End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +3024 -0
- training_loss.png +0 -0
README.md
CHANGED
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
4 |
base_model: Qwen/Qwen2.5-Coder-32B-Instruct
|
5 |
tags:
|
6 |
- llama-factory
|
|
|
7 |
- generated_from_trainer
|
8 |
model-index:
|
9 |
- name: SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_32b_rvvpgds
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_32b_rvvpgds
|
17 |
|
18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct) on
|
19 |
|
20 |
## Model description
|
21 |
|
|
|
4 |
base_model: Qwen/Qwen2.5-Coder-32B-Instruct
|
5 |
tags:
|
6 |
- llama-factory
|
7 |
+
- full
|
8 |
- generated_from_trainer
|
9 |
model-index:
|
10 |
- name: SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_32b_rvvpgds
|
|
|
16 |
|
17 |
# SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source_qwen_code_32b_rvvpgds
|
18 |
|
19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct) on the SWE-BENCH-reproduce-verify-validation-plus-gym-diverse-source dataset.
|
20 |
|
21 |
## Model description
|
22 |
|
all_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9836065573770494,
|
3 |
+
"total_flos": 1.0638855097443942e+17,
|
4 |
+
"train_loss": 0.35510572714145194,
|
5 |
+
"train_runtime": 58523.3937,
|
6 |
+
"train_samples_per_second": 0.175,
|
7 |
+
"train_steps_per_second": 0.007
|
8 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9836065573770494,
|
3 |
+
"total_flos": 1.0638855097443942e+17,
|
4 |
+
"train_loss": 0.35510572714145194,
|
5 |
+
"train_runtime": 58523.3937,
|
6 |
+
"train_samples_per_second": 0.175,
|
7 |
+
"train_steps_per_second": 0.007
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3024 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.9836065573770494,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 426,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.00702576112412178,
|
13 |
+
"grad_norm": 0.18053510785102844,
|
14 |
+
"learning_rate": 2.3255813953488374e-07,
|
15 |
+
"loss": 0.7704,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.01405152224824356,
|
20 |
+
"grad_norm": 0.18344072997570038,
|
21 |
+
"learning_rate": 4.651162790697675e-07,
|
22 |
+
"loss": 0.7172,
|
23 |
+
"step": 2
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.02107728337236534,
|
27 |
+
"grad_norm": 0.17655028402805328,
|
28 |
+
"learning_rate": 6.976744186046513e-07,
|
29 |
+
"loss": 0.7562,
|
30 |
+
"step": 3
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.02810304449648712,
|
34 |
+
"grad_norm": 0.1985616534948349,
|
35 |
+
"learning_rate": 9.30232558139535e-07,
|
36 |
+
"loss": 0.8166,
|
37 |
+
"step": 4
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.0351288056206089,
|
41 |
+
"grad_norm": 0.14133895933628082,
|
42 |
+
"learning_rate": 1.1627906976744188e-06,
|
43 |
+
"loss": 0.668,
|
44 |
+
"step": 5
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.04215456674473068,
|
48 |
+
"grad_norm": 0.17750921845436096,
|
49 |
+
"learning_rate": 1.3953488372093025e-06,
|
50 |
+
"loss": 0.7862,
|
51 |
+
"step": 6
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.04918032786885246,
|
55 |
+
"grad_norm": 0.1687806397676468,
|
56 |
+
"learning_rate": 1.6279069767441862e-06,
|
57 |
+
"loss": 0.7697,
|
58 |
+
"step": 7
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.05620608899297424,
|
62 |
+
"grad_norm": 0.14865685999393463,
|
63 |
+
"learning_rate": 1.86046511627907e-06,
|
64 |
+
"loss": 0.6865,
|
65 |
+
"step": 8
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.06323185011709602,
|
69 |
+
"grad_norm": 0.15419331192970276,
|
70 |
+
"learning_rate": 2.0930232558139536e-06,
|
71 |
+
"loss": 0.7145,
|
72 |
+
"step": 9
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.0702576112412178,
|
76 |
+
"grad_norm": 0.1144702211022377,
|
77 |
+
"learning_rate": 2.3255813953488376e-06,
|
78 |
+
"loss": 0.6761,
|
79 |
+
"step": 10
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.07728337236533958,
|
83 |
+
"grad_norm": 0.12051534652709961,
|
84 |
+
"learning_rate": 2.558139534883721e-06,
|
85 |
+
"loss": 0.7301,
|
86 |
+
"step": 11
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.08430913348946135,
|
90 |
+
"grad_norm": 0.11249984055757523,
|
91 |
+
"learning_rate": 2.790697674418605e-06,
|
92 |
+
"loss": 0.7126,
|
93 |
+
"step": 12
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.09133489461358314,
|
97 |
+
"grad_norm": 0.11579480022192001,
|
98 |
+
"learning_rate": 3.0232558139534885e-06,
|
99 |
+
"loss": 0.6869,
|
100 |
+
"step": 13
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.09836065573770492,
|
104 |
+
"grad_norm": 0.08192656189203262,
|
105 |
+
"learning_rate": 3.2558139534883724e-06,
|
106 |
+
"loss": 0.6698,
|
107 |
+
"step": 14
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.1053864168618267,
|
111 |
+
"grad_norm": 0.08397164940834045,
|
112 |
+
"learning_rate": 3.4883720930232564e-06,
|
113 |
+
"loss": 0.6257,
|
114 |
+
"step": 15
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.11241217798594848,
|
118 |
+
"grad_norm": 0.07774341851472855,
|
119 |
+
"learning_rate": 3.72093023255814e-06,
|
120 |
+
"loss": 0.5955,
|
121 |
+
"step": 16
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.11943793911007025,
|
125 |
+
"grad_norm": 0.08383660763502121,
|
126 |
+
"learning_rate": 3.953488372093024e-06,
|
127 |
+
"loss": 0.5698,
|
128 |
+
"step": 17
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.12646370023419204,
|
132 |
+
"grad_norm": 0.08330123126506805,
|
133 |
+
"learning_rate": 4.186046511627907e-06,
|
134 |
+
"loss": 0.6513,
|
135 |
+
"step": 18
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.13348946135831383,
|
139 |
+
"grad_norm": 0.07081824541091919,
|
140 |
+
"learning_rate": 4.418604651162791e-06,
|
141 |
+
"loss": 0.5878,
|
142 |
+
"step": 19
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.1405152224824356,
|
146 |
+
"grad_norm": 0.09581199288368225,
|
147 |
+
"learning_rate": 4.651162790697675e-06,
|
148 |
+
"loss": 0.6923,
|
149 |
+
"step": 20
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.14754098360655737,
|
153 |
+
"grad_norm": 0.06888371706008911,
|
154 |
+
"learning_rate": 4.883720930232559e-06,
|
155 |
+
"loss": 0.5683,
|
156 |
+
"step": 21
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.15456674473067916,
|
160 |
+
"grad_norm": 0.07578740268945694,
|
161 |
+
"learning_rate": 5.116279069767442e-06,
|
162 |
+
"loss": 0.5844,
|
163 |
+
"step": 22
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.16159250585480095,
|
167 |
+
"grad_norm": 0.0799824595451355,
|
168 |
+
"learning_rate": 5.348837209302326e-06,
|
169 |
+
"loss": 0.6688,
|
170 |
+
"step": 23
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.1686182669789227,
|
174 |
+
"grad_norm": 0.06833522021770477,
|
175 |
+
"learning_rate": 5.58139534883721e-06,
|
176 |
+
"loss": 0.5623,
|
177 |
+
"step": 24
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.1756440281030445,
|
181 |
+
"grad_norm": 0.07280978560447693,
|
182 |
+
"learning_rate": 5.8139534883720935e-06,
|
183 |
+
"loss": 0.6319,
|
184 |
+
"step": 25
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.18266978922716628,
|
188 |
+
"grad_norm": 0.06204582750797272,
|
189 |
+
"learning_rate": 6.046511627906977e-06,
|
190 |
+
"loss": 0.563,
|
191 |
+
"step": 26
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.18969555035128804,
|
195 |
+
"grad_norm": 0.06134812906384468,
|
196 |
+
"learning_rate": 6.279069767441861e-06,
|
197 |
+
"loss": 0.5847,
|
198 |
+
"step": 27
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.19672131147540983,
|
202 |
+
"grad_norm": 0.05748824402689934,
|
203 |
+
"learning_rate": 6.511627906976745e-06,
|
204 |
+
"loss": 0.5311,
|
205 |
+
"step": 28
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.20374707259953162,
|
209 |
+
"grad_norm": 0.06215423345565796,
|
210 |
+
"learning_rate": 6.744186046511628e-06,
|
211 |
+
"loss": 0.5514,
|
212 |
+
"step": 29
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.2107728337236534,
|
216 |
+
"grad_norm": 0.06185278296470642,
|
217 |
+
"learning_rate": 6.976744186046513e-06,
|
218 |
+
"loss": 0.5563,
|
219 |
+
"step": 30
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.21779859484777517,
|
223 |
+
"grad_norm": 0.05966101959347725,
|
224 |
+
"learning_rate": 7.209302325581395e-06,
|
225 |
+
"loss": 0.5758,
|
226 |
+
"step": 31
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.22482435597189696,
|
230 |
+
"grad_norm": 0.06163804978132248,
|
231 |
+
"learning_rate": 7.44186046511628e-06,
|
232 |
+
"loss": 0.5063,
|
233 |
+
"step": 32
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.23185011709601874,
|
237 |
+
"grad_norm": 0.06333377212285995,
|
238 |
+
"learning_rate": 7.674418604651164e-06,
|
239 |
+
"loss": 0.5583,
|
240 |
+
"step": 33
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.2388758782201405,
|
244 |
+
"grad_norm": 0.07251594960689545,
|
245 |
+
"learning_rate": 7.906976744186048e-06,
|
246 |
+
"loss": 0.5886,
|
247 |
+
"step": 34
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.2459016393442623,
|
251 |
+
"grad_norm": 0.0646405890583992,
|
252 |
+
"learning_rate": 8.139534883720931e-06,
|
253 |
+
"loss": 0.5421,
|
254 |
+
"step": 35
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.2529274004683841,
|
258 |
+
"grad_norm": 0.08668368309736252,
|
259 |
+
"learning_rate": 8.372093023255815e-06,
|
260 |
+
"loss": 0.5354,
|
261 |
+
"step": 36
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.25995316159250587,
|
265 |
+
"grad_norm": 0.06333053857088089,
|
266 |
+
"learning_rate": 8.604651162790698e-06,
|
267 |
+
"loss": 0.5425,
|
268 |
+
"step": 37
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.26697892271662765,
|
272 |
+
"grad_norm": 0.053554244339466095,
|
273 |
+
"learning_rate": 8.837209302325582e-06,
|
274 |
+
"loss": 0.4706,
|
275 |
+
"step": 38
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.27400468384074944,
|
279 |
+
"grad_norm": 0.05727384239435196,
|
280 |
+
"learning_rate": 9.069767441860465e-06,
|
281 |
+
"loss": 0.4781,
|
282 |
+
"step": 39
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.2810304449648712,
|
286 |
+
"grad_norm": 0.05613735318183899,
|
287 |
+
"learning_rate": 9.30232558139535e-06,
|
288 |
+
"loss": 0.4938,
|
289 |
+
"step": 40
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.28805620608899296,
|
293 |
+
"grad_norm": 0.05665242299437523,
|
294 |
+
"learning_rate": 9.534883720930234e-06,
|
295 |
+
"loss": 0.516,
|
296 |
+
"step": 41
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.29508196721311475,
|
300 |
+
"grad_norm": 0.06407799571752548,
|
301 |
+
"learning_rate": 9.767441860465117e-06,
|
302 |
+
"loss": 0.5219,
|
303 |
+
"step": 42
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.30210772833723654,
|
307 |
+
"grad_norm": 0.059850167483091354,
|
308 |
+
"learning_rate": 1e-05,
|
309 |
+
"loss": 0.5281,
|
310 |
+
"step": 43
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.3091334894613583,
|
314 |
+
"grad_norm": 0.058536890894174576,
|
315 |
+
"learning_rate": 9.99983179466314e-06,
|
316 |
+
"loss": 0.4646,
|
317 |
+
"step": 44
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.3161592505854801,
|
321 |
+
"grad_norm": 0.053681179881095886,
|
322 |
+
"learning_rate": 9.999327189969768e-06,
|
323 |
+
"loss": 0.4662,
|
324 |
+
"step": 45
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.3231850117096019,
|
328 |
+
"grad_norm": 0.05520308017730713,
|
329 |
+
"learning_rate": 9.998486219870769e-06,
|
330 |
+
"loss": 0.4972,
|
331 |
+
"step": 46
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.33021077283372363,
|
335 |
+
"grad_norm": 0.06100783869624138,
|
336 |
+
"learning_rate": 9.997308940948405e-06,
|
337 |
+
"loss": 0.5449,
|
338 |
+
"step": 47
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.3372365339578454,
|
342 |
+
"grad_norm": 0.05527447536587715,
|
343 |
+
"learning_rate": 9.995795432412513e-06,
|
344 |
+
"loss": 0.517,
|
345 |
+
"step": 48
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.3442622950819672,
|
349 |
+
"grad_norm": 0.06016694754362106,
|
350 |
+
"learning_rate": 9.993945796095183e-06,
|
351 |
+
"loss": 0.4996,
|
352 |
+
"step": 49
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.351288056206089,
|
356 |
+
"grad_norm": 0.061447110027074814,
|
357 |
+
"learning_rate": 9.991760156443892e-06,
|
358 |
+
"loss": 0.4678,
|
359 |
+
"step": 50
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.3583138173302108,
|
363 |
+
"grad_norm": 0.05987882614135742,
|
364 |
+
"learning_rate": 9.989238660513141e-06,
|
365 |
+
"loss": 0.5699,
|
366 |
+
"step": 51
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.36533957845433257,
|
370 |
+
"grad_norm": 0.05668003484606743,
|
371 |
+
"learning_rate": 9.98638147795456e-06,
|
372 |
+
"loss": 0.505,
|
373 |
+
"step": 52
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.37236533957845436,
|
377 |
+
"grad_norm": 0.05434773117303848,
|
378 |
+
"learning_rate": 9.983188801005492e-06,
|
379 |
+
"loss": 0.4742,
|
380 |
+
"step": 53
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.3793911007025761,
|
384 |
+
"grad_norm": 0.05131086707115173,
|
385 |
+
"learning_rate": 9.979660844476056e-06,
|
386 |
+
"loss": 0.5096,
|
387 |
+
"step": 54
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.3864168618266979,
|
391 |
+
"grad_norm": 0.058555662631988525,
|
392 |
+
"learning_rate": 9.975797845734699e-06,
|
393 |
+
"loss": 0.5277,
|
394 |
+
"step": 55
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.39344262295081966,
|
398 |
+
"grad_norm": 0.0570974200963974,
|
399 |
+
"learning_rate": 9.971600064692222e-06,
|
400 |
+
"loss": 0.4801,
|
401 |
+
"step": 56
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.40046838407494145,
|
405 |
+
"grad_norm": 0.056893255561590195,
|
406 |
+
"learning_rate": 9.967067783784297e-06,
|
407 |
+
"loss": 0.4857,
|
408 |
+
"step": 57
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.40749414519906324,
|
412 |
+
"grad_norm": 0.05847252905368805,
|
413 |
+
"learning_rate": 9.962201307952455e-06,
|
414 |
+
"loss": 0.5099,
|
415 |
+
"step": 58
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.41451990632318503,
|
419 |
+
"grad_norm": 0.06154610961675644,
|
420 |
+
"learning_rate": 9.957000964623585e-06,
|
421 |
+
"loss": 0.5099,
|
422 |
+
"step": 59
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.4215456674473068,
|
426 |
+
"grad_norm": 0.05544036626815796,
|
427 |
+
"learning_rate": 9.951467103687879e-06,
|
428 |
+
"loss": 0.4713,
|
429 |
+
"step": 60
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.42857142857142855,
|
433 |
+
"grad_norm": 0.04792679101228714,
|
434 |
+
"learning_rate": 9.945600097475322e-06,
|
435 |
+
"loss": 0.4576,
|
436 |
+
"step": 61
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.43559718969555034,
|
440 |
+
"grad_norm": 0.05695069581270218,
|
441 |
+
"learning_rate": 9.939400340730611e-06,
|
442 |
+
"loss": 0.5169,
|
443 |
+
"step": 62
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.4426229508196721,
|
447 |
+
"grad_norm": 0.049334120005369186,
|
448 |
+
"learning_rate": 9.932868250586619e-06,
|
449 |
+
"loss": 0.4828,
|
450 |
+
"step": 63
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.4496487119437939,
|
454 |
+
"grad_norm": 0.045351915061473846,
|
455 |
+
"learning_rate": 9.926004266536314e-06,
|
456 |
+
"loss": 0.4305,
|
457 |
+
"step": 64
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.4566744730679157,
|
461 |
+
"grad_norm": 0.049619413912296295,
|
462 |
+
"learning_rate": 9.918808850403192e-06,
|
463 |
+
"loss": 0.4892,
|
464 |
+
"step": 65
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.4637002341920375,
|
468 |
+
"grad_norm": 0.05420510098338127,
|
469 |
+
"learning_rate": 9.911282486310214e-06,
|
470 |
+
"loss": 0.4722,
|
471 |
+
"step": 66
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.4707259953161593,
|
475 |
+
"grad_norm": 0.06276430934667587,
|
476 |
+
"learning_rate": 9.903425680647225e-06,
|
477 |
+
"loss": 0.5461,
|
478 |
+
"step": 67
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.477751756440281,
|
482 |
+
"grad_norm": 0.06031210348010063,
|
483 |
+
"learning_rate": 9.895238962036878e-06,
|
484 |
+
"loss": 0.426,
|
485 |
+
"step": 68
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.4847775175644028,
|
489 |
+
"grad_norm": 0.05641568452119827,
|
490 |
+
"learning_rate": 9.88672288129908e-06,
|
491 |
+
"loss": 0.4954,
|
492 |
+
"step": 69
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.4918032786885246,
|
496 |
+
"grad_norm": 0.05691906809806824,
|
497 |
+
"learning_rate": 9.877878011413924e-06,
|
498 |
+
"loss": 0.4511,
|
499 |
+
"step": 70
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.49882903981264637,
|
503 |
+
"grad_norm": 0.06182294711470604,
|
504 |
+
"learning_rate": 9.868704947483134e-06,
|
505 |
+
"loss": 0.5105,
|
506 |
+
"step": 71
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.5058548009367682,
|
510 |
+
"grad_norm": 0.053402017802000046,
|
511 |
+
"learning_rate": 9.859204306690038e-06,
|
512 |
+
"loss": 0.4792,
|
513 |
+
"step": 72
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.5128805620608899,
|
517 |
+
"grad_norm": 0.08146236836910248,
|
518 |
+
"learning_rate": 9.849376728258024e-06,
|
519 |
+
"loss": 0.496,
|
520 |
+
"step": 73
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.5199063231850117,
|
524 |
+
"grad_norm": 0.051524728536605835,
|
525 |
+
"learning_rate": 9.839222873407553e-06,
|
526 |
+
"loss": 0.4674,
|
527 |
+
"step": 74
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.5269320843091335,
|
531 |
+
"grad_norm": 0.059568360447883606,
|
532 |
+
"learning_rate": 9.828743425311654e-06,
|
533 |
+
"loss": 0.4675,
|
534 |
+
"step": 75
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.5339578454332553,
|
538 |
+
"grad_norm": 0.05749664828181267,
|
539 |
+
"learning_rate": 9.817939089049964e-06,
|
540 |
+
"loss": 0.449,
|
541 |
+
"step": 76
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.5409836065573771,
|
545 |
+
"grad_norm": 0.051995862275362015,
|
546 |
+
"learning_rate": 9.806810591561295e-06,
|
547 |
+
"loss": 0.4323,
|
548 |
+
"step": 77
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.5480093676814989,
|
552 |
+
"grad_norm": 0.051814232021570206,
|
553 |
+
"learning_rate": 9.795358681594712e-06,
|
554 |
+
"loss": 0.4724,
|
555 |
+
"step": 78
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.5550351288056206,
|
559 |
+
"grad_norm": 0.056114014238119125,
|
560 |
+
"learning_rate": 9.783584129659162e-06,
|
561 |
+
"loss": 0.4983,
|
562 |
+
"step": 79
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.5620608899297423,
|
566 |
+
"grad_norm": 0.06239385902881622,
|
567 |
+
"learning_rate": 9.771487727971642e-06,
|
568 |
+
"loss": 0.4808,
|
569 |
+
"step": 80
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.5690866510538641,
|
573 |
+
"grad_norm": 0.054439231753349304,
|
574 |
+
"learning_rate": 9.759070290403873e-06,
|
575 |
+
"loss": 0.4642,
|
576 |
+
"step": 81
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.5761124121779859,
|
580 |
+
"grad_norm": 0.05044344440102577,
|
581 |
+
"learning_rate": 9.746332652427566e-06,
|
582 |
+
"loss": 0.4654,
|
583 |
+
"step": 82
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.5831381733021077,
|
587 |
+
"grad_norm": 0.06456462293863297,
|
588 |
+
"learning_rate": 9.733275671058195e-06,
|
589 |
+
"loss": 0.5002,
|
590 |
+
"step": 83
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.5901639344262295,
|
594 |
+
"grad_norm": 0.05796905234456062,
|
595 |
+
"learning_rate": 9.71990022479734e-06,
|
596 |
+
"loss": 0.4735,
|
597 |
+
"step": 84
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.5971896955503513,
|
601 |
+
"grad_norm": 0.055456362664699554,
|
602 |
+
"learning_rate": 9.70620721357358e-06,
|
603 |
+
"loss": 0.4612,
|
604 |
+
"step": 85
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.6042154566744731,
|
608 |
+
"grad_norm": 0.07052718102931976,
|
609 |
+
"learning_rate": 9.69219755868194e-06,
|
610 |
+
"loss": 0.4645,
|
611 |
+
"step": 86
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.6112412177985949,
|
615 |
+
"grad_norm": 0.057100776582956314,
|
616 |
+
"learning_rate": 9.677872202721906e-06,
|
617 |
+
"loss": 0.4949,
|
618 |
+
"step": 87
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.6182669789227166,
|
622 |
+
"grad_norm": 0.056776516139507294,
|
623 |
+
"learning_rate": 9.663232109534011e-06,
|
624 |
+
"loss": 0.5361,
|
625 |
+
"step": 88
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.6252927400468384,
|
629 |
+
"grad_norm": 0.05490226671099663,
|
630 |
+
"learning_rate": 9.648278264134977e-06,
|
631 |
+
"loss": 0.441,
|
632 |
+
"step": 89
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.6323185011709602,
|
636 |
+
"grad_norm": 0.0555921345949173,
|
637 |
+
"learning_rate": 9.633011672651443e-06,
|
638 |
+
"loss": 0.4116,
|
639 |
+
"step": 90
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.639344262295082,
|
643 |
+
"grad_norm": 0.053403131663799286,
|
644 |
+
"learning_rate": 9.617433362252277e-06,
|
645 |
+
"loss": 0.482,
|
646 |
+
"step": 91
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.6463700234192038,
|
650 |
+
"grad_norm": 0.04735114425420761,
|
651 |
+
"learning_rate": 9.601544381079457e-06,
|
652 |
+
"loss": 0.4476,
|
653 |
+
"step": 92
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.6533957845433255,
|
657 |
+
"grad_norm": 0.051030732691287994,
|
658 |
+
"learning_rate": 9.585345798177557e-06,
|
659 |
+
"loss": 0.4997,
|
660 |
+
"step": 93
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.6604215456674473,
|
664 |
+
"grad_norm": 0.054940689355134964,
|
665 |
+
"learning_rate": 9.56883870342181e-06,
|
666 |
+
"loss": 0.4406,
|
667 |
+
"step": 94
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.667447306791569,
|
671 |
+
"grad_norm": 0.06611790508031845,
|
672 |
+
"learning_rate": 9.552024207444794e-06,
|
673 |
+
"loss": 0.4291,
|
674 |
+
"step": 95
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.6744730679156908,
|
678 |
+
"grad_norm": 0.06092395260930061,
|
679 |
+
"learning_rate": 9.534903441561693e-06,
|
680 |
+
"loss": 0.5012,
|
681 |
+
"step": 96
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.6814988290398126,
|
685 |
+
"grad_norm": 0.0560709647834301,
|
686 |
+
"learning_rate": 9.517477557694182e-06,
|
687 |
+
"loss": 0.5025,
|
688 |
+
"step": 97
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.6885245901639344,
|
692 |
+
"grad_norm": 0.06219569593667984,
|
693 |
+
"learning_rate": 9.499747728292928e-06,
|
694 |
+
"loss": 0.5078,
|
695 |
+
"step": 98
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.6955503512880562,
|
699 |
+
"grad_norm": 0.05389304459095001,
|
700 |
+
"learning_rate": 9.481715146258699e-06,
|
701 |
+
"loss": 0.4275,
|
702 |
+
"step": 99
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.702576112412178,
|
706 |
+
"grad_norm": 0.055709365755319595,
|
707 |
+
"learning_rate": 9.463381024862116e-06,
|
708 |
+
"loss": 0.5152,
|
709 |
+
"step": 100
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.7096018735362998,
|
713 |
+
"grad_norm": 0.07741156220436096,
|
714 |
+
"learning_rate": 9.444746597662e-06,
|
715 |
+
"loss": 0.504,
|
716 |
+
"step": 101
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 0.7166276346604216,
|
720 |
+
"grad_norm": 0.049752719700336456,
|
721 |
+
"learning_rate": 9.425813118422393e-06,
|
722 |
+
"loss": 0.4613,
|
723 |
+
"step": 102
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 0.7236533957845434,
|
727 |
+
"grad_norm": 0.056723710149526596,
|
728 |
+
"learning_rate": 9.406581861028199e-06,
|
729 |
+
"loss": 0.5293,
|
730 |
+
"step": 103
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 0.7306791569086651,
|
734 |
+
"grad_norm": 0.05891204625368118,
|
735 |
+
"learning_rate": 9.387054119399466e-06,
|
736 |
+
"loss": 0.4229,
|
737 |
+
"step": 104
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 0.7377049180327869,
|
741 |
+
"grad_norm": 0.053549159318208694,
|
742 |
+
"learning_rate": 9.36723120740434e-06,
|
743 |
+
"loss": 0.4525,
|
744 |
+
"step": 105
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.7447306791569087,
|
748 |
+
"grad_norm": 0.05781884491443634,
|
749 |
+
"learning_rate": 9.347114458770656e-06,
|
750 |
+
"loss": 0.4756,
|
751 |
+
"step": 106
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 0.7517564402810304,
|
755 |
+
"grad_norm": 0.052228234708309174,
|
756 |
+
"learning_rate": 9.326705226996207e-06,
|
757 |
+
"loss": 0.4577,
|
758 |
+
"step": 107
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 0.7587822014051522,
|
762 |
+
"grad_norm": 0.05756888911128044,
|
763 |
+
"learning_rate": 9.306004885257675e-06,
|
764 |
+
"loss": 0.4604,
|
765 |
+
"step": 108
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.765807962529274,
|
769 |
+
"grad_norm": 0.07348482310771942,
|
770 |
+
"learning_rate": 9.28501482631824e-06,
|
771 |
+
"loss": 0.4768,
|
772 |
+
"step": 109
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 0.7728337236533958,
|
776 |
+
"grad_norm": 0.060214102268218994,
|
777 |
+
"learning_rate": 9.26373646243388e-06,
|
778 |
+
"loss": 0.4891,
|
779 |
+
"step": 110
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 0.7798594847775175,
|
783 |
+
"grad_norm": 0.05229140818119049,
|
784 |
+
"learning_rate": 9.242171225258336e-06,
|
785 |
+
"loss": 0.4719,
|
786 |
+
"step": 111
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 0.7868852459016393,
|
790 |
+
"grad_norm": 0.053298093378543854,
|
791 |
+
"learning_rate": 9.220320565746806e-06,
|
792 |
+
"loss": 0.4449,
|
793 |
+
"step": 112
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 0.7939110070257611,
|
797 |
+
"grad_norm": 0.04904168099164963,
|
798 |
+
"learning_rate": 9.198185954058305e-06,
|
799 |
+
"loss": 0.4682,
|
800 |
+
"step": 113
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 0.8009367681498829,
|
804 |
+
"grad_norm": 0.055706869810819626,
|
805 |
+
"learning_rate": 9.175768879456759e-06,
|
806 |
+
"loss": 0.4522,
|
807 |
+
"step": 114
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.8079625292740047,
|
811 |
+
"grad_norm": 0.057728879153728485,
|
812 |
+
"learning_rate": 9.153070850210803e-06,
|
813 |
+
"loss": 0.4465,
|
814 |
+
"step": 115
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 0.8149882903981265,
|
818 |
+
"grad_norm": 0.04915739968419075,
|
819 |
+
"learning_rate": 9.130093393492302e-06,
|
820 |
+
"loss": 0.4254,
|
821 |
+
"step": 116
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 0.8220140515222483,
|
825 |
+
"grad_norm": 0.04908665642142296,
|
826 |
+
"learning_rate": 9.106838055273589e-06,
|
827 |
+
"loss": 0.4492,
|
828 |
+
"step": 117
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 0.8290398126463701,
|
832 |
+
"grad_norm": 0.05924678593873978,
|
833 |
+
"learning_rate": 9.083306400223465e-06,
|
834 |
+
"loss": 0.4821,
|
835 |
+
"step": 118
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 0.8360655737704918,
|
839 |
+
"grad_norm": 0.05097891017794609,
|
840 |
+
"learning_rate": 9.059500011601919e-06,
|
841 |
+
"loss": 0.4454,
|
842 |
+
"step": 119
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 0.8430913348946136,
|
846 |
+
"grad_norm": 0.05687716603279114,
|
847 |
+
"learning_rate": 9.035420491153596e-06,
|
848 |
+
"loss": 0.4329,
|
849 |
+
"step": 120
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.8501170960187353,
|
853 |
+
"grad_norm": 0.0736871287226677,
|
854 |
+
"learning_rate": 9.011069459000035e-06,
|
855 |
+
"loss": 0.4685,
|
856 |
+
"step": 121
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 0.8571428571428571,
|
860 |
+
"grad_norm": 0.05752350762486458,
|
861 |
+
"learning_rate": 8.986448553530665e-06,
|
862 |
+
"loss": 0.4282,
|
863 |
+
"step": 122
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 0.8641686182669789,
|
867 |
+
"grad_norm": 0.05238118767738342,
|
868 |
+
"learning_rate": 8.961559431292562e-06,
|
869 |
+
"loss": 0.4381,
|
870 |
+
"step": 123
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 0.8711943793911007,
|
874 |
+
"grad_norm": 0.055724114179611206,
|
875 |
+
"learning_rate": 8.936403766879003e-06,
|
876 |
+
"loss": 0.4239,
|
877 |
+
"step": 124
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 0.8782201405152225,
|
881 |
+
"grad_norm": 0.05202113091945648,
|
882 |
+
"learning_rate": 8.910983252816794e-06,
|
883 |
+
"loss": 0.4127,
|
884 |
+
"step": 125
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 0.8852459016393442,
|
888 |
+
"grad_norm": 0.05094996467232704,
|
889 |
+
"learning_rate": 8.885299599452381e-06,
|
890 |
+
"loss": 0.447,
|
891 |
+
"step": 126
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.892271662763466,
|
895 |
+
"grad_norm": 0.060567859560251236,
|
896 |
+
"learning_rate": 8.859354534836797e-06,
|
897 |
+
"loss": 0.4433,
|
898 |
+
"step": 127
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 0.8992974238875878,
|
902 |
+
"grad_norm": 0.05051707103848457,
|
903 |
+
"learning_rate": 8.833149804609372e-06,
|
904 |
+
"loss": 0.409,
|
905 |
+
"step": 128
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 0.9063231850117096,
|
909 |
+
"grad_norm": 0.05201994627714157,
|
910 |
+
"learning_rate": 8.806687171880298e-06,
|
911 |
+
"loss": 0.4511,
|
912 |
+
"step": 129
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 0.9133489461358314,
|
916 |
+
"grad_norm": 0.049249548465013504,
|
917 |
+
"learning_rate": 8.779968417111991e-06,
|
918 |
+
"loss": 0.4365,
|
919 |
+
"step": 130
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 0.9203747072599532,
|
923 |
+
"grad_norm": 0.05699831247329712,
|
924 |
+
"learning_rate": 8.752995337999316e-06,
|
925 |
+
"loss": 0.5187,
|
926 |
+
"step": 131
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 0.927400468384075,
|
930 |
+
"grad_norm": 0.05579648166894913,
|
931 |
+
"learning_rate": 8.725769749348612e-06,
|
932 |
+
"loss": 0.4587,
|
933 |
+
"step": 132
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.9344262295081968,
|
937 |
+
"grad_norm": 0.05366150662302971,
|
938 |
+
"learning_rate": 8.698293482955605e-06,
|
939 |
+
"loss": 0.4599,
|
940 |
+
"step": 133
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 0.9414519906323185,
|
944 |
+
"grad_norm": 0.05521006882190704,
|
945 |
+
"learning_rate": 8.670568387482153e-06,
|
946 |
+
"loss": 0.4634,
|
947 |
+
"step": 134
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 0.9484777517564403,
|
951 |
+
"grad_norm": 0.058564692735672,
|
952 |
+
"learning_rate": 8.642596328331864e-06,
|
953 |
+
"loss": 0.4546,
|
954 |
+
"step": 135
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.955503512880562,
|
958 |
+
"grad_norm": 0.22925680875778198,
|
959 |
+
"learning_rate": 8.614379187524593e-06,
|
960 |
+
"loss": 0.4392,
|
961 |
+
"step": 136
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 0.9625292740046838,
|
965 |
+
"grad_norm": 0.05385170876979828,
|
966 |
+
"learning_rate": 8.585918863569806e-06,
|
967 |
+
"loss": 0.4316,
|
968 |
+
"step": 137
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 0.9695550351288056,
|
972 |
+
"grad_norm": 0.05049740523099899,
|
973 |
+
"learning_rate": 8.55721727133886e-06,
|
974 |
+
"loss": 0.4467,
|
975 |
+
"step": 138
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.9765807962529274,
|
979 |
+
"grad_norm": 0.0547189861536026,
|
980 |
+
"learning_rate": 8.528276341936146e-06,
|
981 |
+
"loss": 0.4707,
|
982 |
+
"step": 139
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 0.9836065573770492,
|
986 |
+
"grad_norm": 0.057287830859422684,
|
987 |
+
"learning_rate": 8.499098022569177e-06,
|
988 |
+
"loss": 0.4413,
|
989 |
+
"step": 140
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 0.990632318501171,
|
993 |
+
"grad_norm": 0.05672721937298775,
|
994 |
+
"learning_rate": 8.469684276417568e-06,
|
995 |
+
"loss": 0.485,
|
996 |
+
"step": 141
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 0.9976580796252927,
|
1000 |
+
"grad_norm": 0.05014313757419586,
|
1001 |
+
"learning_rate": 8.440037082500953e-06,
|
1002 |
+
"loss": 0.441,
|
1003 |
+
"step": 142
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 1.0,
|
1007 |
+
"grad_norm": 0.05014313757419586,
|
1008 |
+
"learning_rate": 8.410158435545825e-06,
|
1009 |
+
"loss": 0.4919,
|
1010 |
+
"step": 143
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 1.0070257611241218,
|
1014 |
+
"grad_norm": 0.09903538227081299,
|
1015 |
+
"learning_rate": 8.380050345851338e-06,
|
1016 |
+
"loss": 0.3345,
|
1017 |
+
"step": 144
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 1.0140515222482436,
|
1021 |
+
"grad_norm": 0.04598068818449974,
|
1022 |
+
"learning_rate": 8.349714839154035e-06,
|
1023 |
+
"loss": 0.3456,
|
1024 |
+
"step": 145
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 1.0210772833723654,
|
1028 |
+
"grad_norm": 0.05103905871510506,
|
1029 |
+
"learning_rate": 8.319153956491567e-06,
|
1030 |
+
"loss": 0.3466,
|
1031 |
+
"step": 146
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 1.0281030444964872,
|
1035 |
+
"grad_norm": 0.050713758915662766,
|
1036 |
+
"learning_rate": 8.288369754065362e-06,
|
1037 |
+
"loss": 0.3337,
|
1038 |
+
"step": 147
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 1.035128805620609,
|
1042 |
+
"grad_norm": 0.04552711173892021,
|
1043 |
+
"learning_rate": 8.257364303102275e-06,
|
1044 |
+
"loss": 0.3678,
|
1045 |
+
"step": 148
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 1.0421545667447307,
|
1049 |
+
"grad_norm": 0.05452848598361015,
|
1050 |
+
"learning_rate": 8.226139689715233e-06,
|
1051 |
+
"loss": 0.3589,
|
1052 |
+
"step": 149
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 1.0491803278688525,
|
1056 |
+
"grad_norm": 0.048317231237888336,
|
1057 |
+
"learning_rate": 8.19469801476288e-06,
|
1058 |
+
"loss": 0.3848,
|
1059 |
+
"step": 150
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 1.0562060889929743,
|
1063 |
+
"grad_norm": 0.04934811219573021,
|
1064 |
+
"learning_rate": 8.16304139370823e-06,
|
1065 |
+
"loss": 0.3017,
|
1066 |
+
"step": 151
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 1.063231850117096,
|
1070 |
+
"grad_norm": 0.04967602342367172,
|
1071 |
+
"learning_rate": 8.131171956476328e-06,
|
1072 |
+
"loss": 0.3647,
|
1073 |
+
"step": 152
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 1.0702576112412179,
|
1077 |
+
"grad_norm": 0.048318326473236084,
|
1078 |
+
"learning_rate": 8.09909184731094e-06,
|
1079 |
+
"loss": 0.3386,
|
1080 |
+
"step": 153
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 1.0772833723653397,
|
1084 |
+
"grad_norm": 0.06288264691829681,
|
1085 |
+
"learning_rate": 8.066803224630295e-06,
|
1086 |
+
"loss": 0.334,
|
1087 |
+
"step": 154
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 1.0843091334894615,
|
1091 |
+
"grad_norm": 0.05197209119796753,
|
1092 |
+
"learning_rate": 8.034308260881854e-06,
|
1093 |
+
"loss": 0.3538,
|
1094 |
+
"step": 155
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 1.0913348946135832,
|
1098 |
+
"grad_norm": 0.0542137436568737,
|
1099 |
+
"learning_rate": 8.00160914239615e-06,
|
1100 |
+
"loss": 0.3296,
|
1101 |
+
"step": 156
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 1.098360655737705,
|
1105 |
+
"grad_norm": 0.05365780368447304,
|
1106 |
+
"learning_rate": 7.968708069239672e-06,
|
1107 |
+
"loss": 0.3932,
|
1108 |
+
"step": 157
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 1.1053864168618266,
|
1112 |
+
"grad_norm": 0.060223035514354706,
|
1113 |
+
"learning_rate": 7.935607255066867e-06,
|
1114 |
+
"loss": 0.3679,
|
1115 |
+
"step": 158
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 1.1124121779859484,
|
1119 |
+
"grad_norm": 0.05058826506137848,
|
1120 |
+
"learning_rate": 7.902308926971166e-06,
|
1121 |
+
"loss": 0.3321,
|
1122 |
+
"step": 159
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 1.1194379391100702,
|
1126 |
+
"grad_norm": 0.054012175649404526,
|
1127 |
+
"learning_rate": 7.868815325335168e-06,
|
1128 |
+
"loss": 0.3501,
|
1129 |
+
"step": 160
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 1.126463700234192,
|
1133 |
+
"grad_norm": 0.061976704746484756,
|
1134 |
+
"learning_rate": 7.835128703679896e-06,
|
1135 |
+
"loss": 0.35,
|
1136 |
+
"step": 161
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 1.1334894613583137,
|
1140 |
+
"grad_norm": 0.05166299641132355,
|
1141 |
+
"learning_rate": 7.801251328513164e-06,
|
1142 |
+
"loss": 0.3521,
|
1143 |
+
"step": 162
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 1.1405152224824355,
|
1147 |
+
"grad_norm": 0.05794499069452286,
|
1148 |
+
"learning_rate": 7.767185479177092e-06,
|
1149 |
+
"loss": 0.3318,
|
1150 |
+
"step": 163
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 1.1475409836065573,
|
1154 |
+
"grad_norm": 0.05970500782132149,
|
1155 |
+
"learning_rate": 7.732933447694748e-06,
|
1156 |
+
"loss": 0.3601,
|
1157 |
+
"step": 164
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 1.154566744730679,
|
1161 |
+
"grad_norm": 0.04941580072045326,
|
1162 |
+
"learning_rate": 7.698497538615928e-06,
|
1163 |
+
"loss": 0.321,
|
1164 |
+
"step": 165
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 1.161592505854801,
|
1168 |
+
"grad_norm": 0.061307549476623535,
|
1169 |
+
"learning_rate": 7.663880068862106e-06,
|
1170 |
+
"loss": 0.3103,
|
1171 |
+
"step": 166
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 1.1686182669789227,
|
1175 |
+
"grad_norm": 0.055390097200870514,
|
1176 |
+
"learning_rate": 7.629083367570547e-06,
|
1177 |
+
"loss": 0.3561,
|
1178 |
+
"step": 167
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 1.1756440281030445,
|
1182 |
+
"grad_norm": 0.06362055987119675,
|
1183 |
+
"learning_rate": 7.594109775937595e-06,
|
1184 |
+
"loss": 0.3477,
|
1185 |
+
"step": 168
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 1.1826697892271663,
|
1189 |
+
"grad_norm": 0.05179616063833237,
|
1190 |
+
"learning_rate": 7.558961647061156e-06,
|
1191 |
+
"loss": 0.3534,
|
1192 |
+
"step": 169
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 1.189695550351288,
|
1196 |
+
"grad_norm": 0.0547102652490139,
|
1197 |
+
"learning_rate": 7.5236413457823745e-06,
|
1198 |
+
"loss": 0.3023,
|
1199 |
+
"step": 170
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 1.1967213114754098,
|
1203 |
+
"grad_norm": 0.0540938526391983,
|
1204 |
+
"learning_rate": 7.488151248526518e-06,
|
1205 |
+
"loss": 0.3144,
|
1206 |
+
"step": 171
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 1.2037470725995316,
|
1210 |
+
"grad_norm": 0.04804334416985512,
|
1211 |
+
"learning_rate": 7.452493743143092e-06,
|
1212 |
+
"loss": 0.317,
|
1213 |
+
"step": 172
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 1.2107728337236534,
|
1217 |
+
"grad_norm": 0.05935240909457207,
|
1218 |
+
"learning_rate": 7.416671228745181e-06,
|
1219 |
+
"loss": 0.3801,
|
1220 |
+
"step": 173
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 1.2177985948477752,
|
1224 |
+
"grad_norm": 0.05715787410736084,
|
1225 |
+
"learning_rate": 7.380686115548024e-06,
|
1226 |
+
"loss": 0.3286,
|
1227 |
+
"step": 174
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 1.224824355971897,
|
1231 |
+
"grad_norm": 0.05142643675208092,
|
1232 |
+
"learning_rate": 7.344540824706855e-06,
|
1233 |
+
"loss": 0.3351,
|
1234 |
+
"step": 175
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 1.2318501170960188,
|
1238 |
+
"grad_norm": 0.0654173269867897,
|
1239 |
+
"learning_rate": 7.3082377881540025e-06,
|
1240 |
+
"loss": 0.344,
|
1241 |
+
"step": 176
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 1.2388758782201406,
|
1245 |
+
"grad_norm": 0.05274007469415665,
|
1246 |
+
"learning_rate": 7.271779448435265e-06,
|
1247 |
+
"loss": 0.3427,
|
1248 |
+
"step": 177
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 1.2459016393442623,
|
1252 |
+
"grad_norm": 0.05858859047293663,
|
1253 |
+
"learning_rate": 7.235168258545569e-06,
|
1254 |
+
"loss": 0.3477,
|
1255 |
+
"step": 178
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 1.2529274004683841,
|
1259 |
+
"grad_norm": 0.05919253081083298,
|
1260 |
+
"learning_rate": 7.198406681763925e-06,
|
1261 |
+
"loss": 0.348,
|
1262 |
+
"step": 179
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 1.259953161592506,
|
1266 |
+
"grad_norm": 0.05598839744925499,
|
1267 |
+
"learning_rate": 7.161497191487693e-06,
|
1268 |
+
"loss": 0.3529,
|
1269 |
+
"step": 180
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 1.2669789227166277,
|
1273 |
+
"grad_norm": 0.04682192578911781,
|
1274 |
+
"learning_rate": 7.124442271066174e-06,
|
1275 |
+
"loss": 0.3346,
|
1276 |
+
"step": 181
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 1.2740046838407495,
|
1280 |
+
"grad_norm": 0.05666663125157356,
|
1281 |
+
"learning_rate": 7.087244413633516e-06,
|
1282 |
+
"loss": 0.3372,
|
1283 |
+
"step": 182
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 1.281030444964871,
|
1287 |
+
"grad_norm": 0.05192001909017563,
|
1288 |
+
"learning_rate": 7.049906121940974e-06,
|
1289 |
+
"loss": 0.3495,
|
1290 |
+
"step": 183
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 1.288056206088993,
|
1294 |
+
"grad_norm": 0.056901462376117706,
|
1295 |
+
"learning_rate": 7.012429908188523e-06,
|
1296 |
+
"loss": 0.3189,
|
1297 |
+
"step": 184
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 1.2950819672131146,
|
1301 |
+
"grad_norm": 0.05573394522070885,
|
1302 |
+
"learning_rate": 6.9748182938558225e-06,
|
1303 |
+
"loss": 0.3074,
|
1304 |
+
"step": 185
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 1.3021077283372366,
|
1308 |
+
"grad_norm": 0.051766857504844666,
|
1309 |
+
"learning_rate": 6.937073809532581e-06,
|
1310 |
+
"loss": 0.3801,
|
1311 |
+
"step": 186
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 1.3091334894613582,
|
1315 |
+
"grad_norm": 0.0547168143093586,
|
1316 |
+
"learning_rate": 6.899198994748274e-06,
|
1317 |
+
"loss": 0.3197,
|
1318 |
+
"step": 187
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 1.3161592505854802,
|
1322 |
+
"grad_norm": 0.055122897028923035,
|
1323 |
+
"learning_rate": 6.861196397801297e-06,
|
1324 |
+
"loss": 0.2924,
|
1325 |
+
"step": 188
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 1.3231850117096018,
|
1329 |
+
"grad_norm": 0.04852576553821564,
|
1330 |
+
"learning_rate": 6.823068575587496e-06,
|
1331 |
+
"loss": 0.3215,
|
1332 |
+
"step": 189
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 1.3302107728337236,
|
1336 |
+
"grad_norm": 0.04942547157406807,
|
1337 |
+
"learning_rate": 6.784818093428144e-06,
|
1338 |
+
"loss": 0.31,
|
1339 |
+
"step": 190
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 1.3372365339578454,
|
1343 |
+
"grad_norm": 0.05302649736404419,
|
1344 |
+
"learning_rate": 6.746447524897335e-06,
|
1345 |
+
"loss": 0.3386,
|
1346 |
+
"step": 191
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 1.3442622950819672,
|
1350 |
+
"grad_norm": 0.04981255903840065,
|
1351 |
+
"learning_rate": 6.70795945164883e-06,
|
1352 |
+
"loss": 0.3429,
|
1353 |
+
"step": 192
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 1.351288056206089,
|
1357 |
+
"grad_norm": 0.061598002910614014,
|
1358 |
+
"learning_rate": 6.6693564632423626e-06,
|
1359 |
+
"loss": 0.3444,
|
1360 |
+
"step": 193
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 1.3583138173302107,
|
1364 |
+
"grad_norm": 0.06926045566797256,
|
1365 |
+
"learning_rate": 6.630641156969397e-06,
|
1366 |
+
"loss": 0.3119,
|
1367 |
+
"step": 194
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 1.3653395784543325,
|
1371 |
+
"grad_norm": 0.05165793001651764,
|
1372 |
+
"learning_rate": 6.591816137678388e-06,
|
1373 |
+
"loss": 0.3309,
|
1374 |
+
"step": 195
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 1.3723653395784543,
|
1378 |
+
"grad_norm": 0.05770070105791092,
|
1379 |
+
"learning_rate": 6.552884017599517e-06,
|
1380 |
+
"loss": 0.3527,
|
1381 |
+
"step": 196
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 1.379391100702576,
|
1385 |
+
"grad_norm": 0.06011437624692917,
|
1386 |
+
"learning_rate": 6.513847416168929e-06,
|
1387 |
+
"loss": 0.3574,
|
1388 |
+
"step": 197
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 1.3864168618266979,
|
1392 |
+
"grad_norm": 0.05262259393930435,
|
1393 |
+
"learning_rate": 6.474708959852504e-06,
|
1394 |
+
"loss": 0.2922,
|
1395 |
+
"step": 198
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 1.3934426229508197,
|
1399 |
+
"grad_norm": 0.05124557390809059,
|
1400 |
+
"learning_rate": 6.435471281969133e-06,
|
1401 |
+
"loss": 0.3053,
|
1402 |
+
"step": 199
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 1.4004683840749415,
|
1406 |
+
"grad_norm": 0.054529063403606415,
|
1407 |
+
"learning_rate": 6.396137022513545e-06,
|
1408 |
+
"loss": 0.3311,
|
1409 |
+
"step": 200
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 1.4074941451990632,
|
1413 |
+
"grad_norm": 0.05387987568974495,
|
1414 |
+
"learning_rate": 6.3567088279786885e-06,
|
1415 |
+
"loss": 0.3508,
|
1416 |
+
"step": 201
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 1.414519906323185,
|
1420 |
+
"grad_norm": 0.052831970155239105,
|
1421 |
+
"learning_rate": 6.317189351177657e-06,
|
1422 |
+
"loss": 0.3425,
|
1423 |
+
"step": 202
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 1.4215456674473068,
|
1427 |
+
"grad_norm": 0.0602416917681694,
|
1428 |
+
"learning_rate": 6.277581251065217e-06,
|
1429 |
+
"loss": 0.3448,
|
1430 |
+
"step": 203
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 1.4285714285714286,
|
1434 |
+
"grad_norm": 0.04873733967542648,
|
1435 |
+
"learning_rate": 6.237887192558894e-06,
|
1436 |
+
"loss": 0.361,
|
1437 |
+
"step": 204
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 1.4355971896955504,
|
1441 |
+
"grad_norm": 0.05855415388941765,
|
1442 |
+
"learning_rate": 6.198109846359682e-06,
|
1443 |
+
"loss": 0.3324,
|
1444 |
+
"step": 205
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 1.4426229508196722,
|
1448 |
+
"grad_norm": 0.05669476091861725,
|
1449 |
+
"learning_rate": 6.15825188877235e-06,
|
1450 |
+
"loss": 0.3892,
|
1451 |
+
"step": 206
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 1.449648711943794,
|
1455 |
+
"grad_norm": 0.0551174134016037,
|
1456 |
+
"learning_rate": 6.118316001525368e-06,
|
1457 |
+
"loss": 0.3579,
|
1458 |
+
"step": 207
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 1.4566744730679158,
|
1462 |
+
"grad_norm": 0.05648890882730484,
|
1463 |
+
"learning_rate": 6.078304871590485e-06,
|
1464 |
+
"loss": 0.3183,
|
1465 |
+
"step": 208
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 1.4637002341920375,
|
1469 |
+
"grad_norm": 0.05304768308997154,
|
1470 |
+
"learning_rate": 6.038221191001935e-06,
|
1471 |
+
"loss": 0.3346,
|
1472 |
+
"step": 209
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 1.4707259953161593,
|
1476 |
+
"grad_norm": 0.04858344793319702,
|
1477 |
+
"learning_rate": 5.998067656675318e-06,
|
1478 |
+
"loss": 0.3545,
|
1479 |
+
"step": 210
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 1.4777517564402811,
|
1483 |
+
"grad_norm": 0.060655366629362106,
|
1484 |
+
"learning_rate": 5.95784697022614e-06,
|
1485 |
+
"loss": 0.3321,
|
1486 |
+
"step": 211
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 1.4847775175644027,
|
1490 |
+
"grad_norm": 0.053603459149599075,
|
1491 |
+
"learning_rate": 5.917561837788046e-06,
|
1492 |
+
"loss": 0.3308,
|
1493 |
+
"step": 212
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 1.4918032786885247,
|
1497 |
+
"grad_norm": 0.04807604104280472,
|
1498 |
+
"learning_rate": 5.877214969830746e-06,
|
1499 |
+
"loss": 0.2965,
|
1500 |
+
"step": 213
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 1.4988290398126463,
|
1504 |
+
"grad_norm": 0.05581279471516609,
|
1505 |
+
"learning_rate": 5.836809080977644e-06,
|
1506 |
+
"loss": 0.3358,
|
1507 |
+
"step": 214
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 1.5058548009367683,
|
1511 |
+
"grad_norm": 0.054010357707738876,
|
1512 |
+
"learning_rate": 5.7963468898232026e-06,
|
1513 |
+
"loss": 0.3361,
|
1514 |
+
"step": 215
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 1.5128805620608898,
|
1518 |
+
"grad_norm": 0.06058075278997421,
|
1519 |
+
"learning_rate": 5.755831118750016e-06,
|
1520 |
+
"loss": 0.3657,
|
1521 |
+
"step": 216
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 1.5199063231850118,
|
1525 |
+
"grad_norm": 0.05514070391654968,
|
1526 |
+
"learning_rate": 5.715264493745652e-06,
|
1527 |
+
"loss": 0.3355,
|
1528 |
+
"step": 217
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 1.5269320843091334,
|
1532 |
+
"grad_norm": 0.056756455451250076,
|
1533 |
+
"learning_rate": 5.6746497442192425e-06,
|
1534 |
+
"loss": 0.3322,
|
1535 |
+
"step": 218
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 1.5339578454332554,
|
1539 |
+
"grad_norm": 0.051219794899225235,
|
1540 |
+
"learning_rate": 5.633989602817837e-06,
|
1541 |
+
"loss": 0.314,
|
1542 |
+
"step": 219
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 1.540983606557377,
|
1546 |
+
"grad_norm": 0.05676623433828354,
|
1547 |
+
"learning_rate": 5.593286805242549e-06,
|
1548 |
+
"loss": 0.3442,
|
1549 |
+
"step": 220
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 1.548009367681499,
|
1553 |
+
"grad_norm": 0.059845082461833954,
|
1554 |
+
"learning_rate": 5.552544090064487e-06,
|
1555 |
+
"loss": 0.3441,
|
1556 |
+
"step": 221
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 1.5550351288056206,
|
1560 |
+
"grad_norm": 0.05877413973212242,
|
1561 |
+
"learning_rate": 5.5117641985405055e-06,
|
1562 |
+
"loss": 0.3165,
|
1563 |
+
"step": 222
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 1.5620608899297423,
|
1567 |
+
"grad_norm": 0.04743430018424988,
|
1568 |
+
"learning_rate": 5.47094987442876e-06,
|
1569 |
+
"loss": 0.3253,
|
1570 |
+
"step": 223
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 1.5690866510538641,
|
1574 |
+
"grad_norm": 0.05630961060523987,
|
1575 |
+
"learning_rate": 5.430103863804107e-06,
|
1576 |
+
"loss": 0.3384,
|
1577 |
+
"step": 224
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 1.576112412177986,
|
1581 |
+
"grad_norm": 0.05597255006432533,
|
1582 |
+
"learning_rate": 5.389228914873334e-06,
|
1583 |
+
"loss": 0.3488,
|
1584 |
+
"step": 225
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 1.5831381733021077,
|
1588 |
+
"grad_norm": 0.045843351632356644,
|
1589 |
+
"learning_rate": 5.348327777790262e-06,
|
1590 |
+
"loss": 0.3169,
|
1591 |
+
"step": 226
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 1.5901639344262295,
|
1595 |
+
"grad_norm": 0.049752555787563324,
|
1596 |
+
"learning_rate": 5.307403204470711e-06,
|
1597 |
+
"loss": 0.3393,
|
1598 |
+
"step": 227
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 1.5971896955503513,
|
1602 |
+
"grad_norm": 0.059289898723363876,
|
1603 |
+
"learning_rate": 5.266457948407336e-06,
|
1604 |
+
"loss": 0.3531,
|
1605 |
+
"step": 228
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 1.604215456674473,
|
1609 |
+
"grad_norm": 0.05609236657619476,
|
1610 |
+
"learning_rate": 5.2254947644843735e-06,
|
1611 |
+
"loss": 0.3384,
|
1612 |
+
"step": 229
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 1.6112412177985949,
|
1616 |
+
"grad_norm": 0.053165484219789505,
|
1617 |
+
"learning_rate": 5.18451640879228e-06,
|
1618 |
+
"loss": 0.32,
|
1619 |
+
"step": 230
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 1.6182669789227166,
|
1623 |
+
"grad_norm": 0.05502665415406227,
|
1624 |
+
"learning_rate": 5.14352563844231e-06,
|
1625 |
+
"loss": 0.2832,
|
1626 |
+
"step": 231
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 1.6252927400468384,
|
1630 |
+
"grad_norm": 0.04645336791872978,
|
1631 |
+
"learning_rate": 5.1025252113809945e-06,
|
1632 |
+
"loss": 0.3047,
|
1633 |
+
"step": 232
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 1.6323185011709602,
|
1637 |
+
"grad_norm": 0.05163678526878357,
|
1638 |
+
"learning_rate": 5.061517886204592e-06,
|
1639 |
+
"loss": 0.3143,
|
1640 |
+
"step": 233
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 1.639344262295082,
|
1644 |
+
"grad_norm": 0.05668428912758827,
|
1645 |
+
"learning_rate": 5.02050642197348e-06,
|
1646 |
+
"loss": 0.334,
|
1647 |
+
"step": 234
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 1.6463700234192038,
|
1651 |
+
"grad_norm": 0.062112439423799515,
|
1652 |
+
"learning_rate": 4.979493578026523e-06,
|
1653 |
+
"loss": 0.3259,
|
1654 |
+
"step": 235
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 1.6533957845433256,
|
1658 |
+
"grad_norm": 0.050369080156087875,
|
1659 |
+
"learning_rate": 4.9384821137954106e-06,
|
1660 |
+
"loss": 0.3506,
|
1661 |
+
"step": 236
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 1.6604215456674472,
|
1665 |
+
"grad_norm": 0.0596189871430397,
|
1666 |
+
"learning_rate": 4.897474788619007e-06,
|
1667 |
+
"loss": 0.3211,
|
1668 |
+
"step": 237
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 1.6674473067915692,
|
1672 |
+
"grad_norm": 0.05669752135872841,
|
1673 |
+
"learning_rate": 4.856474361557692e-06,
|
1674 |
+
"loss": 0.3037,
|
1675 |
+
"step": 238
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 1.6744730679156907,
|
1679 |
+
"grad_norm": 0.048729464411735535,
|
1680 |
+
"learning_rate": 4.815483591207721e-06,
|
1681 |
+
"loss": 0.3546,
|
1682 |
+
"step": 239
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 1.6814988290398127,
|
1686 |
+
"grad_norm": 0.05216846615076065,
|
1687 |
+
"learning_rate": 4.774505235515628e-06,
|
1688 |
+
"loss": 0.3619,
|
1689 |
+
"step": 240
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 1.6885245901639343,
|
1693 |
+
"grad_norm": 0.05519293621182442,
|
1694 |
+
"learning_rate": 4.733542051592665e-06,
|
1695 |
+
"loss": 0.3097,
|
1696 |
+
"step": 241
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 1.6955503512880563,
|
1700 |
+
"grad_norm": 0.05697103589773178,
|
1701 |
+
"learning_rate": 4.69259679552929e-06,
|
1702 |
+
"loss": 0.343,
|
1703 |
+
"step": 242
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 1.7025761124121779,
|
1707 |
+
"grad_norm": 0.057407498359680176,
|
1708 |
+
"learning_rate": 4.651672222209738e-06,
|
1709 |
+
"loss": 0.3594,
|
1710 |
+
"step": 243
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 1.7096018735362999,
|
1714 |
+
"grad_norm": 0.0562685951590538,
|
1715 |
+
"learning_rate": 4.6107710851266695e-06,
|
1716 |
+
"loss": 0.3212,
|
1717 |
+
"step": 244
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 1.7166276346604215,
|
1721 |
+
"grad_norm": 0.0513756088912487,
|
1722 |
+
"learning_rate": 4.5698961361958955e-06,
|
1723 |
+
"loss": 0.3164,
|
1724 |
+
"step": 245
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 1.7236533957845435,
|
1728 |
+
"grad_norm": 0.05124720558524132,
|
1729 |
+
"learning_rate": 4.529050125571241e-06,
|
1730 |
+
"loss": 0.3212,
|
1731 |
+
"step": 246
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 1.730679156908665,
|
1735 |
+
"grad_norm": 0.052108533680438995,
|
1736 |
+
"learning_rate": 4.488235801459495e-06,
|
1737 |
+
"loss": 0.3056,
|
1738 |
+
"step": 247
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 1.737704918032787,
|
1742 |
+
"grad_norm": 0.058752771466970444,
|
1743 |
+
"learning_rate": 4.447455909935513e-06,
|
1744 |
+
"loss": 0.3539,
|
1745 |
+
"step": 248
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 1.7447306791569086,
|
1749 |
+
"grad_norm": 0.053762953728437424,
|
1750 |
+
"learning_rate": 4.4067131947574515e-06,
|
1751 |
+
"loss": 0.3157,
|
1752 |
+
"step": 249
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 1.7517564402810304,
|
1756 |
+
"grad_norm": 0.04892832413315773,
|
1757 |
+
"learning_rate": 4.3660103971821635e-06,
|
1758 |
+
"loss": 0.3279,
|
1759 |
+
"step": 250
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 1.7587822014051522,
|
1763 |
+
"grad_norm": 0.0599125474691391,
|
1764 |
+
"learning_rate": 4.3253502557807575e-06,
|
1765 |
+
"loss": 0.3255,
|
1766 |
+
"step": 251
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 1.765807962529274,
|
1770 |
+
"grad_norm": 0.05481419712305069,
|
1771 |
+
"learning_rate": 4.28473550625435e-06,
|
1772 |
+
"loss": 0.3502,
|
1773 |
+
"step": 252
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 1.7728337236533958,
|
1777 |
+
"grad_norm": 0.05331256613135338,
|
1778 |
+
"learning_rate": 4.244168881249986e-06,
|
1779 |
+
"loss": 0.3565,
|
1780 |
+
"step": 253
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 1.7798594847775175,
|
1784 |
+
"grad_norm": 0.053473543375730515,
|
1785 |
+
"learning_rate": 4.203653110176798e-06,
|
1786 |
+
"loss": 0.2904,
|
1787 |
+
"step": 254
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 1.7868852459016393,
|
1791 |
+
"grad_norm": 0.053232353180646896,
|
1792 |
+
"learning_rate": 4.163190919022357e-06,
|
1793 |
+
"loss": 0.321,
|
1794 |
+
"step": 255
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 1.7939110070257611,
|
1798 |
+
"grad_norm": 0.05323438718914986,
|
1799 |
+
"learning_rate": 4.122785030169256e-06,
|
1800 |
+
"loss": 0.3562,
|
1801 |
+
"step": 256
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 1.800936768149883,
|
1805 |
+
"grad_norm": 0.049881111830472946,
|
1806 |
+
"learning_rate": 4.082438162211955e-06,
|
1807 |
+
"loss": 0.3218,
|
1808 |
+
"step": 257
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 1.8079625292740047,
|
1812 |
+
"grad_norm": 0.2167116105556488,
|
1813 |
+
"learning_rate": 4.042153029773861e-06,
|
1814 |
+
"loss": 0.3218,
|
1815 |
+
"step": 258
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 1.8149882903981265,
|
1819 |
+
"grad_norm": 0.051224831491708755,
|
1820 |
+
"learning_rate": 4.001932343324683e-06,
|
1821 |
+
"loss": 0.3317,
|
1822 |
+
"step": 259
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 1.8220140515222483,
|
1826 |
+
"grad_norm": 0.10497119277715683,
|
1827 |
+
"learning_rate": 3.961778808998066e-06,
|
1828 |
+
"loss": 0.3676,
|
1829 |
+
"step": 260
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 1.82903981264637,
|
1833 |
+
"grad_norm": 0.05344713106751442,
|
1834 |
+
"learning_rate": 3.921695128409517e-06,
|
1835 |
+
"loss": 0.3619,
|
1836 |
+
"step": 261
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 1.8360655737704918,
|
1840 |
+
"grad_norm": 0.04895958676934242,
|
1841 |
+
"learning_rate": 3.8816839984746334e-06,
|
1842 |
+
"loss": 0.2942,
|
1843 |
+
"step": 262
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 1.8430913348946136,
|
1847 |
+
"grad_norm": 0.05228881910443306,
|
1848 |
+
"learning_rate": 3.841748111227652e-06,
|
1849 |
+
"loss": 0.3259,
|
1850 |
+
"step": 263
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 1.8501170960187352,
|
1854 |
+
"grad_norm": 0.04963938146829605,
|
1855 |
+
"learning_rate": 3.8018901536403198e-06,
|
1856 |
+
"loss": 0.3582,
|
1857 |
+
"step": 264
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 1.8571428571428572,
|
1861 |
+
"grad_norm": 0.056527718901634216,
|
1862 |
+
"learning_rate": 3.762112807441108e-06,
|
1863 |
+
"loss": 0.3681,
|
1864 |
+
"step": 265
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 1.8641686182669788,
|
1868 |
+
"grad_norm": 0.05749901384115219,
|
1869 |
+
"learning_rate": 3.7224187489347847e-06,
|
1870 |
+
"loss": 0.3329,
|
1871 |
+
"step": 266
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 1.8711943793911008,
|
1875 |
+
"grad_norm": 0.05997895449399948,
|
1876 |
+
"learning_rate": 3.682810648822343e-06,
|
1877 |
+
"loss": 0.3636,
|
1878 |
+
"step": 267
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 1.8782201405152223,
|
1882 |
+
"grad_norm": 0.05262213572859764,
|
1883 |
+
"learning_rate": 3.6432911720213127e-06,
|
1884 |
+
"loss": 0.3473,
|
1885 |
+
"step": 268
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 1.8852459016393444,
|
1889 |
+
"grad_norm": 0.05787108838558197,
|
1890 |
+
"learning_rate": 3.6038629774864563e-06,
|
1891 |
+
"loss": 0.3231,
|
1892 |
+
"step": 269
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 1.892271662763466,
|
1896 |
+
"grad_norm": 0.06611717492341995,
|
1897 |
+
"learning_rate": 3.56452871803087e-06,
|
1898 |
+
"loss": 0.3687,
|
1899 |
+
"step": 270
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 1.899297423887588,
|
1903 |
+
"grad_norm": 0.056811776012182236,
|
1904 |
+
"learning_rate": 3.525291040147498e-06,
|
1905 |
+
"loss": 0.298,
|
1906 |
+
"step": 271
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 1.9063231850117095,
|
1910 |
+
"grad_norm": 0.0592680424451828,
|
1911 |
+
"learning_rate": 3.486152583831072e-06,
|
1912 |
+
"loss": 0.3253,
|
1913 |
+
"step": 272
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 1.9133489461358315,
|
1917 |
+
"grad_norm": 0.05636563152074814,
|
1918 |
+
"learning_rate": 3.447115982400485e-06,
|
1919 |
+
"loss": 0.3421,
|
1920 |
+
"step": 273
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 1.920374707259953,
|
1924 |
+
"grad_norm": 0.05371502414345741,
|
1925 |
+
"learning_rate": 3.4081838623216124e-06,
|
1926 |
+
"loss": 0.3643,
|
1927 |
+
"step": 274
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 1.927400468384075,
|
1931 |
+
"grad_norm": 0.05231010541319847,
|
1932 |
+
"learning_rate": 3.3693588430306035e-06,
|
1933 |
+
"loss": 0.2898,
|
1934 |
+
"step": 275
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 1.9344262295081966,
|
1938 |
+
"grad_norm": 0.055993061512708664,
|
1939 |
+
"learning_rate": 3.330643536757638e-06,
|
1940 |
+
"loss": 0.3231,
|
1941 |
+
"step": 276
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 1.9414519906323187,
|
1945 |
+
"grad_norm": 0.05135795846581459,
|
1946 |
+
"learning_rate": 3.2920405483511702e-06,
|
1947 |
+
"loss": 0.3241,
|
1948 |
+
"step": 277
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 1.9484777517564402,
|
1952 |
+
"grad_norm": 0.05628650262951851,
|
1953 |
+
"learning_rate": 3.253552475102668e-06,
|
1954 |
+
"loss": 0.3325,
|
1955 |
+
"step": 278
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 1.955503512880562,
|
1959 |
+
"grad_norm": 0.05219852551817894,
|
1960 |
+
"learning_rate": 3.215181906571858e-06,
|
1961 |
+
"loss": 0.3499,
|
1962 |
+
"step": 279
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 1.9625292740046838,
|
1966 |
+
"grad_norm": 0.05198737978935242,
|
1967 |
+
"learning_rate": 3.1769314244125056e-06,
|
1968 |
+
"loss": 0.3252,
|
1969 |
+
"step": 280
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 1.9695550351288056,
|
1973 |
+
"grad_norm": 0.0620955154299736,
|
1974 |
+
"learning_rate": 3.1388036021987047e-06,
|
1975 |
+
"loss": 0.38,
|
1976 |
+
"step": 281
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 1.9765807962529274,
|
1980 |
+
"grad_norm": 0.05170467123389244,
|
1981 |
+
"learning_rate": 3.100801005251727e-06,
|
1982 |
+
"loss": 0.3206,
|
1983 |
+
"step": 282
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 1.9836065573770492,
|
1987 |
+
"grad_norm": 0.05948978289961815,
|
1988 |
+
"learning_rate": 3.0629261904674206e-06,
|
1989 |
+
"loss": 0.3355,
|
1990 |
+
"step": 283
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 1.990632318501171,
|
1994 |
+
"grad_norm": 0.055122461169958115,
|
1995 |
+
"learning_rate": 3.025181706144178e-06,
|
1996 |
+
"loss": 0.2983,
|
1997 |
+
"step": 284
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 1.9976580796252927,
|
2001 |
+
"grad_norm": 0.05636753514409065,
|
2002 |
+
"learning_rate": 2.987570091811479e-06,
|
2003 |
+
"loss": 0.3636,
|
2004 |
+
"step": 285
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 2.0,
|
2008 |
+
"grad_norm": 0.05636753514409065,
|
2009 |
+
"learning_rate": 2.9500938780590276e-06,
|
2010 |
+
"loss": 0.3186,
|
2011 |
+
"step": 286
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 2.0070257611241216,
|
2015 |
+
"grad_norm": 0.10458986461162567,
|
2016 |
+
"learning_rate": 2.9127555863664857e-06,
|
2017 |
+
"loss": 0.254,
|
2018 |
+
"step": 287
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 2.0140515222482436,
|
2022 |
+
"grad_norm": 0.059524934738874435,
|
2023 |
+
"learning_rate": 2.8755577289338267e-06,
|
2024 |
+
"loss": 0.2405,
|
2025 |
+
"step": 288
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 2.021077283372365,
|
2029 |
+
"grad_norm": 0.0579247772693634,
|
2030 |
+
"learning_rate": 2.838502808512309e-06,
|
2031 |
+
"loss": 0.2159,
|
2032 |
+
"step": 289
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 2.028103044496487,
|
2036 |
+
"grad_norm": 0.06685949116945267,
|
2037 |
+
"learning_rate": 2.801593318236078e-06,
|
2038 |
+
"loss": 0.2268,
|
2039 |
+
"step": 290
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 2.0351288056206087,
|
2043 |
+
"grad_norm": 0.05411679297685623,
|
2044 |
+
"learning_rate": 2.764831741454432e-06,
|
2045 |
+
"loss": 0.2579,
|
2046 |
+
"step": 291
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 2.0421545667447307,
|
2050 |
+
"grad_norm": 0.058176521211862564,
|
2051 |
+
"learning_rate": 2.7282205515647348e-06,
|
2052 |
+
"loss": 0.195,
|
2053 |
+
"step": 292
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 2.0491803278688523,
|
2057 |
+
"grad_norm": 0.051991887390613556,
|
2058 |
+
"learning_rate": 2.6917622118459975e-06,
|
2059 |
+
"loss": 0.2002,
|
2060 |
+
"step": 293
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 2.0562060889929743,
|
2064 |
+
"grad_norm": 0.06430847942829132,
|
2065 |
+
"learning_rate": 2.655459175293146e-06,
|
2066 |
+
"loss": 0.1863,
|
2067 |
+
"step": 294
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 2.063231850117096,
|
2071 |
+
"grad_norm": 0.06031821295619011,
|
2072 |
+
"learning_rate": 2.6193138844519785e-06,
|
2073 |
+
"loss": 0.2521,
|
2074 |
+
"step": 295
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 2.070257611241218,
|
2078 |
+
"grad_norm": 0.07283060252666473,
|
2079 |
+
"learning_rate": 2.58332877125482e-06,
|
2080 |
+
"loss": 0.2138,
|
2081 |
+
"step": 296
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 2.0772833723653394,
|
2085 |
+
"grad_norm": 0.05646536499261856,
|
2086 |
+
"learning_rate": 2.5475062568569077e-06,
|
2087 |
+
"loss": 0.2463,
|
2088 |
+
"step": 297
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 2.0843091334894615,
|
2092 |
+
"grad_norm": 0.0762052983045578,
|
2093 |
+
"learning_rate": 2.511848751473485e-06,
|
2094 |
+
"loss": 0.2135,
|
2095 |
+
"step": 298
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 2.091334894613583,
|
2099 |
+
"grad_norm": 0.04977189749479294,
|
2100 |
+
"learning_rate": 2.476358654217627e-06,
|
2101 |
+
"loss": 0.1957,
|
2102 |
+
"step": 299
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 2.098360655737705,
|
2106 |
+
"grad_norm": 0.05424179881811142,
|
2107 |
+
"learning_rate": 2.4410383529388448e-06,
|
2108 |
+
"loss": 0.2154,
|
2109 |
+
"step": 300
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 2.1053864168618266,
|
2113 |
+
"grad_norm": 0.05405616760253906,
|
2114 |
+
"learning_rate": 2.405890224062406e-06,
|
2115 |
+
"loss": 0.2171,
|
2116 |
+
"step": 301
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 2.1124121779859486,
|
2120 |
+
"grad_norm": 0.050341855734586716,
|
2121 |
+
"learning_rate": 2.370916632429455e-06,
|
2122 |
+
"loss": 0.1926,
|
2123 |
+
"step": 302
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 2.11943793911007,
|
2127 |
+
"grad_norm": 0.05347389727830887,
|
2128 |
+
"learning_rate": 2.336119931137897e-06,
|
2129 |
+
"loss": 0.2238,
|
2130 |
+
"step": 303
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 2.126463700234192,
|
2134 |
+
"grad_norm": 0.061148934066295624,
|
2135 |
+
"learning_rate": 2.3015024613840742e-06,
|
2136 |
+
"loss": 0.2218,
|
2137 |
+
"step": 304
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 2.1334894613583137,
|
2141 |
+
"grad_norm": 0.056503091007471085,
|
2142 |
+
"learning_rate": 2.2670665523052534e-06,
|
2143 |
+
"loss": 0.2147,
|
2144 |
+
"step": 305
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 2.1405152224824358,
|
2148 |
+
"grad_norm": 0.05360401049256325,
|
2149 |
+
"learning_rate": 2.2328145208229096e-06,
|
2150 |
+
"loss": 0.1889,
|
2151 |
+
"step": 306
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 2.1475409836065573,
|
2155 |
+
"grad_norm": 0.05166396126151085,
|
2156 |
+
"learning_rate": 2.1987486714868384e-06,
|
2157 |
+
"loss": 0.1911,
|
2158 |
+
"step": 307
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 2.1545667447306793,
|
2162 |
+
"grad_norm": 0.05574238300323486,
|
2163 |
+
"learning_rate": 2.164871296320106e-06,
|
2164 |
+
"loss": 0.2184,
|
2165 |
+
"step": 308
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 2.161592505854801,
|
2169 |
+
"grad_norm": 0.05423097312450409,
|
2170 |
+
"learning_rate": 2.1311846746648325e-06,
|
2171 |
+
"loss": 0.1904,
|
2172 |
+
"step": 309
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 2.168618266978923,
|
2176 |
+
"grad_norm": 0.04529114067554474,
|
2177 |
+
"learning_rate": 2.097691073028836e-06,
|
2178 |
+
"loss": 0.2049,
|
2179 |
+
"step": 310
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 2.1756440281030445,
|
2183 |
+
"grad_norm": 0.050600163638591766,
|
2184 |
+
"learning_rate": 2.064392744933135e-06,
|
2185 |
+
"loss": 0.2136,
|
2186 |
+
"step": 311
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 2.1826697892271665,
|
2190 |
+
"grad_norm": 0.04993721842765808,
|
2191 |
+
"learning_rate": 2.0312919307603286e-06,
|
2192 |
+
"loss": 0.1835,
|
2193 |
+
"step": 312
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 2.189695550351288,
|
2197 |
+
"grad_norm": 0.050447553396224976,
|
2198 |
+
"learning_rate": 1.998390857603853e-06,
|
2199 |
+
"loss": 0.202,
|
2200 |
+
"step": 313
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 2.19672131147541,
|
2204 |
+
"grad_norm": 0.05513535067439079,
|
2205 |
+
"learning_rate": 1.965691739118146e-06,
|
2206 |
+
"loss": 0.1964,
|
2207 |
+
"step": 314
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 2.2037470725995316,
|
2211 |
+
"grad_norm": 0.05925602465867996,
|
2212 |
+
"learning_rate": 1.9331967753697077e-06,
|
2213 |
+
"loss": 0.1935,
|
2214 |
+
"step": 315
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 2.210772833723653,
|
2218 |
+
"grad_norm": 0.05441617593169212,
|
2219 |
+
"learning_rate": 1.9009081526890622e-06,
|
2220 |
+
"loss": 0.1894,
|
2221 |
+
"step": 316
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 2.217798594847775,
|
2225 |
+
"grad_norm": 0.06225774437189102,
|
2226 |
+
"learning_rate": 1.8688280435236732e-06,
|
2227 |
+
"loss": 0.1849,
|
2228 |
+
"step": 317
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 2.2248243559718968,
|
2232 |
+
"grad_norm": 0.05718431621789932,
|
2233 |
+
"learning_rate": 1.8369586062917693e-06,
|
2234 |
+
"loss": 0.2327,
|
2235 |
+
"step": 318
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 2.2318501170960188,
|
2239 |
+
"grad_norm": 0.05463183671236038,
|
2240 |
+
"learning_rate": 1.8053019852371195e-06,
|
2241 |
+
"loss": 0.2159,
|
2242 |
+
"step": 319
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 2.2388758782201403,
|
2246 |
+
"grad_norm": 0.05712981894612312,
|
2247 |
+
"learning_rate": 1.7738603102847696e-06,
|
2248 |
+
"loss": 0.2057,
|
2249 |
+
"step": 320
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 2.2459016393442623,
|
2253 |
+
"grad_norm": 0.05346499755978584,
|
2254 |
+
"learning_rate": 1.7426356968977265e-06,
|
2255 |
+
"loss": 0.2406,
|
2256 |
+
"step": 321
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 2.252927400468384,
|
2260 |
+
"grad_norm": 0.06216704845428467,
|
2261 |
+
"learning_rate": 1.711630245934638e-06,
|
2262 |
+
"loss": 0.2307,
|
2263 |
+
"step": 322
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 2.259953161592506,
|
2267 |
+
"grad_norm": 0.04996800422668457,
|
2268 |
+
"learning_rate": 1.6808460435084316e-06,
|
2269 |
+
"loss": 0.2145,
|
2270 |
+
"step": 323
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 2.2669789227166275,
|
2274 |
+
"grad_norm": 0.05448152869939804,
|
2275 |
+
"learning_rate": 1.6502851608459668e-06,
|
2276 |
+
"loss": 0.2114,
|
2277 |
+
"step": 324
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 2.2740046838407495,
|
2281 |
+
"grad_norm": 0.05008429288864136,
|
2282 |
+
"learning_rate": 1.6199496541486647e-06,
|
2283 |
+
"loss": 0.1881,
|
2284 |
+
"step": 325
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 2.281030444964871,
|
2288 |
+
"grad_norm": 0.04923682287335396,
|
2289 |
+
"learning_rate": 1.589841564454176e-06,
|
2290 |
+
"loss": 0.2114,
|
2291 |
+
"step": 326
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 2.288056206088993,
|
2295 |
+
"grad_norm": 0.04919268935918808,
|
2296 |
+
"learning_rate": 1.5599629174990482e-06,
|
2297 |
+
"loss": 0.2149,
|
2298 |
+
"step": 327
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 2.2950819672131146,
|
2302 |
+
"grad_norm": 0.05411992594599724,
|
2303 |
+
"learning_rate": 1.5303157235824323e-06,
|
2304 |
+
"loss": 0.2034,
|
2305 |
+
"step": 328
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 2.3021077283372366,
|
2309 |
+
"grad_norm": 0.047855135053396225,
|
2310 |
+
"learning_rate": 1.5009019774308249e-06,
|
2311 |
+
"loss": 0.1977,
|
2312 |
+
"step": 329
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 2.309133489461358,
|
2316 |
+
"grad_norm": 0.05060090869665146,
|
2317 |
+
"learning_rate": 1.471723658063856e-06,
|
2318 |
+
"loss": 0.226,
|
2319 |
+
"step": 330
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 2.3161592505854802,
|
2323 |
+
"grad_norm": 0.05514250695705414,
|
2324 |
+
"learning_rate": 1.4427827286611412e-06,
|
2325 |
+
"loss": 0.2213,
|
2326 |
+
"step": 331
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 2.323185011709602,
|
2330 |
+
"grad_norm": 0.05435176566243172,
|
2331 |
+
"learning_rate": 1.4140811364301931e-06,
|
2332 |
+
"loss": 0.1782,
|
2333 |
+
"step": 332
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 2.330210772833724,
|
2337 |
+
"grad_norm": 0.05435670167207718,
|
2338 |
+
"learning_rate": 1.385620812475409e-06,
|
2339 |
+
"loss": 0.2196,
|
2340 |
+
"step": 333
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 2.3372365339578454,
|
2344 |
+
"grad_norm": 0.060686711221933365,
|
2345 |
+
"learning_rate": 1.3574036716681366e-06,
|
2346 |
+
"loss": 0.2386,
|
2347 |
+
"step": 334
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 2.3442622950819674,
|
2351 |
+
"grad_norm": 0.045962344855070114,
|
2352 |
+
"learning_rate": 1.3294316125178474e-06,
|
2353 |
+
"loss": 0.2073,
|
2354 |
+
"step": 335
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 2.351288056206089,
|
2358 |
+
"grad_norm": 0.05140933021903038,
|
2359 |
+
"learning_rate": 1.301706517044395e-06,
|
2360 |
+
"loss": 0.2094,
|
2361 |
+
"step": 336
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 2.358313817330211,
|
2365 |
+
"grad_norm": 0.052996646612882614,
|
2366 |
+
"learning_rate": 1.2742302506513894e-06,
|
2367 |
+
"loss": 0.2183,
|
2368 |
+
"step": 337
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 2.3653395784543325,
|
2372 |
+
"grad_norm": 0.060029152780771255,
|
2373 |
+
"learning_rate": 1.247004662000686e-06,
|
2374 |
+
"loss": 0.2298,
|
2375 |
+
"step": 338
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 2.3723653395784545,
|
2379 |
+
"grad_norm": 0.0602019801735878,
|
2380 |
+
"learning_rate": 1.2200315828880094e-06,
|
2381 |
+
"loss": 0.2162,
|
2382 |
+
"step": 339
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 2.379391100702576,
|
2386 |
+
"grad_norm": 0.0632694885134697,
|
2387 |
+
"learning_rate": 1.1933128281197042e-06,
|
2388 |
+
"loss": 0.2032,
|
2389 |
+
"step": 340
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 2.3864168618266977,
|
2393 |
+
"grad_norm": 0.05728806555271149,
|
2394 |
+
"learning_rate": 1.166850195390628e-06,
|
2395 |
+
"loss": 0.238,
|
2396 |
+
"step": 341
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 2.3934426229508197,
|
2400 |
+
"grad_norm": 0.05771248787641525,
|
2401 |
+
"learning_rate": 1.1406454651632042e-06,
|
2402 |
+
"loss": 0.2385,
|
2403 |
+
"step": 342
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 2.4004683840749417,
|
2407 |
+
"grad_norm": 0.09708061069250107,
|
2408 |
+
"learning_rate": 1.1147004005476192e-06,
|
2409 |
+
"loss": 0.2367,
|
2410 |
+
"step": 343
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 2.4074941451990632,
|
2414 |
+
"grad_norm": 0.05926092714071274,
|
2415 |
+
"learning_rate": 1.089016747183208e-06,
|
2416 |
+
"loss": 0.1787,
|
2417 |
+
"step": 344
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 2.414519906323185,
|
2421 |
+
"grad_norm": 0.05011197552084923,
|
2422 |
+
"learning_rate": 1.063596233120997e-06,
|
2423 |
+
"loss": 0.2119,
|
2424 |
+
"step": 345
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 2.421545667447307,
|
2428 |
+
"grad_norm": 0.05037648603320122,
|
2429 |
+
"learning_rate": 1.03844056870744e-06,
|
2430 |
+
"loss": 0.2177,
|
2431 |
+
"step": 346
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 2.4285714285714284,
|
2435 |
+
"grad_norm": 0.05578481778502464,
|
2436 |
+
"learning_rate": 1.013551446469337e-06,
|
2437 |
+
"loss": 0.1988,
|
2438 |
+
"step": 347
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 2.4355971896955504,
|
2442 |
+
"grad_norm": 0.05308612063527107,
|
2443 |
+
"learning_rate": 9.889305409999656e-07,
|
2444 |
+
"loss": 0.2006,
|
2445 |
+
"step": 348
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 2.442622950819672,
|
2449 |
+
"grad_norm": 0.05964551866054535,
|
2450 |
+
"learning_rate": 9.64579508846405e-07,
|
2451 |
+
"loss": 0.2598,
|
2452 |
+
"step": 349
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 2.449648711943794,
|
2456 |
+
"grad_norm": 0.057544246315956116,
|
2457 |
+
"learning_rate": 9.40499988398082e-07,
|
2458 |
+
"loss": 0.2095,
|
2459 |
+
"step": 350
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 2.4566744730679155,
|
2463 |
+
"grad_norm": 0.05650878697633743,
|
2464 |
+
"learning_rate": 9.166935997765364e-07,
|
2465 |
+
"loss": 0.2347,
|
2466 |
+
"step": 351
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 2.4637002341920375,
|
2470 |
+
"grad_norm": 0.07671873271465302,
|
2471 |
+
"learning_rate": 8.93161944726414e-07,
|
2472 |
+
"loss": 0.2145,
|
2473 |
+
"step": 352
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 2.470725995316159,
|
2477 |
+
"grad_norm": 0.0586552731692791,
|
2478 |
+
"learning_rate": 8.699066065077005e-07,
|
2479 |
+
"loss": 0.2078,
|
2480 |
+
"step": 353
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 2.477751756440281,
|
2484 |
+
"grad_norm": 0.045289915055036545,
|
2485 |
+
"learning_rate": 8.469291497891979e-07,
|
2486 |
+
"loss": 0.2129,
|
2487 |
+
"step": 354
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 2.4847775175644027,
|
2491 |
+
"grad_norm": 0.05548376590013504,
|
2492 |
+
"learning_rate": 8.242311205432418e-07,
|
2493 |
+
"loss": 0.1944,
|
2494 |
+
"step": 355
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 2.4918032786885247,
|
2498 |
+
"grad_norm": 0.04866700619459152,
|
2499 |
+
"learning_rate": 8.018140459416962e-07,
|
2500 |
+
"loss": 0.2228,
|
2501 |
+
"step": 356
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 2.4988290398126463,
|
2505 |
+
"grad_norm": 0.05934524163603783,
|
2506 |
+
"learning_rate": 7.796794342531949e-07,
|
2507 |
+
"loss": 0.2155,
|
2508 |
+
"step": 357
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 2.5058548009367683,
|
2512 |
+
"grad_norm": 0.056303609162569046,
|
2513 |
+
"learning_rate": 7.57828774741664e-07,
|
2514 |
+
"loss": 0.2048,
|
2515 |
+
"step": 358
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 2.51288056206089,
|
2519 |
+
"grad_norm": 0.05041569098830223,
|
2520 |
+
"learning_rate": 7.362635375661225e-07,
|
2521 |
+
"loss": 0.224,
|
2522 |
+
"step": 359
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 2.519906323185012,
|
2526 |
+
"grad_norm": 0.05846166983246803,
|
2527 |
+
"learning_rate": 7.149851736817609e-07,
|
2528 |
+
"loss": 0.2502,
|
2529 |
+
"step": 360
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 2.5269320843091334,
|
2533 |
+
"grad_norm": 0.0575077049434185,
|
2534 |
+
"learning_rate": 6.939951147423269e-07,
|
2535 |
+
"loss": 0.1915,
|
2536 |
+
"step": 361
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 2.5339578454332554,
|
2540 |
+
"grad_norm": 0.05687614157795906,
|
2541 |
+
"learning_rate": 6.732947730037936e-07,
|
2542 |
+
"loss": 0.2019,
|
2543 |
+
"step": 362
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 2.540983606557377,
|
2547 |
+
"grad_norm": 0.05574915185570717,
|
2548 |
+
"learning_rate": 6.52885541229345e-07,
|
2549 |
+
"loss": 0.2085,
|
2550 |
+
"step": 363
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 2.548009367681499,
|
2554 |
+
"grad_norm": 0.05865642428398132,
|
2555 |
+
"learning_rate": 6.327687925956616e-07,
|
2556 |
+
"loss": 0.1975,
|
2557 |
+
"step": 364
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 2.5550351288056206,
|
2561 |
+
"grad_norm": 0.04809865728020668,
|
2562 |
+
"learning_rate": 6.12945880600535e-07,
|
2563 |
+
"loss": 0.1823,
|
2564 |
+
"step": 365
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 2.562060889929742,
|
2568 |
+
"grad_norm": 0.05387086421251297,
|
2569 |
+
"learning_rate": 5.93418138971803e-07,
|
2570 |
+
"loss": 0.2394,
|
2571 |
+
"step": 366
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 2.569086651053864,
|
2575 |
+
"grad_norm": 0.05231115594506264,
|
2576 |
+
"learning_rate": 5.741868815776081e-07,
|
2577 |
+
"loss": 0.2235,
|
2578 |
+
"step": 367
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 2.576112412177986,
|
2582 |
+
"grad_norm": 0.07717155665159225,
|
2583 |
+
"learning_rate": 5.552534023380024e-07,
|
2584 |
+
"loss": 0.2279,
|
2585 |
+
"step": 368
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 2.5831381733021077,
|
2589 |
+
"grad_norm": 0.05653085187077522,
|
2590 |
+
"learning_rate": 5.366189751378858e-07,
|
2591 |
+
"loss": 0.1869,
|
2592 |
+
"step": 369
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 2.5901639344262293,
|
2596 |
+
"grad_norm": 0.05067911744117737,
|
2597 |
+
"learning_rate": 5.18284853741301e-07,
|
2598 |
+
"loss": 0.2376,
|
2599 |
+
"step": 370
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 2.5971896955503513,
|
2603 |
+
"grad_norm": 0.05797469988465309,
|
2604 |
+
"learning_rate": 5.002522717070751e-07,
|
2605 |
+
"loss": 0.1973,
|
2606 |
+
"step": 371
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 2.6042154566744733,
|
2610 |
+
"grad_norm": 0.05374491214752197,
|
2611 |
+
"learning_rate": 4.8252244230582e-07,
|
2612 |
+
"loss": 0.2014,
|
2613 |
+
"step": 372
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 2.611241217798595,
|
2617 |
+
"grad_norm": 0.05649881809949875,
|
2618 |
+
"learning_rate": 4.6509655843830827e-07,
|
2619 |
+
"loss": 0.2027,
|
2620 |
+
"step": 373
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 2.6182669789227164,
|
2624 |
+
"grad_norm": 0.054554861038923264,
|
2625 |
+
"learning_rate": 4.4797579255520585e-07,
|
2626 |
+
"loss": 0.2092,
|
2627 |
+
"step": 374
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 2.6252927400468384,
|
2631 |
+
"grad_norm": 0.054530441761016846,
|
2632 |
+
"learning_rate": 4.311612965781903e-07,
|
2633 |
+
"loss": 0.2196,
|
2634 |
+
"step": 375
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 2.6323185011709604,
|
2638 |
+
"grad_norm": 0.0523928701877594,
|
2639 |
+
"learning_rate": 4.1465420182244476e-07,
|
2640 |
+
"loss": 0.2104,
|
2641 |
+
"step": 376
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 2.639344262295082,
|
2645 |
+
"grad_norm": 0.057155121117830276,
|
2646 |
+
"learning_rate": 3.984556189205441e-07,
|
2647 |
+
"loss": 0.1966,
|
2648 |
+
"step": 377
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 2.6463700234192036,
|
2652 |
+
"grad_norm": 0.05527928099036217,
|
2653 |
+
"learning_rate": 3.8256663774772383e-07,
|
2654 |
+
"loss": 0.235,
|
2655 |
+
"step": 378
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 2.6533957845433256,
|
2659 |
+
"grad_norm": 0.05507628992199898,
|
2660 |
+
"learning_rate": 3.669883273485575e-07,
|
2661 |
+
"loss": 0.2284,
|
2662 |
+
"step": 379
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 2.660421545667447,
|
2666 |
+
"grad_norm": 0.05377492681145668,
|
2667 |
+
"learning_rate": 3.5172173586502543e-07,
|
2668 |
+
"loss": 0.1792,
|
2669 |
+
"step": 380
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 2.667447306791569,
|
2673 |
+
"grad_norm": 0.05631445720791817,
|
2674 |
+
"learning_rate": 3.3676789046599045e-07,
|
2675 |
+
"loss": 0.1963,
|
2676 |
+
"step": 381
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 2.6744730679156907,
|
2680 |
+
"grad_norm": 0.05693857744336128,
|
2681 |
+
"learning_rate": 3.2212779727809504e-07,
|
2682 |
+
"loss": 0.2265,
|
2683 |
+
"step": 382
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 2.6814988290398127,
|
2687 |
+
"grad_norm": 0.052210524678230286,
|
2688 |
+
"learning_rate": 3.0780244131806193e-07,
|
2689 |
+
"loss": 0.2292,
|
2690 |
+
"step": 383
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 2.6885245901639343,
|
2694 |
+
"grad_norm": 0.06421205401420593,
|
2695 |
+
"learning_rate": 2.937927864264206e-07,
|
2696 |
+
"loss": 0.2247,
|
2697 |
+
"step": 384
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 2.6955503512880563,
|
2701 |
+
"grad_norm": 0.054567910730838776,
|
2702 |
+
"learning_rate": 2.800997752026596e-07,
|
2703 |
+
"loss": 0.2115,
|
2704 |
+
"step": 385
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 2.702576112412178,
|
2708 |
+
"grad_norm": 0.05900174379348755,
|
2709 |
+
"learning_rate": 2.667243289418059e-07,
|
2710 |
+
"loss": 0.2014,
|
2711 |
+
"step": 386
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 2.7096018735363,
|
2715 |
+
"grad_norm": 0.05543851852416992,
|
2716 |
+
"learning_rate": 2.5366734757243496e-07,
|
2717 |
+
"loss": 0.1795,
|
2718 |
+
"step": 387
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 2.7166276346604215,
|
2722 |
+
"grad_norm": 0.05802013352513313,
|
2723 |
+
"learning_rate": 2.4092970959612885e-07,
|
2724 |
+
"loss": 0.195,
|
2725 |
+
"step": 388
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 2.7236533957845435,
|
2729 |
+
"grad_norm": 0.04725894704461098,
|
2730 |
+
"learning_rate": 2.2851227202836002e-07,
|
2731 |
+
"loss": 0.2301,
|
2732 |
+
"step": 389
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 2.730679156908665,
|
2736 |
+
"grad_norm": 0.05542716756463051,
|
2737 |
+
"learning_rate": 2.1641587034083756e-07,
|
2738 |
+
"loss": 0.2115,
|
2739 |
+
"step": 390
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 2.737704918032787,
|
2743 |
+
"grad_norm": 0.048591192811727524,
|
2744 |
+
"learning_rate": 2.0464131840528978e-07,
|
2745 |
+
"loss": 0.1922,
|
2746 |
+
"step": 391
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 2.7447306791569086,
|
2750 |
+
"grad_norm": 0.04870427027344704,
|
2751 |
+
"learning_rate": 1.9318940843870594e-07,
|
2752 |
+
"loss": 0.1852,
|
2753 |
+
"step": 392
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 2.7517564402810306,
|
2757 |
+
"grad_norm": 0.05185515433549881,
|
2758 |
+
"learning_rate": 1.8206091095003543e-07,
|
2759 |
+
"loss": 0.2278,
|
2760 |
+
"step": 393
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 2.758782201405152,
|
2764 |
+
"grad_norm": 0.056988395750522614,
|
2765 |
+
"learning_rate": 1.7125657468834656e-07,
|
2766 |
+
"loss": 0.1981,
|
2767 |
+
"step": 394
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 2.7658079625292737,
|
2771 |
+
"grad_norm": 0.050774361938238144,
|
2772 |
+
"learning_rate": 1.6077712659244792e-07,
|
2773 |
+
"loss": 0.2147,
|
2774 |
+
"step": 395
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 2.7728337236533958,
|
2778 |
+
"grad_norm": 0.053753916174173355,
|
2779 |
+
"learning_rate": 1.5062327174197645e-07,
|
2780 |
+
"loss": 0.1922,
|
2781 |
+
"step": 396
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 2.7798594847775178,
|
2785 |
+
"grad_norm": 0.06190885975956917,
|
2786 |
+
"learning_rate": 1.4079569330996412e-07,
|
2787 |
+
"loss": 0.2409,
|
2788 |
+
"step": 397
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 2.7868852459016393,
|
2792 |
+
"grad_norm": 0.05707972124218941,
|
2793 |
+
"learning_rate": 1.3129505251686603e-07,
|
2794 |
+
"loss": 0.1889,
|
2795 |
+
"step": 398
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 2.793911007025761,
|
2799 |
+
"grad_norm": 0.05355476588010788,
|
2800 |
+
"learning_rate": 1.2212198858607694e-07,
|
2801 |
+
"loss": 0.2026,
|
2802 |
+
"step": 399
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 2.800936768149883,
|
2806 |
+
"grad_norm": 0.055988870561122894,
|
2807 |
+
"learning_rate": 1.1327711870091963e-07,
|
2808 |
+
"loss": 0.2114,
|
2809 |
+
"step": 400
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 2.807962529274005,
|
2813 |
+
"grad_norm": 0.056202538311481476,
|
2814 |
+
"learning_rate": 1.0476103796312254e-07,
|
2815 |
+
"loss": 0.2217,
|
2816 |
+
"step": 401
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 2.8149882903981265,
|
2820 |
+
"grad_norm": 0.06274823844432831,
|
2821 |
+
"learning_rate": 9.657431935277629e-08,
|
2822 |
+
"loss": 0.2113,
|
2823 |
+
"step": 402
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 2.822014051522248,
|
2827 |
+
"grad_norm": 0.049836281687021255,
|
2828 |
+
"learning_rate": 8.871751368978554e-08,
|
2829 |
+
"loss": 0.2032,
|
2830 |
+
"step": 403
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 2.82903981264637,
|
2834 |
+
"grad_norm": 0.05938475951552391,
|
2835 |
+
"learning_rate": 8.119114959680929e-08,
|
2836 |
+
"loss": 0.19,
|
2837 |
+
"step": 404
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 2.836065573770492,
|
2841 |
+
"grad_norm": 0.06170409545302391,
|
2842 |
+
"learning_rate": 7.399573346368871e-08,
|
2843 |
+
"loss": 0.2377,
|
2844 |
+
"step": 405
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 2.8430913348946136,
|
2848 |
+
"grad_norm": 0.05599289387464523,
|
2849 |
+
"learning_rate": 6.713174941338163e-08,
|
2850 |
+
"loss": 0.2307,
|
2851 |
+
"step": 406
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 2.850117096018735,
|
2855 |
+
"grad_norm": 0.05716150254011154,
|
2856 |
+
"learning_rate": 6.05996592693886e-08,
|
2857 |
+
"loss": 0.1509,
|
2858 |
+
"step": 407
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 2.857142857142857,
|
2862 |
+
"grad_norm": 0.04617927595973015,
|
2863 |
+
"learning_rate": 5.439990252467886e-08,
|
2864 |
+
"loss": 0.2429,
|
2865 |
+
"step": 408
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 2.8641686182669788,
|
2869 |
+
"grad_norm": 0.05094977468252182,
|
2870 |
+
"learning_rate": 4.853289631212066e-08,
|
2871 |
+
"loss": 0.2302,
|
2872 |
+
"step": 409
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 2.871194379391101,
|
2876 |
+
"grad_norm": 0.05809462442994118,
|
2877 |
+
"learning_rate": 4.299903537641703e-08,
|
2878 |
+
"loss": 0.1985,
|
2879 |
+
"step": 410
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 2.8782201405152223,
|
2883 |
+
"grad_norm": 0.0512741394340992,
|
2884 |
+
"learning_rate": 3.779869204754427e-08,
|
2885 |
+
"loss": 0.2277,
|
2886 |
+
"step": 411
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 2.8852459016393444,
|
2890 |
+
"grad_norm": 0.06057005748152733,
|
2891 |
+
"learning_rate": 3.2932216215704195e-08,
|
2892 |
+
"loss": 0.1957,
|
2893 |
+
"step": 412
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 2.892271662763466,
|
2897 |
+
"grad_norm": 0.05182475969195366,
|
2898 |
+
"learning_rate": 2.8399935307778516e-08,
|
2899 |
+
"loss": 0.1988,
|
2900 |
+
"step": 413
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 2.899297423887588,
|
2904 |
+
"grad_norm": 0.05228479206562042,
|
2905 |
+
"learning_rate": 2.420215426530259e-08,
|
2906 |
+
"loss": 0.1873,
|
2907 |
+
"step": 414
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 2.9063231850117095,
|
2911 |
+
"grad_norm": 0.05834292992949486,
|
2912 |
+
"learning_rate": 2.0339155523945164e-08,
|
2913 |
+
"loss": 0.1845,
|
2914 |
+
"step": 415
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 2.9133489461358315,
|
2918 |
+
"grad_norm": 0.053505126386880875,
|
2919 |
+
"learning_rate": 1.681119899450856e-08,
|
2920 |
+
"loss": 0.208,
|
2921 |
+
"step": 416
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 2.920374707259953,
|
2925 |
+
"grad_norm": 0.05727040395140648,
|
2926 |
+
"learning_rate": 1.3618522045439897e-08,
|
2927 |
+
"loss": 0.2026,
|
2928 |
+
"step": 417
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 2.927400468384075,
|
2932 |
+
"grad_norm": 0.15337811410427094,
|
2933 |
+
"learning_rate": 1.0761339486859424e-08,
|
2934 |
+
"loss": 0.2024,
|
2935 |
+
"step": 418
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 2.9344262295081966,
|
2939 |
+
"grad_norm": 0.05246254801750183,
|
2940 |
+
"learning_rate": 8.239843556108739e-09,
|
2941 |
+
"loss": 0.1698,
|
2942 |
+
"step": 419
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 2.9414519906323187,
|
2946 |
+
"grad_norm": 0.06088126823306084,
|
2947 |
+
"learning_rate": 6.054203904817812e-09,
|
2948 |
+
"loss": 0.2337,
|
2949 |
+
"step": 420
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 2.9484777517564402,
|
2953 |
+
"grad_norm": 0.05842256173491478,
|
2954 |
+
"learning_rate": 4.204567587486885e-09,
|
2955 |
+
"loss": 0.2318,
|
2956 |
+
"step": 421
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 2.9555035128805622,
|
2960 |
+
"grad_norm": 0.05337844043970108,
|
2961 |
+
"learning_rate": 2.6910590515966117e-09,
|
2962 |
+
"loss": 0.1856,
|
2963 |
+
"step": 422
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 2.962529274004684,
|
2967 |
+
"grad_norm": 0.05508222058415413,
|
2968 |
+
"learning_rate": 1.5137801292325338e-09,
|
2969 |
+
"loss": 0.227,
|
2970 |
+
"step": 423
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 2.9695550351288054,
|
2974 |
+
"grad_norm": 0.05756290256977081,
|
2975 |
+
"learning_rate": 6.728100302327844e-10,
|
2976 |
+
"loss": 0.1856,
|
2977 |
+
"step": 424
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 2.9765807962529274,
|
2981 |
+
"grad_norm": 0.05223892256617546,
|
2982 |
+
"learning_rate": 1.6820533686179308e-10,
|
2983 |
+
"loss": 0.2299,
|
2984 |
+
"step": 425
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 2.9836065573770494,
|
2988 |
+
"grad_norm": 0.0553586483001709,
|
2989 |
+
"learning_rate": 0.0,
|
2990 |
+
"loss": 0.1971,
|
2991 |
+
"step": 426
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 2.9836065573770494,
|
2995 |
+
"step": 426,
|
2996 |
+
"total_flos": 1.0638855097443942e+17,
|
2997 |
+
"train_loss": 0.35510572714145194,
|
2998 |
+
"train_runtime": 58523.3937,
|
2999 |
+
"train_samples_per_second": 0.175,
|
3000 |
+
"train_steps_per_second": 0.007
|
3001 |
+
}
|
3002 |
+
],
|
3003 |
+
"logging_steps": 1,
|
3004 |
+
"max_steps": 426,
|
3005 |
+
"num_input_tokens_seen": 0,
|
3006 |
+
"num_train_epochs": 3,
|
3007 |
+
"save_steps": 200,
|
3008 |
+
"stateful_callbacks": {
|
3009 |
+
"TrainerControl": {
|
3010 |
+
"args": {
|
3011 |
+
"should_epoch_stop": false,
|
3012 |
+
"should_evaluate": false,
|
3013 |
+
"should_log": false,
|
3014 |
+
"should_save": true,
|
3015 |
+
"should_training_stop": true
|
3016 |
+
},
|
3017 |
+
"attributes": {}
|
3018 |
+
}
|
3019 |
+
},
|
3020 |
+
"total_flos": 1.0638855097443942e+17,
|
3021 |
+
"train_batch_size": 1,
|
3022 |
+
"trial_name": null,
|
3023 |
+
"trial_params": null
|
3024 |
+
}
|
training_loss.png
ADDED
![]() |