 3.266107319893463e-08,
|
3530 |
+
"loss": 0.5051,
|
3531 |
+
"step": 503
|
3532 |
+
},
|
3533 |
+
{
|
3534 |
+
"epoch": 2.897841726618705,
|
3535 |
+
"grad_norm": 1.3022201279271626,
|
3536 |
+
"learning_rate": 2.89352096275175e-08,
|
3537 |
+
"loss": 0.5139,
|
3538 |
+
"step": 504
|
3539 |
+
},
|
3540 |
+
{
|
3541 |
+
"epoch": 2.903597122302158,
|
3542 |
+
"grad_norm": 1.4268625992886872,
|
3543 |
+
"learning_rate": 2.5434310441773135e-08,
|
3544 |
+
"loss": 0.5046,
|
3545 |
+
"step": 505
|
3546 |
+
},
|
3547 |
+
{
|
3548 |
+
"epoch": 2.9093525179856115,
|
3549 |
+
"grad_norm": 1.4424116607391237,
|
3550 |
+
"learning_rate": 2.2158534074083193e-08,
|
3551 |
+
"loss": 0.4081,
|
3552 |
+
"step": 506
|
3553 |
+
},
|
3554 |
+
{
|
3555 |
+
"epoch": 2.9151079136690647,
|
3556 |
+
"grad_norm": 1.3971332050198229,
|
3557 |
+
"learning_rate": 1.910802876894824e-08,
|
3558 |
+
"loss": 0.4497,
|
3559 |
+
"step": 507
|
3560 |
+
},
|
3561 |
+
{
|
3562 |
+
"epoch": 2.920863309352518,
|
3563 |
+
"grad_norm": 1.4467175991043593,
|
3564 |
+
"learning_rate": 1.6282932576279775e-08,
|
3565 |
+
"loss": 0.5532,
|
3566 |
+
"step": 508
|
3567 |
+
},
|
3568 |
+
{
|
3569 |
+
"epoch": 2.9266187050359713,
|
3570 |
+
"grad_norm": 1.4295698837971194,
|
3571 |
+
"learning_rate": 1.3683373345150796e-08,
|
3572 |
+
"loss": 0.4512,
|
3573 |
+
"step": 509
|
3574 |
+
},
|
3575 |
+
{
|
3576 |
+
"epoch": 2.9323741007194246,
|
3577 |
+
"grad_norm": 1.3783042972879613,
|
3578 |
+
"learning_rate": 1.1309468718013194e-08,
|
3579 |
+
"loss": 0.5683,
|
3580 |
+
"step": 510
|
3581 |
+
},
|
3582 |
+
{
|
3583 |
+
"epoch": 2.938129496402878,
|
3584 |
+
"grad_norm": 1.462143528479705,
|
3585 |
+
"learning_rate": 9.16132612537035e-09,
|
3586 |
+
"loss": 0.5451,
|
3587 |
+
"step": 511
|
3588 |
+
},
|
3589 |
+
{
|
3590 |
+
"epoch": 2.9438848920863308,
|
3591 |
+
"grad_norm": 1.4618255084244922,
|
3592 |
+
"learning_rate": 7.2390427809176934e-09,
|
3593 |
+
"loss": 0.4716,
|
3594 |
+
"step": 512
|
3595 |
+
},
|
3596 |
+
{
|
3597 |
+
"epoch": 2.949640287769784,
|
3598 |
+
"grad_norm": 1.3455189640301384,
|
3599 |
+
"learning_rate": 5.542705677143434e-09,
|
3600 |
+
"loss": 0.5137,
|
3601 |
+
"step": 513
|
3602 |
+
},
|
3603 |
+
{
|
3604 |
+
"epoch": 2.9553956834532373,
|
3605 |
+
"grad_norm": 1.4802472454891864,
|
3606 |
+
"learning_rate": 4.072391581388946e-09,
|
3607 |
+
"loss": 0.4704,
|
3608 |
+
"step": 514
|
3609 |
+
},
|
3610 |
+
{
|
3611 |
+
"epoch": 2.9611510791366906,
|
3612 |
+
"grad_norm": 1.489811405234316,
|
3613 |
+
"learning_rate": 2.8281670323798693e-09,
|
3614 |
+
"loss": 0.5358,
|
3615 |
+
"step": 515
|
3616 |
+
},
|
3617 |
+
{
|
3618 |
+
"epoch": 2.966906474820144,
|
3619 |
+
"grad_norm": 1.4937820827189587,
|
3620 |
+
"learning_rate": 1.8100883372085266e-09,
|
3621 |
+
"loss": 0.5253,
|
3622 |
+
"step": 516
|
3623 |
+
},
|
3624 |
+
{
|
3625 |
+
"epoch": 2.972661870503597,
|
3626 |
+
"grad_norm": 1.4382704408115423,
|
3627 |
+
"learning_rate": 1.0182015687909552e-09,
|
3628 |
+
"loss": 0.4914,
|
3629 |
+
"step": 517
|
3630 |
+
},
|
3631 |
+
{
|
3632 |
+
"epoch": 2.9784172661870505,
|
3633 |
+
"grad_norm": 1.404392320887915,
|
3634 |
+
"learning_rate": 4.5254256377913474e-10,
|
3635 |
+
"loss": 0.5768,
|
3636 |
+
"step": 518
|
3637 |
+
},
|
3638 |
+
{
|
3639 |
+
"epoch": 2.9841726618705033,
|
3640 |
+
"grad_norm": 1.5388011799217256,
|
3641 |
+
"learning_rate": 1.1313692094117034e-10,
|
3642 |
+
"loss": 0.4779,
|
3643 |
+
"step": 519
|
3644 |
+
},
|
3645 |
+
{
|
3646 |
+
"epoch": 2.9841726618705033,
|
3647 |
+
"step": 519,
|
3648 |
+
"total_flos": 4.226722994984059e+17,
|
3649 |
+
"train_loss": 0.7501044547282203,
|
3650 |
+
"train_runtime": 58208.7442,
|
3651 |
+
"train_samples_per_second": 0.43,
|
3652 |
+
"train_steps_per_second": 0.009
|
3653 |
+
}
|
3654 |
+
],
|
3655 |
+
"logging_steps": 1,
|
3656 |
+
"max_steps": 519,
|
3657 |
+
"num_input_tokens_seen": 0,
|
3658 |
+
"num_train_epochs": 3,
|
3659 |
+
"save_steps": 200,
|
3660 |
+
"stateful_callbacks": {
|
3661 |
+
"TrainerControl": {
|
3662 |
+
"args": {
|
3663 |
+
"should_epoch_stop": false,
|
3664 |
+
"should_evaluate": false,
|
3665 |
+
"should_log": false,
|
3666 |
+
"should_save": true,
|
3667 |
+
"should_training_stop": true
|
3668 |
+
},
|
3669 |
+
"attributes": {}
|
3670 |
+
}
|
3671 |
+
},
|
3672 |
+
"total_flos": 4.226722994984059e+17,
|
3673 |
+
"train_batch_size": 1,
|
3674 |
+
"trial_name": null,
|
3675 |
+
"trial_params": null
|
3676 |
+
}
|
training_loss.png
ADDED
![]() |