End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2863 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: 1k_globalbatchsize32_lr5e6_epochs13
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# 1k_globalbatchsize32_lr5e6_epochs13
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: 1k_globalbatchsize32_lr5e6_epochs13
|
|
|
|
| 16 |
|
| 17 |
# 1k_globalbatchsize32_lr5e6_epochs13
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/openthoughts_1000 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.896,
|
| 3 |
+
"total_flos": 2.2087303787769037e+17,
|
| 4 |
+
"train_loss": 0.3277998325731559,
|
| 5 |
+
"train_runtime": 13876.437,
|
| 6 |
+
"train_samples_per_second": 0.933,
|
| 7 |
+
"train_steps_per_second": 0.029
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.896,
|
| 3 |
+
"total_flos": 2.2087303787769037e+17,
|
| 4 |
+
"train_loss": 0.3277998325731559,
|
| 5 |
+
"train_runtime": 13876.437,
|
| 6 |
+
"train_samples_per_second": 0.933,
|
| 7 |
+
"train_steps_per_second": 0.029
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2863 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 12.896,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 403,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.032,
|
| 13 |
+
"grad_norm": 5.849556277204242,
|
| 14 |
+
"learning_rate": 1.2195121951219514e-07,
|
| 15 |
+
"loss": 0.8186,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.064,
|
| 20 |
+
"grad_norm": 5.897026099693046,
|
| 21 |
+
"learning_rate": 2.439024390243903e-07,
|
| 22 |
+
"loss": 0.8458,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.096,
|
| 27 |
+
"grad_norm": 5.722749931039787,
|
| 28 |
+
"learning_rate": 3.6585365853658536e-07,
|
| 29 |
+
"loss": 0.8311,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.128,
|
| 34 |
+
"grad_norm": 5.946871790898374,
|
| 35 |
+
"learning_rate": 4.878048780487805e-07,
|
| 36 |
+
"loss": 0.8718,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16,
|
| 41 |
+
"grad_norm": 6.128494806604928,
|
| 42 |
+
"learning_rate": 6.097560975609757e-07,
|
| 43 |
+
"loss": 0.8189,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.192,
|
| 48 |
+
"grad_norm": 6.113984717476297,
|
| 49 |
+
"learning_rate": 7.317073170731707e-07,
|
| 50 |
+
"loss": 0.9245,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.224,
|
| 55 |
+
"grad_norm": 5.942358219370905,
|
| 56 |
+
"learning_rate": 8.53658536585366e-07,
|
| 57 |
+
"loss": 0.8893,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.256,
|
| 62 |
+
"grad_norm": 5.1631628044313445,
|
| 63 |
+
"learning_rate": 9.75609756097561e-07,
|
| 64 |
+
"loss": 0.8441,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.288,
|
| 69 |
+
"grad_norm": 5.459078869746953,
|
| 70 |
+
"learning_rate": 1.0975609756097562e-06,
|
| 71 |
+
"loss": 0.8262,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.32,
|
| 76 |
+
"grad_norm": 4.612510949957037,
|
| 77 |
+
"learning_rate": 1.2195121951219514e-06,
|
| 78 |
+
"loss": 0.8212,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.352,
|
| 83 |
+
"grad_norm": 4.373746027707626,
|
| 84 |
+
"learning_rate": 1.3414634146341465e-06,
|
| 85 |
+
"loss": 0.8676,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.384,
|
| 90 |
+
"grad_norm": 4.306516033326923,
|
| 91 |
+
"learning_rate": 1.4634146341463414e-06,
|
| 92 |
+
"loss": 0.8373,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.416,
|
| 97 |
+
"grad_norm": 2.785144681347322,
|
| 98 |
+
"learning_rate": 1.5853658536585368e-06,
|
| 99 |
+
"loss": 0.7595,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.448,
|
| 104 |
+
"grad_norm": 2.6350200531690215,
|
| 105 |
+
"learning_rate": 1.707317073170732e-06,
|
| 106 |
+
"loss": 0.8025,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.48,
|
| 111 |
+
"grad_norm": 2.6075651371895483,
|
| 112 |
+
"learning_rate": 1.8292682926829268e-06,
|
| 113 |
+
"loss": 0.766,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.512,
|
| 118 |
+
"grad_norm": 2.0659475356567647,
|
| 119 |
+
"learning_rate": 1.951219512195122e-06,
|
| 120 |
+
"loss": 0.7789,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.544,
|
| 125 |
+
"grad_norm": 2.018183976562194,
|
| 126 |
+
"learning_rate": 2.073170731707317e-06,
|
| 127 |
+
"loss": 0.7775,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.576,
|
| 132 |
+
"grad_norm": 2.4992713997752665,
|
| 133 |
+
"learning_rate": 2.1951219512195125e-06,
|
| 134 |
+
"loss": 0.7508,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.608,
|
| 139 |
+
"grad_norm": 3.4160867498236844,
|
| 140 |
+
"learning_rate": 2.317073170731708e-06,
|
| 141 |
+
"loss": 0.7831,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.64,
|
| 146 |
+
"grad_norm": 3.3995778959217184,
|
| 147 |
+
"learning_rate": 2.4390243902439027e-06,
|
| 148 |
+
"loss": 0.7051,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.672,
|
| 153 |
+
"grad_norm": 3.356913620677719,
|
| 154 |
+
"learning_rate": 2.5609756097560977e-06,
|
| 155 |
+
"loss": 0.735,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.704,
|
| 160 |
+
"grad_norm": 3.4485792373305495,
|
| 161 |
+
"learning_rate": 2.682926829268293e-06,
|
| 162 |
+
"loss": 0.7504,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.736,
|
| 167 |
+
"grad_norm": 2.775219211115156,
|
| 168 |
+
"learning_rate": 2.8048780487804884e-06,
|
| 169 |
+
"loss": 0.6851,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.768,
|
| 174 |
+
"grad_norm": 2.9699753753960527,
|
| 175 |
+
"learning_rate": 2.926829268292683e-06,
|
| 176 |
+
"loss": 0.791,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8,
|
| 181 |
+
"grad_norm": 1.9046018244729659,
|
| 182 |
+
"learning_rate": 3.0487804878048782e-06,
|
| 183 |
+
"loss": 0.7116,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.832,
|
| 188 |
+
"grad_norm": 1.7584248199754853,
|
| 189 |
+
"learning_rate": 3.1707317073170736e-06,
|
| 190 |
+
"loss": 0.7682,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.864,
|
| 195 |
+
"grad_norm": 1.4870843121046768,
|
| 196 |
+
"learning_rate": 3.292682926829269e-06,
|
| 197 |
+
"loss": 0.6955,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.896,
|
| 202 |
+
"grad_norm": 1.5791059474876092,
|
| 203 |
+
"learning_rate": 3.414634146341464e-06,
|
| 204 |
+
"loss": 0.6821,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.928,
|
| 209 |
+
"grad_norm": 1.5015383013859356,
|
| 210 |
+
"learning_rate": 3.5365853658536588e-06,
|
| 211 |
+
"loss": 0.701,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.96,
|
| 216 |
+
"grad_norm": 1.5353940501247298,
|
| 217 |
+
"learning_rate": 3.6585365853658537e-06,
|
| 218 |
+
"loss": 0.7222,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.992,
|
| 223 |
+
"grad_norm": 1.4677052617262567,
|
| 224 |
+
"learning_rate": 3.780487804878049e-06,
|
| 225 |
+
"loss": 0.6838,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.024,
|
| 230 |
+
"grad_norm": 2.1766405956777755,
|
| 231 |
+
"learning_rate": 3.902439024390244e-06,
|
| 232 |
+
"loss": 1.0542,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.056,
|
| 237 |
+
"grad_norm": 1.1928687462071357,
|
| 238 |
+
"learning_rate": 4.024390243902439e-06,
|
| 239 |
+
"loss": 0.6235,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.088,
|
| 244 |
+
"grad_norm": 1.0286099846913035,
|
| 245 |
+
"learning_rate": 4.146341463414634e-06,
|
| 246 |
+
"loss": 0.6258,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.12,
|
| 251 |
+
"grad_norm": 1.0747784822991218,
|
| 252 |
+
"learning_rate": 4.268292682926829e-06,
|
| 253 |
+
"loss": 0.6702,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.152,
|
| 258 |
+
"grad_norm": 0.962470776270371,
|
| 259 |
+
"learning_rate": 4.390243902439025e-06,
|
| 260 |
+
"loss": 0.5741,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.184,
|
| 265 |
+
"grad_norm": 1.2656865970358744,
|
| 266 |
+
"learning_rate": 4.51219512195122e-06,
|
| 267 |
+
"loss": 0.691,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.216,
|
| 272 |
+
"grad_norm": 1.2632482244976766,
|
| 273 |
+
"learning_rate": 4.634146341463416e-06,
|
| 274 |
+
"loss": 0.6774,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.248,
|
| 279 |
+
"grad_norm": 1.1753881829414252,
|
| 280 |
+
"learning_rate": 4.75609756097561e-06,
|
| 281 |
+
"loss": 0.5534,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.28,
|
| 286 |
+
"grad_norm": 0.8508889194005277,
|
| 287 |
+
"learning_rate": 4.8780487804878055e-06,
|
| 288 |
+
"loss": 0.5971,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.312,
|
| 293 |
+
"grad_norm": 0.802644049353535,
|
| 294 |
+
"learning_rate": 5e-06,
|
| 295 |
+
"loss": 0.5202,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3439999999999999,
|
| 300 |
+
"grad_norm": 0.9800768008764194,
|
| 301 |
+
"learning_rate": 4.999905856597241e-06,
|
| 302 |
+
"loss": 0.6705,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.376,
|
| 307 |
+
"grad_norm": 0.803110222182918,
|
| 308 |
+
"learning_rate": 4.999623433479346e-06,
|
| 309 |
+
"loss": 0.5315,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.408,
|
| 314 |
+
"grad_norm": 0.9477706564190158,
|
| 315 |
+
"learning_rate": 4.999152751916936e-06,
|
| 316 |
+
"loss": 0.6445,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.44,
|
| 321 |
+
"grad_norm": 1.0284577855915737,
|
| 322 |
+
"learning_rate": 4.99849384735926e-06,
|
| 323 |
+
"loss": 0.6769,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.472,
|
| 328 |
+
"grad_norm": 0.7970856486838346,
|
| 329 |
+
"learning_rate": 4.997646769431532e-06,
|
| 330 |
+
"loss": 0.5371,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.504,
|
| 335 |
+
"grad_norm": 0.8094391056732192,
|
| 336 |
+
"learning_rate": 4.9966115819311926e-06,
|
| 337 |
+
"loss": 0.6055,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.536,
|
| 342 |
+
"grad_norm": 0.8257228282004593,
|
| 343 |
+
"learning_rate": 4.9953883628231e-06,
|
| 344 |
+
"loss": 0.6152,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.568,
|
| 349 |
+
"grad_norm": 0.9519135973480538,
|
| 350 |
+
"learning_rate": 4.99397720423366e-06,
|
| 351 |
+
"loss": 0.6453,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.6,
|
| 356 |
+
"grad_norm": 0.8032344909768069,
|
| 357 |
+
"learning_rate": 4.992378212443891e-06,
|
| 358 |
+
"loss": 0.5899,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.6320000000000001,
|
| 363 |
+
"grad_norm": 0.7928432845271344,
|
| 364 |
+
"learning_rate": 4.990591507881416e-06,
|
| 365 |
+
"loss": 0.571,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6640000000000001,
|
| 370 |
+
"grad_norm": 0.872945757213797,
|
| 371 |
+
"learning_rate": 4.988617225111392e-06,
|
| 372 |
+
"loss": 0.6221,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.696,
|
| 377 |
+
"grad_norm": 0.9085065700458688,
|
| 378 |
+
"learning_rate": 4.986455512826377e-06,
|
| 379 |
+
"loss": 0.609,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.728,
|
| 384 |
+
"grad_norm": 0.7651931804678502,
|
| 385 |
+
"learning_rate": 4.984106533835132e-06,
|
| 386 |
+
"loss": 0.5305,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.76,
|
| 391 |
+
"grad_norm": 0.8227356576229764,
|
| 392 |
+
"learning_rate": 4.981570465050357e-06,
|
| 393 |
+
"loss": 0.5493,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.792,
|
| 398 |
+
"grad_norm": 0.822727070066304,
|
| 399 |
+
"learning_rate": 4.978847497475369e-06,
|
| 400 |
+
"loss": 0.6164,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.8239999999999998,
|
| 405 |
+
"grad_norm": 0.7160048714802999,
|
| 406 |
+
"learning_rate": 4.975937836189712e-06,
|
| 407 |
+
"loss": 0.5401,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8559999999999999,
|
| 412 |
+
"grad_norm": 0.763028118096705,
|
| 413 |
+
"learning_rate": 4.97284170033372e-06,
|
| 414 |
+
"loss": 0.5871,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.888,
|
| 419 |
+
"grad_norm": 0.6832461165931352,
|
| 420 |
+
"learning_rate": 4.969559323092004e-06,
|
| 421 |
+
"loss": 0.5104,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.92,
|
| 426 |
+
"grad_norm": 0.7468945242633849,
|
| 427 |
+
"learning_rate": 4.966090951675893e-06,
|
| 428 |
+
"loss": 0.5707,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.952,
|
| 433 |
+
"grad_norm": 0.8806093646704142,
|
| 434 |
+
"learning_rate": 4.962436847304818e-06,
|
| 435 |
+
"loss": 0.6179,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.984,
|
| 440 |
+
"grad_norm": 0.7207029671285448,
|
| 441 |
+
"learning_rate": 4.958597285186635e-06,
|
| 442 |
+
"loss": 0.5897,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.016,
|
| 447 |
+
"grad_norm": 1.4244038383210325,
|
| 448 |
+
"learning_rate": 4.954572554496897e-06,
|
| 449 |
+
"loss": 0.855,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.048,
|
| 454 |
+
"grad_norm": 0.773835823048524,
|
| 455 |
+
"learning_rate": 4.950362958357078e-06,
|
| 456 |
+
"loss": 0.5093,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.08,
|
| 461 |
+
"grad_norm": 0.7863551578984473,
|
| 462 |
+
"learning_rate": 4.945968813811743e-06,
|
| 463 |
+
"loss": 0.5478,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.112,
|
| 468 |
+
"grad_norm": 0.6640818760415559,
|
| 469 |
+
"learning_rate": 4.941390451804668e-06,
|
| 470 |
+
"loss": 0.4982,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.144,
|
| 475 |
+
"grad_norm": 0.7850080843731436,
|
| 476 |
+
"learning_rate": 4.936628217153914e-06,
|
| 477 |
+
"loss": 0.5422,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.176,
|
| 482 |
+
"grad_norm": 0.5703622243364901,
|
| 483 |
+
"learning_rate": 4.931682468525863e-06,
|
| 484 |
+
"loss": 0.413,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.208,
|
| 489 |
+
"grad_norm": 0.7111763921998362,
|
| 490 |
+
"learning_rate": 4.9265535784081965e-06,
|
| 491 |
+
"loss": 0.502,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.24,
|
| 496 |
+
"grad_norm": 0.7477270270033164,
|
| 497 |
+
"learning_rate": 4.92124193308185e-06,
|
| 498 |
+
"loss": 0.5625,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.2720000000000002,
|
| 503 |
+
"grad_norm": 0.709007804007645,
|
| 504 |
+
"learning_rate": 4.915747932591916e-06,
|
| 505 |
+
"loss": 0.505,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.304,
|
| 510 |
+
"grad_norm": 0.6774734628700982,
|
| 511 |
+
"learning_rate": 4.910071990717516e-06,
|
| 512 |
+
"loss": 0.5561,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.336,
|
| 517 |
+
"grad_norm": 0.6580560490507651,
|
| 518 |
+
"learning_rate": 4.9042145349406335e-06,
|
| 519 |
+
"loss": 0.4677,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.368,
|
| 524 |
+
"grad_norm": 0.6346401612374466,
|
| 525 |
+
"learning_rate": 4.898176006413925e-06,
|
| 526 |
+
"loss": 0.5179,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.4,
|
| 531 |
+
"grad_norm": 0.6080248615332154,
|
| 532 |
+
"learning_rate": 4.891956859927489e-06,
|
| 533 |
+
"loss": 0.434,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.432,
|
| 538 |
+
"grad_norm": 0.8389178422634704,
|
| 539 |
+
"learning_rate": 4.885557563874614e-06,
|
| 540 |
+
"loss": 0.6016,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.464,
|
| 545 |
+
"grad_norm": 0.615230334543301,
|
| 546 |
+
"learning_rate": 4.8789786002165055e-06,
|
| 547 |
+
"loss": 0.4484,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.496,
|
| 552 |
+
"grad_norm": 0.5863460239871948,
|
| 553 |
+
"learning_rate": 4.872220464445983e-06,
|
| 554 |
+
"loss": 0.4766,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.528,
|
| 559 |
+
"grad_norm": 0.6361411281245383,
|
| 560 |
+
"learning_rate": 4.865283665550167e-06,
|
| 561 |
+
"loss": 0.4833,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.56,
|
| 566 |
+
"grad_norm": 0.7099406086449909,
|
| 567 |
+
"learning_rate": 4.8581687259721375e-06,
|
| 568 |
+
"loss": 0.5876,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.592,
|
| 573 |
+
"grad_norm": 0.7220569534344881,
|
| 574 |
+
"learning_rate": 4.850876181571592e-06,
|
| 575 |
+
"loss": 0.5517,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.624,
|
| 580 |
+
"grad_norm": 0.6183366502306967,
|
| 581 |
+
"learning_rate": 4.843406581584487e-06,
|
| 582 |
+
"loss": 0.4711,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.656,
|
| 587 |
+
"grad_norm": 0.6825890552051384,
|
| 588 |
+
"learning_rate": 4.83576048858167e-06,
|
| 589 |
+
"loss": 0.4951,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.6879999999999997,
|
| 594 |
+
"grad_norm": 0.7434791701964563,
|
| 595 |
+
"learning_rate": 4.8279384784265124e-06,
|
| 596 |
+
"loss": 0.552,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.7199999999999998,
|
| 601 |
+
"grad_norm": 0.7081425720692629,
|
| 602 |
+
"learning_rate": 4.8199411402315356e-06,
|
| 603 |
+
"loss": 0.5769,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.752,
|
| 608 |
+
"grad_norm": 0.5087385999506354,
|
| 609 |
+
"learning_rate": 4.811769076314044e-06,
|
| 610 |
+
"loss": 0.3774,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.784,
|
| 615 |
+
"grad_norm": 0.6776053182453752,
|
| 616 |
+
"learning_rate": 4.803422902150762e-06,
|
| 617 |
+
"loss": 0.5108,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.816,
|
| 622 |
+
"grad_norm": 0.6706814223980743,
|
| 623 |
+
"learning_rate": 4.794903246331477e-06,
|
| 624 |
+
"loss": 0.4945,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.848,
|
| 629 |
+
"grad_norm": 0.5764102991256048,
|
| 630 |
+
"learning_rate": 4.786210750511701e-06,
|
| 631 |
+
"loss": 0.4616,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.88,
|
| 636 |
+
"grad_norm": 0.705063277061744,
|
| 637 |
+
"learning_rate": 4.777346069364343e-06,
|
| 638 |
+
"loss": 0.6117,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.912,
|
| 643 |
+
"grad_norm": 0.5603983430629818,
|
| 644 |
+
"learning_rate": 4.7683098705303995e-06,
|
| 645 |
+
"loss": 0.4452,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.944,
|
| 650 |
+
"grad_norm": 0.6965726864559006,
|
| 651 |
+
"learning_rate": 4.7591028345686765e-06,
|
| 652 |
+
"loss": 0.5453,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.976,
|
| 657 |
+
"grad_norm": 0.6264744127555706,
|
| 658 |
+
"learning_rate": 4.749725654904529e-06,
|
| 659 |
+
"loss": 0.4177,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.008,
|
| 664 |
+
"grad_norm": 1.279958151295519,
|
| 665 |
+
"learning_rate": 4.740179037777639e-06,
|
| 666 |
+
"loss": 0.8889,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.04,
|
| 671 |
+
"grad_norm": 0.673611932133756,
|
| 672 |
+
"learning_rate": 4.730463702188824e-06,
|
| 673 |
+
"loss": 0.509,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.072,
|
| 678 |
+
"grad_norm": 0.6258573615607128,
|
| 679 |
+
"learning_rate": 4.720580379845884e-06,
|
| 680 |
+
"loss": 0.4415,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.104,
|
| 685 |
+
"grad_norm": 0.6666909339994116,
|
| 686 |
+
"learning_rate": 4.710529815108496e-06,
|
| 687 |
+
"loss": 0.3991,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.136,
|
| 692 |
+
"grad_norm": 0.6905162988704795,
|
| 693 |
+
"learning_rate": 4.700312764932151e-06,
|
| 694 |
+
"loss": 0.4603,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.168,
|
| 699 |
+
"grad_norm": 0.713118856359383,
|
| 700 |
+
"learning_rate": 4.689929998811145e-06,
|
| 701 |
+
"loss": 0.5021,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.2,
|
| 706 |
+
"grad_norm": 0.5985502187319915,
|
| 707 |
+
"learning_rate": 4.679382298720625e-06,
|
| 708 |
+
"loss": 0.375,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.232,
|
| 713 |
+
"grad_norm": 0.7727805896468829,
|
| 714 |
+
"learning_rate": 4.668670459057693e-06,
|
| 715 |
+
"loss": 0.5022,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.2640000000000002,
|
| 720 |
+
"grad_norm": 0.5767775270382344,
|
| 721 |
+
"learning_rate": 4.657795286581576e-06,
|
| 722 |
+
"loss": 0.372,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.296,
|
| 727 |
+
"grad_norm": 0.6858125911692157,
|
| 728 |
+
"learning_rate": 4.64675760035287e-06,
|
| 729 |
+
"loss": 0.454,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.328,
|
| 734 |
+
"grad_norm": 0.6451556869300004,
|
| 735 |
+
"learning_rate": 4.635558231671846e-06,
|
| 736 |
+
"loss": 0.513,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.36,
|
| 741 |
+
"grad_norm": 0.47121399780666584,
|
| 742 |
+
"learning_rate": 4.624198024015845e-06,
|
| 743 |
+
"loss": 0.3442,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.392,
|
| 748 |
+
"grad_norm": 0.7749014985719241,
|
| 749 |
+
"learning_rate": 4.612677832975751e-06,
|
| 750 |
+
"loss": 0.5,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.424,
|
| 755 |
+
"grad_norm": 0.6346536197956437,
|
| 756 |
+
"learning_rate": 4.6009985261915536e-06,
|
| 757 |
+
"loss": 0.4228,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.456,
|
| 762 |
+
"grad_norm": 0.6577199203857297,
|
| 763 |
+
"learning_rate": 4.5891609832869964e-06,
|
| 764 |
+
"loss": 0.4441,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.488,
|
| 769 |
+
"grad_norm": 0.6797699170201807,
|
| 770 |
+
"learning_rate": 4.577166095803336e-06,
|
| 771 |
+
"loss": 0.434,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.52,
|
| 776 |
+
"grad_norm": 0.6652749132431918,
|
| 777 |
+
"learning_rate": 4.565014767132191e-06,
|
| 778 |
+
"loss": 0.4128,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.552,
|
| 783 |
+
"grad_norm": 0.7280504438280285,
|
| 784 |
+
"learning_rate": 4.552707912447504e-06,
|
| 785 |
+
"loss": 0.4117,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.584,
|
| 790 |
+
"grad_norm": 0.6901309155371619,
|
| 791 |
+
"learning_rate": 4.540246458636619e-06,
|
| 792 |
+
"loss": 0.4972,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.616,
|
| 797 |
+
"grad_norm": 0.6539420051609891,
|
| 798 |
+
"learning_rate": 4.527631344230466e-06,
|
| 799 |
+
"loss": 0.4346,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.648,
|
| 804 |
+
"grad_norm": 0.7276764867026899,
|
| 805 |
+
"learning_rate": 4.514863519332882e-06,
|
| 806 |
+
"loss": 0.4698,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.68,
|
| 811 |
+
"grad_norm": 0.6175907671780341,
|
| 812 |
+
"learning_rate": 4.501943945549054e-06,
|
| 813 |
+
"loss": 0.4136,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.7119999999999997,
|
| 818 |
+
"grad_norm": 0.6023557542009865,
|
| 819 |
+
"learning_rate": 4.488873595913092e-06,
|
| 820 |
+
"loss": 0.4511,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.7439999999999998,
|
| 825 |
+
"grad_norm": 0.6301322985613332,
|
| 826 |
+
"learning_rate": 4.475653454814746e-06,
|
| 827 |
+
"loss": 0.441,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.776,
|
| 832 |
+
"grad_norm": 0.6340733352505695,
|
| 833 |
+
"learning_rate": 4.4622845179252735e-06,
|
| 834 |
+
"loss": 0.4027,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.808,
|
| 839 |
+
"grad_norm": 0.6432287187869242,
|
| 840 |
+
"learning_rate": 4.44876779212244e-06,
|
| 841 |
+
"loss": 0.3986,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.84,
|
| 846 |
+
"grad_norm": 0.6594993068804657,
|
| 847 |
+
"learning_rate": 4.435104295414697e-06,
|
| 848 |
+
"loss": 0.452,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.872,
|
| 853 |
+
"grad_norm": 0.6445718630882695,
|
| 854 |
+
"learning_rate": 4.421295056864501e-06,
|
| 855 |
+
"loss": 0.4504,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.904,
|
| 860 |
+
"grad_norm": 0.6238995635064682,
|
| 861 |
+
"learning_rate": 4.407341116510818e-06,
|
| 862 |
+
"loss": 0.4558,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.936,
|
| 867 |
+
"grad_norm": 0.5359484609283722,
|
| 868 |
+
"learning_rate": 4.3932435252907914e-06,
|
| 869 |
+
"loss": 0.3728,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.968,
|
| 874 |
+
"grad_norm": 0.7283748341793856,
|
| 875 |
+
"learning_rate": 4.379003344960585e-06,
|
| 876 |
+
"loss": 0.5181,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.0,
|
| 881 |
+
"grad_norm": 1.237058634537511,
|
| 882 |
+
"learning_rate": 4.364621648015426e-06,
|
| 883 |
+
"loss": 0.6485,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.032,
|
| 888 |
+
"grad_norm": 0.7189112417262469,
|
| 889 |
+
"learning_rate": 4.3500995176088235e-06,
|
| 890 |
+
"loss": 0.3787,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.064,
|
| 895 |
+
"grad_norm": 0.6376959092529467,
|
| 896 |
+
"learning_rate": 4.335438047470996e-06,
|
| 897 |
+
"loss": 0.3928,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.096,
|
| 902 |
+
"grad_norm": 0.6926295760270299,
|
| 903 |
+
"learning_rate": 4.320638341826494e-06,
|
| 904 |
+
"loss": 0.3467,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.128,
|
| 909 |
+
"grad_norm": 0.7759472194751662,
|
| 910 |
+
"learning_rate": 4.305701515311037e-06,
|
| 911 |
+
"loss": 0.3417,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.16,
|
| 916 |
+
"grad_norm": 0.7903567022198654,
|
| 917 |
+
"learning_rate": 4.290628692887564e-06,
|
| 918 |
+
"loss": 0.3988,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.192,
|
| 923 |
+
"grad_norm": 0.6775976478128614,
|
| 924 |
+
"learning_rate": 4.27542100976151e-06,
|
| 925 |
+
"loss": 0.3882,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.224,
|
| 930 |
+
"grad_norm": 0.6978160974410718,
|
| 931 |
+
"learning_rate": 4.260079611295303e-06,
|
| 932 |
+
"loss": 0.3785,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.256,
|
| 937 |
+
"grad_norm": 0.6543473089069344,
|
| 938 |
+
"learning_rate": 4.244605652922108e-06,
|
| 939 |
+
"loss": 0.4232,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.288,
|
| 944 |
+
"grad_norm": 0.567510016178576,
|
| 945 |
+
"learning_rate": 4.229000300058802e-06,
|
| 946 |
+
"loss": 0.3328,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.32,
|
| 951 |
+
"grad_norm": 0.7026260353156972,
|
| 952 |
+
"learning_rate": 4.2132647280182e-06,
|
| 953 |
+
"loss": 0.3689,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.352,
|
| 958 |
+
"grad_norm": 0.7595463041990811,
|
| 959 |
+
"learning_rate": 4.197400121920539e-06,
|
| 960 |
+
"loss": 0.4105,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.384,
|
| 965 |
+
"grad_norm": 0.6784378790316485,
|
| 966 |
+
"learning_rate": 4.1814076766042206e-06,
|
| 967 |
+
"loss": 0.3584,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.416,
|
| 972 |
+
"grad_norm": 0.7040399997283757,
|
| 973 |
+
"learning_rate": 4.165288596535821e-06,
|
| 974 |
+
"loss": 0.3428,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.448,
|
| 979 |
+
"grad_norm": 0.6811902663377835,
|
| 980 |
+
"learning_rate": 4.149044095719377e-06,
|
| 981 |
+
"loss": 0.3268,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.48,
|
| 986 |
+
"grad_norm": 0.6292342873011069,
|
| 987 |
+
"learning_rate": 4.132675397604956e-06,
|
| 988 |
+
"loss": 0.4163,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.5120000000000005,
|
| 993 |
+
"grad_norm": 0.5986205237032173,
|
| 994 |
+
"learning_rate": 4.11618373499651e-06,
|
| 995 |
+
"loss": 0.3724,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.5440000000000005,
|
| 1000 |
+
"grad_norm": 0.7366511370121315,
|
| 1001 |
+
"learning_rate": 4.099570349959025e-06,
|
| 1002 |
+
"loss": 0.3471,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.576,
|
| 1007 |
+
"grad_norm": 0.624561699567186,
|
| 1008 |
+
"learning_rate": 4.082836493724981e-06,
|
| 1009 |
+
"loss": 0.3576,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.608,
|
| 1014 |
+
"grad_norm": 0.6571595063682137,
|
| 1015 |
+
"learning_rate": 4.065983426600113e-06,
|
| 1016 |
+
"loss": 0.3186,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.64,
|
| 1021 |
+
"grad_norm": 0.6579294724596673,
|
| 1022 |
+
"learning_rate": 4.0490124178684884e-06,
|
| 1023 |
+
"loss": 0.3618,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.672,
|
| 1028 |
+
"grad_norm": 0.7226442228936304,
|
| 1029 |
+
"learning_rate": 4.031924745696916e-06,
|
| 1030 |
+
"loss": 0.3731,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.704,
|
| 1035 |
+
"grad_norm": 0.6526801907173936,
|
| 1036 |
+
"learning_rate": 4.014721697038678e-06,
|
| 1037 |
+
"loss": 0.4246,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.736,
|
| 1042 |
+
"grad_norm": 0.6159750078996794,
|
| 1043 |
+
"learning_rate": 3.997404567536606e-06,
|
| 1044 |
+
"loss": 0.3988,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.768,
|
| 1049 |
+
"grad_norm": 0.6284609689319339,
|
| 1050 |
+
"learning_rate": 3.979974661425497e-06,
|
| 1051 |
+
"loss": 0.407,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.8,
|
| 1056 |
+
"grad_norm": 0.6632101552174875,
|
| 1057 |
+
"learning_rate": 3.962433291433889e-06,
|
| 1058 |
+
"loss": 0.4004,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.832,
|
| 1063 |
+
"grad_norm": 0.6515542500448573,
|
| 1064 |
+
"learning_rate": 3.944781778685189e-06,
|
| 1065 |
+
"loss": 0.3787,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.864,
|
| 1070 |
+
"grad_norm": 0.6971261871159161,
|
| 1071 |
+
"learning_rate": 3.927021452598177e-06,
|
| 1072 |
+
"loss": 0.355,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.896,
|
| 1077 |
+
"grad_norm": 0.5783267304385598,
|
| 1078 |
+
"learning_rate": 3.909153650786878e-06,
|
| 1079 |
+
"loss": 0.3647,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.928,
|
| 1084 |
+
"grad_norm": 0.6547335054539042,
|
| 1085 |
+
"learning_rate": 3.891179718959822e-06,
|
| 1086 |
+
"loss": 0.3614,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.96,
|
| 1091 |
+
"grad_norm": 0.6422534350091944,
|
| 1092 |
+
"learning_rate": 3.873101010818692e-06,
|
| 1093 |
+
"loss": 0.3927,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.992,
|
| 1098 |
+
"grad_norm": 0.7002809249283289,
|
| 1099 |
+
"learning_rate": 3.8549188879563685e-06,
|
| 1100 |
+
"loss": 0.3952,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.024,
|
| 1105 |
+
"grad_norm": 1.541829058821354,
|
| 1106 |
+
"learning_rate": 3.836634719754385e-06,
|
| 1107 |
+
"loss": 0.456,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.056,
|
| 1112 |
+
"grad_norm": 0.8605198975053161,
|
| 1113 |
+
"learning_rate": 3.818249883279791e-06,
|
| 1114 |
+
"loss": 0.3757,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.088,
|
| 1119 |
+
"grad_norm": 0.7016324550366518,
|
| 1120 |
+
"learning_rate": 3.7997657631814366e-06,
|
| 1121 |
+
"loss": 0.3035,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.12,
|
| 1126 |
+
"grad_norm": 1.328738315381159,
|
| 1127 |
+
"learning_rate": 3.781183751585693e-06,
|
| 1128 |
+
"loss": 0.3321,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.152,
|
| 1133 |
+
"grad_norm": 0.9946678564822509,
|
| 1134 |
+
"learning_rate": 3.762505247991601e-06,
|
| 1135 |
+
"loss": 0.3305,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.184,
|
| 1140 |
+
"grad_norm": 0.666249844083357,
|
| 1141 |
+
"learning_rate": 3.7437316591654726e-06,
|
| 1142 |
+
"loss": 0.2945,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.216,
|
| 1147 |
+
"grad_norm": 0.7788017018090492,
|
| 1148 |
+
"learning_rate": 3.724864399034932e-06,
|
| 1149 |
+
"loss": 0.3239,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.248,
|
| 1154 |
+
"grad_norm": 0.8140243452638297,
|
| 1155 |
+
"learning_rate": 3.7059048885824367e-06,
|
| 1156 |
+
"loss": 0.3396,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.28,
|
| 1161 |
+
"grad_norm": 0.6712905040344191,
|
| 1162 |
+
"learning_rate": 3.686854555738249e-06,
|
| 1163 |
+
"loss": 0.3093,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.312,
|
| 1168 |
+
"grad_norm": 0.6078748033907079,
|
| 1169 |
+
"learning_rate": 3.6677148352728947e-06,
|
| 1170 |
+
"loss": 0.2804,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.344,
|
| 1175 |
+
"grad_norm": 0.7165018792349958,
|
| 1176 |
+
"learning_rate": 3.6484871686891044e-06,
|
| 1177 |
+
"loss": 0.3633,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.376,
|
| 1182 |
+
"grad_norm": 0.8417077431991168,
|
| 1183 |
+
"learning_rate": 3.629173004113245e-06,
|
| 1184 |
+
"loss": 0.3713,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.408,
|
| 1189 |
+
"grad_norm": 0.8058193310329876,
|
| 1190 |
+
"learning_rate": 3.609773796186256e-06,
|
| 1191 |
+
"loss": 0.3072,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.44,
|
| 1196 |
+
"grad_norm": 0.7072462535931962,
|
| 1197 |
+
"learning_rate": 3.590291005954094e-06,
|
| 1198 |
+
"loss": 0.3269,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.4719999999999995,
|
| 1203 |
+
"grad_norm": 0.5292063382688933,
|
| 1204 |
+
"learning_rate": 3.570726100757693e-06,
|
| 1205 |
+
"loss": 0.2453,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.504,
|
| 1210 |
+
"grad_norm": 0.6996808757407292,
|
| 1211 |
+
"learning_rate": 3.5510805541224536e-06,
|
| 1212 |
+
"loss": 0.3099,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.536,
|
| 1217 |
+
"grad_norm": 0.8116009713344682,
|
| 1218 |
+
"learning_rate": 3.5313558456472623e-06,
|
| 1219 |
+
"loss": 0.363,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.568,
|
| 1224 |
+
"grad_norm": 0.6416602092341779,
|
| 1225 |
+
"learning_rate": 3.511553460893059e-06,
|
| 1226 |
+
"loss": 0.2769,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.6,
|
| 1231 |
+
"grad_norm": 0.6808214771084572,
|
| 1232 |
+
"learning_rate": 3.4916748912709506e-06,
|
| 1233 |
+
"loss": 0.3369,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.632,
|
| 1238 |
+
"grad_norm": 0.7350438523202055,
|
| 1239 |
+
"learning_rate": 3.4717216339298852e-06,
|
| 1240 |
+
"loss": 0.3278,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.664,
|
| 1245 |
+
"grad_norm": 0.8128907625049909,
|
| 1246 |
+
"learning_rate": 3.4516951916438974e-06,
|
| 1247 |
+
"loss": 0.3162,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.696,
|
| 1252 |
+
"grad_norm": 0.7145536698935532,
|
| 1253 |
+
"learning_rate": 3.4315970726989244e-06,
|
| 1254 |
+
"loss": 0.3166,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.728,
|
| 1259 |
+
"grad_norm": 0.7686705822658882,
|
| 1260 |
+
"learning_rate": 3.4114287907792115e-06,
|
| 1261 |
+
"loss": 0.3537,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.76,
|
| 1266 |
+
"grad_norm": 0.5815129260863527,
|
| 1267 |
+
"learning_rate": 3.3911918648533094e-06,
|
| 1268 |
+
"loss": 0.2218,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.792,
|
| 1273 |
+
"grad_norm": 0.6822703900197954,
|
| 1274 |
+
"learning_rate": 3.3708878190596724e-06,
|
| 1275 |
+
"loss": 0.343,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.824,
|
| 1280 |
+
"grad_norm": 0.6349682181482348,
|
| 1281 |
+
"learning_rate": 3.3505181825918685e-06,
|
| 1282 |
+
"loss": 0.2944,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.856,
|
| 1287 |
+
"grad_norm": 0.584111392203079,
|
| 1288 |
+
"learning_rate": 3.33008448958341e-06,
|
| 1289 |
+
"loss": 0.2838,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.888,
|
| 1294 |
+
"grad_norm": 0.6895618507758892,
|
| 1295 |
+
"learning_rate": 3.3095882789922102e-06,
|
| 1296 |
+
"loss": 0.3411,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.92,
|
| 1301 |
+
"grad_norm": 0.6812997509973548,
|
| 1302 |
+
"learning_rate": 3.289031094484675e-06,
|
| 1303 |
+
"loss": 0.3241,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.952,
|
| 1308 |
+
"grad_norm": 0.7219147628714953,
|
| 1309 |
+
"learning_rate": 3.2684144843194453e-06,
|
| 1310 |
+
"loss": 0.3361,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.984,
|
| 1315 |
+
"grad_norm": 0.6575456923543451,
|
| 1316 |
+
"learning_rate": 3.2477400012307885e-06,
|
| 1317 |
+
"loss": 0.2697,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 6.016,
|
| 1322 |
+
"grad_norm": 1.9361117400650483,
|
| 1323 |
+
"learning_rate": 3.2270092023116564e-06,
|
| 1324 |
+
"loss": 0.4916,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.048,
|
| 1329 |
+
"grad_norm": 0.8624146111834964,
|
| 1330 |
+
"learning_rate": 3.206223648896409e-06,
|
| 1331 |
+
"loss": 0.2441,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.08,
|
| 1336 |
+
"grad_norm": 0.6181079277130646,
|
| 1337 |
+
"learning_rate": 3.1853849064432296e-06,
|
| 1338 |
+
"loss": 0.238,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 6.112,
|
| 1343 |
+
"grad_norm": 0.8279914714122331,
|
| 1344 |
+
"learning_rate": 3.1644945444162155e-06,
|
| 1345 |
+
"loss": 0.2805,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.144,
|
| 1350 |
+
"grad_norm": 1.115725524421143,
|
| 1351 |
+
"learning_rate": 3.14355413616718e-06,
|
| 1352 |
+
"loss": 0.3259,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.176,
|
| 1357 |
+
"grad_norm": 0.7753636384497341,
|
| 1358 |
+
"learning_rate": 3.1225652588171534e-06,
|
| 1359 |
+
"loss": 0.2343,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.208,
|
| 1364 |
+
"grad_norm": 0.6333875659930861,
|
| 1365 |
+
"learning_rate": 3.1015294931376035e-06,
|
| 1366 |
+
"loss": 0.2695,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.24,
|
| 1371 |
+
"grad_norm": 0.8271269604878655,
|
| 1372 |
+
"learning_rate": 3.080448423431377e-06,
|
| 1373 |
+
"loss": 0.3782,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.272,
|
| 1378 |
+
"grad_norm": 0.7046561959257681,
|
| 1379 |
+
"learning_rate": 3.059323637413385e-06,
|
| 1380 |
+
"loss": 0.2541,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.304,
|
| 1385 |
+
"grad_norm": 0.7476879983154219,
|
| 1386 |
+
"learning_rate": 3.0381567260910166e-06,
|
| 1387 |
+
"loss": 0.2847,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.336,
|
| 1392 |
+
"grad_norm": 0.6329581264969493,
|
| 1393 |
+
"learning_rate": 3.016949283644319e-06,
|
| 1394 |
+
"loss": 0.2234,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.368,
|
| 1399 |
+
"grad_norm": 0.6899558720221424,
|
| 1400 |
+
"learning_rate": 2.9957029073059276e-06,
|
| 1401 |
+
"loss": 0.2377,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.4,
|
| 1406 |
+
"grad_norm": 0.8616157956876241,
|
| 1407 |
+
"learning_rate": 2.9744191972407754e-06,
|
| 1408 |
+
"loss": 0.2668,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.432,
|
| 1413 |
+
"grad_norm": 0.7419471681162961,
|
| 1414 |
+
"learning_rate": 2.9530997564255728e-06,
|
| 1415 |
+
"loss": 0.2021,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.464,
|
| 1420 |
+
"grad_norm": 0.9215679785568135,
|
| 1421 |
+
"learning_rate": 2.931746190528082e-06,
|
| 1422 |
+
"loss": 0.2637,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.496,
|
| 1427 |
+
"grad_norm": 0.765620021531001,
|
| 1428 |
+
"learning_rate": 2.9103601077861875e-06,
|
| 1429 |
+
"loss": 0.2598,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.5280000000000005,
|
| 1434 |
+
"grad_norm": 0.6871931050751866,
|
| 1435 |
+
"learning_rate": 2.888943118886771e-06,
|
| 1436 |
+
"loss": 0.2493,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.5600000000000005,
|
| 1441 |
+
"grad_norm": 0.8539836511097245,
|
| 1442 |
+
"learning_rate": 2.8674968368444004e-06,
|
| 1443 |
+
"loss": 0.2887,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.592,
|
| 1448 |
+
"grad_norm": 0.6703604513060638,
|
| 1449 |
+
"learning_rate": 2.8460228768798507e-06,
|
| 1450 |
+
"loss": 0.2162,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.624,
|
| 1455 |
+
"grad_norm": 0.6918871279471751,
|
| 1456 |
+
"learning_rate": 2.8245228562984518e-06,
|
| 1457 |
+
"loss": 0.2583,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.656,
|
| 1462 |
+
"grad_norm": 0.7160118882799036,
|
| 1463 |
+
"learning_rate": 2.80299839436828e-06,
|
| 1464 |
+
"loss": 0.2746,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.688,
|
| 1469 |
+
"grad_norm": 0.7666212843255689,
|
| 1470 |
+
"learning_rate": 2.781451112198208e-06,
|
| 1471 |
+
"loss": 0.2026,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.72,
|
| 1476 |
+
"grad_norm": 0.7480471545142591,
|
| 1477 |
+
"learning_rate": 2.7598826326158045e-06,
|
| 1478 |
+
"loss": 0.2233,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.752,
|
| 1483 |
+
"grad_norm": 0.6698124360013661,
|
| 1484 |
+
"learning_rate": 2.738294580045119e-06,
|
| 1485 |
+
"loss": 0.2196,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.784,
|
| 1490 |
+
"grad_norm": 0.7687920922845868,
|
| 1491 |
+
"learning_rate": 2.7166885803843347e-06,
|
| 1492 |
+
"loss": 0.2842,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.816,
|
| 1497 |
+
"grad_norm": 0.6690138590098085,
|
| 1498 |
+
"learning_rate": 2.695066260883313e-06,
|
| 1499 |
+
"loss": 0.2726,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.848,
|
| 1504 |
+
"grad_norm": 0.6356726755744938,
|
| 1505 |
+
"learning_rate": 2.67342925002104e-06,
|
| 1506 |
+
"loss": 0.2784,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.88,
|
| 1511 |
+
"grad_norm": 0.7278700477526574,
|
| 1512 |
+
"learning_rate": 2.65177917738298e-06,
|
| 1513 |
+
"loss": 0.2213,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.912,
|
| 1518 |
+
"grad_norm": 0.7920194138682944,
|
| 1519 |
+
"learning_rate": 2.6301176735383382e-06,
|
| 1520 |
+
"loss": 0.2218,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.944,
|
| 1525 |
+
"grad_norm": 0.6542992329887045,
|
| 1526 |
+
"learning_rate": 2.6084463699172594e-06,
|
| 1527 |
+
"loss": 0.2299,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.976,
|
| 1532 |
+
"grad_norm": 0.6655707546871679,
|
| 1533 |
+
"learning_rate": 2.586766898687955e-06,
|
| 1534 |
+
"loss": 0.251,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.008,
|
| 1539 |
+
"grad_norm": 1.6254077534613807,
|
| 1540 |
+
"learning_rate": 2.56508089263378e-06,
|
| 1541 |
+
"loss": 0.4276,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.04,
|
| 1546 |
+
"grad_norm": 0.7442411262872054,
|
| 1547 |
+
"learning_rate": 2.5433899850302552e-06,
|
| 1548 |
+
"loss": 0.192,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.072,
|
| 1553 |
+
"grad_norm": 0.6624456030796024,
|
| 1554 |
+
"learning_rate": 2.521695809522061e-06,
|
| 1555 |
+
"loss": 0.239,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.104,
|
| 1560 |
+
"grad_norm": 0.7058768600313585,
|
| 1561 |
+
"learning_rate": 2.5e-06,
|
| 1562 |
+
"loss": 0.1949,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.136,
|
| 1567 |
+
"grad_norm": 0.8810850224640968,
|
| 1568 |
+
"learning_rate": 2.4783041904779386e-06,
|
| 1569 |
+
"loss": 0.1902,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.168,
|
| 1574 |
+
"grad_norm": 0.8028975590134214,
|
| 1575 |
+
"learning_rate": 2.4566100149697456e-06,
|
| 1576 |
+
"loss": 0.26,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.2,
|
| 1581 |
+
"grad_norm": 0.6830163337607535,
|
| 1582 |
+
"learning_rate": 2.4349191073662203e-06,
|
| 1583 |
+
"loss": 0.2409,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.232,
|
| 1588 |
+
"grad_norm": 0.6064918380966279,
|
| 1589 |
+
"learning_rate": 2.4132331013120454e-06,
|
| 1590 |
+
"loss": 0.1543,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.264,
|
| 1595 |
+
"grad_norm": 0.8748849905978056,
|
| 1596 |
+
"learning_rate": 2.3915536300827414e-06,
|
| 1597 |
+
"loss": 0.2591,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.296,
|
| 1602 |
+
"grad_norm": 0.6651366346439012,
|
| 1603 |
+
"learning_rate": 2.3698823264616635e-06,
|
| 1604 |
+
"loss": 0.2163,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.328,
|
| 1609 |
+
"grad_norm": 0.7188473278111397,
|
| 1610 |
+
"learning_rate": 2.348220822617021e-06,
|
| 1611 |
+
"loss": 0.2176,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.36,
|
| 1616 |
+
"grad_norm": 0.7020229021241178,
|
| 1617 |
+
"learning_rate": 2.32657074997896e-06,
|
| 1618 |
+
"loss": 0.1984,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.392,
|
| 1623 |
+
"grad_norm": 0.66211258097335,
|
| 1624 |
+
"learning_rate": 2.3049337391166884e-06,
|
| 1625 |
+
"loss": 0.1867,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.424,
|
| 1630 |
+
"grad_norm": 0.9573373282078618,
|
| 1631 |
+
"learning_rate": 2.2833114196156657e-06,
|
| 1632 |
+
"loss": 0.2402,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.456,
|
| 1637 |
+
"grad_norm": 0.7357127100021363,
|
| 1638 |
+
"learning_rate": 2.261705419954882e-06,
|
| 1639 |
+
"loss": 0.231,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.4879999999999995,
|
| 1644 |
+
"grad_norm": 0.7585910321098319,
|
| 1645 |
+
"learning_rate": 2.2401173673841963e-06,
|
| 1646 |
+
"loss": 0.2043,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.52,
|
| 1651 |
+
"grad_norm": 0.6870683771667501,
|
| 1652 |
+
"learning_rate": 2.2185488878017934e-06,
|
| 1653 |
+
"loss": 0.1779,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.552,
|
| 1658 |
+
"grad_norm": 0.7676301772753004,
|
| 1659 |
+
"learning_rate": 2.1970016056317202e-06,
|
| 1660 |
+
"loss": 0.2107,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.584,
|
| 1665 |
+
"grad_norm": 0.7845976366946071,
|
| 1666 |
+
"learning_rate": 2.1754771437015495e-06,
|
| 1667 |
+
"loss": 0.1945,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.616,
|
| 1672 |
+
"grad_norm": 0.6962788867671439,
|
| 1673 |
+
"learning_rate": 2.1539771231201497e-06,
|
| 1674 |
+
"loss": 0.2064,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.648,
|
| 1679 |
+
"grad_norm": 0.7183769801965757,
|
| 1680 |
+
"learning_rate": 2.1325031631555996e-06,
|
| 1681 |
+
"loss": 0.1935,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.68,
|
| 1686 |
+
"grad_norm": 0.7471815960185075,
|
| 1687 |
+
"learning_rate": 2.11105688111323e-06,
|
| 1688 |
+
"loss": 0.2048,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.712,
|
| 1693 |
+
"grad_norm": 0.9190154307124283,
|
| 1694 |
+
"learning_rate": 2.0896398922138124e-06,
|
| 1695 |
+
"loss": 0.1969,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.744,
|
| 1700 |
+
"grad_norm": 0.7243350103336731,
|
| 1701 |
+
"learning_rate": 2.0682538094719183e-06,
|
| 1702 |
+
"loss": 0.2069,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.776,
|
| 1707 |
+
"grad_norm": 0.749579471338467,
|
| 1708 |
+
"learning_rate": 2.0469002435744285e-06,
|
| 1709 |
+
"loss": 0.2671,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.808,
|
| 1714 |
+
"grad_norm": 0.6522335743524954,
|
| 1715 |
+
"learning_rate": 2.0255808027592263e-06,
|
| 1716 |
+
"loss": 0.2016,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.84,
|
| 1721 |
+
"grad_norm": 0.7471642009931441,
|
| 1722 |
+
"learning_rate": 2.004297092694073e-06,
|
| 1723 |
+
"loss": 0.1799,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.872,
|
| 1728 |
+
"grad_norm": 0.8329249368508794,
|
| 1729 |
+
"learning_rate": 1.9830507163556815e-06,
|
| 1730 |
+
"loss": 0.2102,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.904,
|
| 1735 |
+
"grad_norm": 0.6853964773958638,
|
| 1736 |
+
"learning_rate": 1.9618432739089843e-06,
|
| 1737 |
+
"loss": 0.2185,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.936,
|
| 1742 |
+
"grad_norm": 0.6681326615863377,
|
| 1743 |
+
"learning_rate": 1.9406763625866155e-06,
|
| 1744 |
+
"loss": 0.1915,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.968,
|
| 1749 |
+
"grad_norm": 0.7444442305306697,
|
| 1750 |
+
"learning_rate": 1.9195515765686237e-06,
|
| 1751 |
+
"loss": 0.2099,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 8.0,
|
| 1756 |
+
"grad_norm": 1.4566062560033979,
|
| 1757 |
+
"learning_rate": 1.8984705068623976e-06,
|
| 1758 |
+
"loss": 0.4011,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 8.032,
|
| 1763 |
+
"grad_norm": 0.807906744593382,
|
| 1764 |
+
"learning_rate": 1.8774347411828472e-06,
|
| 1765 |
+
"loss": 0.1878,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 8.064,
|
| 1770 |
+
"grad_norm": 0.6688228009880165,
|
| 1771 |
+
"learning_rate": 1.8564458638328203e-06,
|
| 1772 |
+
"loss": 0.1922,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 8.096,
|
| 1777 |
+
"grad_norm": 0.6946399190105664,
|
| 1778 |
+
"learning_rate": 1.835505455583786e-06,
|
| 1779 |
+
"loss": 0.148,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 8.128,
|
| 1784 |
+
"grad_norm": 0.8018686319524767,
|
| 1785 |
+
"learning_rate": 1.8146150935567712e-06,
|
| 1786 |
+
"loss": 0.2095,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 8.16,
|
| 1791 |
+
"grad_norm": 0.7979997938637724,
|
| 1792 |
+
"learning_rate": 1.7937763511035904e-06,
|
| 1793 |
+
"loss": 0.1921,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 8.192,
|
| 1798 |
+
"grad_norm": 0.8896660800631546,
|
| 1799 |
+
"learning_rate": 1.7729907976883443e-06,
|
| 1800 |
+
"loss": 0.1718,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.224,
|
| 1805 |
+
"grad_norm": 0.7699230180050226,
|
| 1806 |
+
"learning_rate": 1.7522599987692113e-06,
|
| 1807 |
+
"loss": 0.1362,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 8.256,
|
| 1812 |
+
"grad_norm": 0.6852316009614421,
|
| 1813 |
+
"learning_rate": 1.7315855156805558e-06,
|
| 1814 |
+
"loss": 0.1742,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.288,
|
| 1819 |
+
"grad_norm": 0.8184547361678618,
|
| 1820 |
+
"learning_rate": 1.7109689055153261e-06,
|
| 1821 |
+
"loss": 0.2072,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.32,
|
| 1826 |
+
"grad_norm": 0.7175234706304453,
|
| 1827 |
+
"learning_rate": 1.690411721007791e-06,
|
| 1828 |
+
"loss": 0.1904,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 8.352,
|
| 1833 |
+
"grad_norm": 0.7102437999512361,
|
| 1834 |
+
"learning_rate": 1.6699155104165903e-06,
|
| 1835 |
+
"loss": 0.2296,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 8.384,
|
| 1840 |
+
"grad_norm": 0.7101121069921504,
|
| 1841 |
+
"learning_rate": 1.6494818174081317e-06,
|
| 1842 |
+
"loss": 0.1656,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.416,
|
| 1847 |
+
"grad_norm": 0.7051481377369857,
|
| 1848 |
+
"learning_rate": 1.6291121809403287e-06,
|
| 1849 |
+
"loss": 0.1434,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 8.448,
|
| 1854 |
+
"grad_norm": 0.7867966792973216,
|
| 1855 |
+
"learning_rate": 1.6088081351466908e-06,
|
| 1856 |
+
"loss": 0.15,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.48,
|
| 1861 |
+
"grad_norm": 0.779555799957025,
|
| 1862 |
+
"learning_rate": 1.588571209220789e-06,
|
| 1863 |
+
"loss": 0.1529,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.512,
|
| 1868 |
+
"grad_norm": 0.7151181263869698,
|
| 1869 |
+
"learning_rate": 1.5684029273010762e-06,
|
| 1870 |
+
"loss": 0.1864,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 8.544,
|
| 1875 |
+
"grad_norm": 0.6669233131764054,
|
| 1876 |
+
"learning_rate": 1.5483048083561036e-06,
|
| 1877 |
+
"loss": 0.1519,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.576,
|
| 1882 |
+
"grad_norm": 0.7095219149224068,
|
| 1883 |
+
"learning_rate": 1.5282783660701154e-06,
|
| 1884 |
+
"loss": 0.1392,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.608,
|
| 1889 |
+
"grad_norm": 0.7343062096757953,
|
| 1890 |
+
"learning_rate": 1.5083251087290506e-06,
|
| 1891 |
+
"loss": 0.1706,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.64,
|
| 1896 |
+
"grad_norm": 0.6725850104265888,
|
| 1897 |
+
"learning_rate": 1.4884465391069415e-06,
|
| 1898 |
+
"loss": 0.1928,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.672,
|
| 1903 |
+
"grad_norm": 0.7121028680803699,
|
| 1904 |
+
"learning_rate": 1.4686441543527374e-06,
|
| 1905 |
+
"loss": 0.1608,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.704,
|
| 1910 |
+
"grad_norm": 0.6788157788458534,
|
| 1911 |
+
"learning_rate": 1.4489194458775468e-06,
|
| 1912 |
+
"loss": 0.1737,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.736,
|
| 1917 |
+
"grad_norm": 0.7240495902038667,
|
| 1918 |
+
"learning_rate": 1.4292738992423066e-06,
|
| 1919 |
+
"loss": 0.1674,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.768,
|
| 1924 |
+
"grad_norm": 0.7360321208085325,
|
| 1925 |
+
"learning_rate": 1.409708994045907e-06,
|
| 1926 |
+
"loss": 0.1816,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.8,
|
| 1931 |
+
"grad_norm": 0.6902317343584977,
|
| 1932 |
+
"learning_rate": 1.3902262038137449e-06,
|
| 1933 |
+
"loss": 0.1476,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.832,
|
| 1938 |
+
"grad_norm": 0.7372464472923361,
|
| 1939 |
+
"learning_rate": 1.3708269958867565e-06,
|
| 1940 |
+
"loss": 0.1899,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.864,
|
| 1945 |
+
"grad_norm": 0.6479442020833953,
|
| 1946 |
+
"learning_rate": 1.3515128313108966e-06,
|
| 1947 |
+
"loss": 0.1789,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.896,
|
| 1952 |
+
"grad_norm": 1.4963792930934798,
|
| 1953 |
+
"learning_rate": 1.3322851647271057e-06,
|
| 1954 |
+
"loss": 0.1663,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 8.928,
|
| 1959 |
+
"grad_norm": 0.8005763202829219,
|
| 1960 |
+
"learning_rate": 1.3131454442617523e-06,
|
| 1961 |
+
"loss": 0.1855,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.96,
|
| 1966 |
+
"grad_norm": 0.7589380859711553,
|
| 1967 |
+
"learning_rate": 1.2940951114175637e-06,
|
| 1968 |
+
"loss": 0.1758,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 8.992,
|
| 1973 |
+
"grad_norm": 0.7559641171217635,
|
| 1974 |
+
"learning_rate": 1.275135600965068e-06,
|
| 1975 |
+
"loss": 0.1863,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 9.024,
|
| 1980 |
+
"grad_norm": 1.5938062678433065,
|
| 1981 |
+
"learning_rate": 1.2562683408345279e-06,
|
| 1982 |
+
"loss": 0.2434,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 9.056,
|
| 1987 |
+
"grad_norm": 0.701006167797008,
|
| 1988 |
+
"learning_rate": 1.237494752008399e-06,
|
| 1989 |
+
"loss": 0.1359,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 9.088,
|
| 1994 |
+
"grad_norm": 0.711810966761551,
|
| 1995 |
+
"learning_rate": 1.2188162484143077e-06,
|
| 1996 |
+
"loss": 0.1765,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 9.12,
|
| 2001 |
+
"grad_norm": 0.6996994520287924,
|
| 2002 |
+
"learning_rate": 1.2002342368185638e-06,
|
| 2003 |
+
"loss": 0.1495,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 9.152,
|
| 2008 |
+
"grad_norm": 0.7069981659814839,
|
| 2009 |
+
"learning_rate": 1.18175011672021e-06,
|
| 2010 |
+
"loss": 0.1405,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 9.184,
|
| 2015 |
+
"grad_norm": 0.8231991171658294,
|
| 2016 |
+
"learning_rate": 1.163365280245615e-06,
|
| 2017 |
+
"loss": 0.1323,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 9.216,
|
| 2022 |
+
"grad_norm": 0.8796942711698479,
|
| 2023 |
+
"learning_rate": 1.1450811120436319e-06,
|
| 2024 |
+
"loss": 0.2059,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 9.248,
|
| 2029 |
+
"grad_norm": 0.6122559474891257,
|
| 2030 |
+
"learning_rate": 1.1268989891813085e-06,
|
| 2031 |
+
"loss": 0.0822,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 9.28,
|
| 2036 |
+
"grad_norm": 0.6979533439001652,
|
| 2037 |
+
"learning_rate": 1.1088202810401789e-06,
|
| 2038 |
+
"loss": 0.1545,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 9.312,
|
| 2043 |
+
"grad_norm": 0.6800422735298531,
|
| 2044 |
+
"learning_rate": 1.0908463492131227e-06,
|
| 2045 |
+
"loss": 0.1369,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 9.344,
|
| 2050 |
+
"grad_norm": 0.710398063279894,
|
| 2051 |
+
"learning_rate": 1.0729785474018243e-06,
|
| 2052 |
+
"loss": 0.1371,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 9.376,
|
| 2057 |
+
"grad_norm": 0.8205147959211968,
|
| 2058 |
+
"learning_rate": 1.0552182213148119e-06,
|
| 2059 |
+
"loss": 0.1666,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 9.408,
|
| 2064 |
+
"grad_norm": 0.6760959754671624,
|
| 2065 |
+
"learning_rate": 1.0375667085661115e-06,
|
| 2066 |
+
"loss": 0.1243,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 9.44,
|
| 2071 |
+
"grad_norm": 0.632589652861321,
|
| 2072 |
+
"learning_rate": 1.020025338574504e-06,
|
| 2073 |
+
"loss": 0.1065,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 9.472,
|
| 2078 |
+
"grad_norm": 0.705025934687606,
|
| 2079 |
+
"learning_rate": 1.0025954324633949e-06,
|
| 2080 |
+
"loss": 0.1742,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 9.504,
|
| 2085 |
+
"grad_norm": 0.7345035982463306,
|
| 2086 |
+
"learning_rate": 9.852783029613224e-07,
|
| 2087 |
+
"loss": 0.1533,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.536,
|
| 2092 |
+
"grad_norm": 0.7178327431036359,
|
| 2093 |
+
"learning_rate": 9.680752543030844e-07,
|
| 2094 |
+
"loss": 0.1605,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 9.568,
|
| 2099 |
+
"grad_norm": 0.6819410695986909,
|
| 2100 |
+
"learning_rate": 9.509875821315126e-07,
|
| 2101 |
+
"loss": 0.161,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 9.6,
|
| 2106 |
+
"grad_norm": 0.7234416739774964,
|
| 2107 |
+
"learning_rate": 9.340165733998877e-07,
|
| 2108 |
+
"loss": 0.1598,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.632,
|
| 2113 |
+
"grad_norm": 0.73130167347675,
|
| 2114 |
+
"learning_rate": 9.171635062750189e-07,
|
| 2115 |
+
"loss": 0.1635,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 9.664,
|
| 2120 |
+
"grad_norm": 0.658131522430436,
|
| 2121 |
+
"learning_rate": 9.004296500409759e-07,
|
| 2122 |
+
"loss": 0.0895,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.696,
|
| 2127 |
+
"grad_norm": 0.6947321589176428,
|
| 2128 |
+
"learning_rate": 8.838162650034912e-07,
|
| 2129 |
+
"loss": 0.1322,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.728,
|
| 2134 |
+
"grad_norm": 0.5871125396728818,
|
| 2135 |
+
"learning_rate": 8.673246023950449e-07,
|
| 2136 |
+
"loss": 0.1195,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 9.76,
|
| 2141 |
+
"grad_norm": 0.7283051233869737,
|
| 2142 |
+
"learning_rate": 8.509559042806237e-07,
|
| 2143 |
+
"loss": 0.1897,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.792,
|
| 2148 |
+
"grad_norm": 0.7889585744548698,
|
| 2149 |
+
"learning_rate": 8.347114034641807e-07,
|
| 2150 |
+
"loss": 0.1941,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 9.824,
|
| 2155 |
+
"grad_norm": 0.6250961539692981,
|
| 2156 |
+
"learning_rate": 8.185923233957802e-07,
|
| 2157 |
+
"loss": 0.1534,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 9.856,
|
| 2162 |
+
"grad_norm": 0.7087141973379683,
|
| 2163 |
+
"learning_rate": 8.025998780794622e-07,
|
| 2164 |
+
"loss": 0.1353,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 9.888,
|
| 2169 |
+
"grad_norm": 0.680515928146072,
|
| 2170 |
+
"learning_rate": 7.867352719818008e-07,
|
| 2171 |
+
"loss": 0.1408,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 9.92,
|
| 2176 |
+
"grad_norm": 0.6675163947700695,
|
| 2177 |
+
"learning_rate": 7.709996999411984e-07,
|
| 2178 |
+
"loss": 0.1548,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 9.952,
|
| 2183 |
+
"grad_norm": 0.6766792515322745,
|
| 2184 |
+
"learning_rate": 7.553943470778927e-07,
|
| 2185 |
+
"loss": 0.1442,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 9.984,
|
| 2190 |
+
"grad_norm": 0.6566160333248258,
|
| 2191 |
+
"learning_rate": 7.399203887046977e-07,
|
| 2192 |
+
"loss": 0.0959,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 10.016,
|
| 2197 |
+
"grad_norm": 1.666679527816923,
|
| 2198 |
+
"learning_rate": 7.245789902384908e-07,
|
| 2199 |
+
"loss": 0.1971,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 10.048,
|
| 2204 |
+
"grad_norm": 0.6356380096069124,
|
| 2205 |
+
"learning_rate": 7.093713071124361e-07,
|
| 2206 |
+
"loss": 0.0995,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 10.08,
|
| 2211 |
+
"grad_norm": 0.7577099898340935,
|
| 2212 |
+
"learning_rate": 6.942984846889639e-07,
|
| 2213 |
+
"loss": 0.1613,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 10.112,
|
| 2218 |
+
"grad_norm": 0.5351324688835172,
|
| 2219 |
+
"learning_rate": 6.793616581735063e-07,
|
| 2220 |
+
"loss": 0.1133,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 10.144,
|
| 2225 |
+
"grad_norm": 0.6796708376296814,
|
| 2226 |
+
"learning_rate": 6.645619525290043e-07,
|
| 2227 |
+
"loss": 0.1333,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 10.176,
|
| 2232 |
+
"grad_norm": 0.6209416411542945,
|
| 2233 |
+
"learning_rate": 6.499004823911772e-07,
|
| 2234 |
+
"loss": 0.1122,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 10.208,
|
| 2239 |
+
"grad_norm": 0.7628687626514484,
|
| 2240 |
+
"learning_rate": 6.353783519845752e-07,
|
| 2241 |
+
"loss": 0.1616,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 10.24,
|
| 2246 |
+
"grad_norm": 0.6641480174740878,
|
| 2247 |
+
"learning_rate": 6.209966550394162e-07,
|
| 2248 |
+
"loss": 0.1147,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 10.272,
|
| 2253 |
+
"grad_norm": 0.8460363259730189,
|
| 2254 |
+
"learning_rate": 6.067564747092095e-07,
|
| 2255 |
+
"loss": 0.1201,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 10.304,
|
| 2260 |
+
"grad_norm": 0.5625566771720574,
|
| 2261 |
+
"learning_rate": 5.926588834891823e-07,
|
| 2262 |
+
"loss": 0.0827,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 10.336,
|
| 2267 |
+
"grad_norm": 0.7384654603091378,
|
| 2268 |
+
"learning_rate": 5.787049431354996e-07,
|
| 2269 |
+
"loss": 0.1626,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 10.368,
|
| 2274 |
+
"grad_norm": 0.6626117899249079,
|
| 2275 |
+
"learning_rate": 5.648957045853043e-07,
|
| 2276 |
+
"loss": 0.1403,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 10.4,
|
| 2281 |
+
"grad_norm": 0.61825633629735,
|
| 2282 |
+
"learning_rate": 5.512322078775603e-07,
|
| 2283 |
+
"loss": 0.1175,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 10.432,
|
| 2288 |
+
"grad_norm": 0.6702571182619494,
|
| 2289 |
+
"learning_rate": 5.377154820747271e-07,
|
| 2290 |
+
"loss": 0.1248,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 10.464,
|
| 2295 |
+
"grad_norm": 0.647331031269689,
|
| 2296 |
+
"learning_rate": 5.243465451852548e-07,
|
| 2297 |
+
"loss": 0.1534,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 10.496,
|
| 2302 |
+
"grad_norm": 0.6426713733629653,
|
| 2303 |
+
"learning_rate": 5.111264040869093e-07,
|
| 2304 |
+
"loss": 0.1201,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 10.528,
|
| 2309 |
+
"grad_norm": 0.7085063709660798,
|
| 2310 |
+
"learning_rate": 4.980560544509467e-07,
|
| 2311 |
+
"loss": 0.1804,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 10.56,
|
| 2316 |
+
"grad_norm": 0.5968202164061807,
|
| 2317 |
+
"learning_rate": 4.85136480667118e-07,
|
| 2318 |
+
"loss": 0.0845,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 10.592,
|
| 2323 |
+
"grad_norm": 0.6309930342886219,
|
| 2324 |
+
"learning_rate": 4.723686557695351e-07,
|
| 2325 |
+
"loss": 0.1359,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 10.624,
|
| 2330 |
+
"grad_norm": 0.7374835893325036,
|
| 2331 |
+
"learning_rate": 4.5975354136338164e-07,
|
| 2332 |
+
"loss": 0.1338,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 10.656,
|
| 2337 |
+
"grad_norm": 0.6457702995416986,
|
| 2338 |
+
"learning_rate": 4.472920875524958e-07,
|
| 2339 |
+
"loss": 0.0792,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 10.688,
|
| 2344 |
+
"grad_norm": 0.7537820690789252,
|
| 2345 |
+
"learning_rate": 4.3498523286780973e-07,
|
| 2346 |
+
"loss": 0.1812,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 10.72,
|
| 2351 |
+
"grad_norm": 0.6479476153254174,
|
| 2352 |
+
"learning_rate": 4.228339041966645e-07,
|
| 2353 |
+
"loss": 0.1024,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 10.752,
|
| 2358 |
+
"grad_norm": 0.637060896337492,
|
| 2359 |
+
"learning_rate": 4.108390167130044e-07,
|
| 2360 |
+
"loss": 0.1257,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 10.784,
|
| 2365 |
+
"grad_norm": 0.7973703777912577,
|
| 2366 |
+
"learning_rate": 3.9900147380844716e-07,
|
| 2367 |
+
"loss": 0.1469,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 10.816,
|
| 2372 |
+
"grad_norm": 0.6487907789946615,
|
| 2373 |
+
"learning_rate": 3.8732216702424915e-07,
|
| 2374 |
+
"loss": 0.0819,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 10.848,
|
| 2379 |
+
"grad_norm": 0.7597496101463044,
|
| 2380 |
+
"learning_rate": 3.7580197598415523e-07,
|
| 2381 |
+
"loss": 0.1024,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 10.88,
|
| 2386 |
+
"grad_norm": 0.7144533623534853,
|
| 2387 |
+
"learning_rate": 3.644417683281551e-07,
|
| 2388 |
+
"loss": 0.1473,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 10.912,
|
| 2393 |
+
"grad_norm": 0.5931719058935477,
|
| 2394 |
+
"learning_rate": 3.532423996471307e-07,
|
| 2395 |
+
"loss": 0.1143,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 10.943999999999999,
|
| 2400 |
+
"grad_norm": 0.6817689947273881,
|
| 2401 |
+
"learning_rate": 3.422047134184242e-07,
|
| 2402 |
+
"loss": 0.1472,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 10.975999999999999,
|
| 2407 |
+
"grad_norm": 0.6648780783348059,
|
| 2408 |
+
"learning_rate": 3.3132954094230813e-07,
|
| 2409 |
+
"loss": 0.1167,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 11.008,
|
| 2414 |
+
"grad_norm": 1.3316050451480825,
|
| 2415 |
+
"learning_rate": 3.2061770127937544e-07,
|
| 2416 |
+
"loss": 0.2483,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 11.04,
|
| 2421 |
+
"grad_norm": 0.6101504773978235,
|
| 2422 |
+
"learning_rate": 3.1007000118885523e-07,
|
| 2423 |
+
"loss": 0.1096,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 11.072,
|
| 2428 |
+
"grad_norm": 0.6472396011770947,
|
| 2429 |
+
"learning_rate": 2.9968723506784954e-07,
|
| 2430 |
+
"loss": 0.1102,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 11.104,
|
| 2435 |
+
"grad_norm": 0.6841951132605096,
|
| 2436 |
+
"learning_rate": 2.894701848915052e-07,
|
| 2437 |
+
"loss": 0.1367,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 11.136,
|
| 2442 |
+
"grad_norm": 0.5210353146864419,
|
| 2443 |
+
"learning_rate": 2.794196201541169e-07,
|
| 2444 |
+
"loss": 0.0796,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 11.168,
|
| 2449 |
+
"grad_norm": 0.6874392373209909,
|
| 2450 |
+
"learning_rate": 2.6953629781117654e-07,
|
| 2451 |
+
"loss": 0.1534,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 11.2,
|
| 2456 |
+
"grad_norm": 0.5383282895471744,
|
| 2457 |
+
"learning_rate": 2.598209622223616e-07,
|
| 2458 |
+
"loss": 0.086,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 11.232,
|
| 2463 |
+
"grad_norm": 0.6186561693838234,
|
| 2464 |
+
"learning_rate": 2.5027434509547143e-07,
|
| 2465 |
+
"loss": 0.0912,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 11.264,
|
| 2470 |
+
"grad_norm": 0.6508840344008129,
|
| 2471 |
+
"learning_rate": 2.408971654313244e-07,
|
| 2472 |
+
"loss": 0.1095,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 11.296,
|
| 2477 |
+
"grad_norm": 0.6777750107313413,
|
| 2478 |
+
"learning_rate": 2.3169012946960083e-07,
|
| 2479 |
+
"loss": 0.1439,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 11.328,
|
| 2484 |
+
"grad_norm": 0.5477692551314409,
|
| 2485 |
+
"learning_rate": 2.2265393063565787e-07,
|
| 2486 |
+
"loss": 0.0868,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 11.36,
|
| 2491 |
+
"grad_norm": 0.5869701558205148,
|
| 2492 |
+
"learning_rate": 2.1378924948829926e-07,
|
| 2493 |
+
"loss": 0.0941,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 11.392,
|
| 2498 |
+
"grad_norm": 0.7737711045633633,
|
| 2499 |
+
"learning_rate": 2.0509675366852333e-07,
|
| 2500 |
+
"loss": 0.1659,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 11.424,
|
| 2505 |
+
"grad_norm": 0.7284238102506699,
|
| 2506 |
+
"learning_rate": 1.96577097849239e-07,
|
| 2507 |
+
"loss": 0.1175,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 11.456,
|
| 2512 |
+
"grad_norm": 0.6026534091298396,
|
| 2513 |
+
"learning_rate": 1.8823092368595642e-07,
|
| 2514 |
+
"loss": 0.1044,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 11.488,
|
| 2519 |
+
"grad_norm": 0.645979672754451,
|
| 2520 |
+
"learning_rate": 1.8005885976846522e-07,
|
| 2521 |
+
"loss": 0.098,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 11.52,
|
| 2526 |
+
"grad_norm": 0.5839131758494809,
|
| 2527 |
+
"learning_rate": 1.7206152157348806e-07,
|
| 2528 |
+
"loss": 0.1059,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 11.552,
|
| 2533 |
+
"grad_norm": 0.6459615979459279,
|
| 2534 |
+
"learning_rate": 1.642395114183301e-07,
|
| 2535 |
+
"loss": 0.1243,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 11.584,
|
| 2540 |
+
"grad_norm": 0.6486777186241907,
|
| 2541 |
+
"learning_rate": 1.565934184155138e-07,
|
| 2542 |
+
"loss": 0.1113,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 11.616,
|
| 2547 |
+
"grad_norm": 0.6355899345873391,
|
| 2548 |
+
"learning_rate": 1.4912381842840884e-07,
|
| 2549 |
+
"loss": 0.1017,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 11.648,
|
| 2554 |
+
"grad_norm": 0.6498560023039535,
|
| 2555 |
+
"learning_rate": 1.4183127402786324e-07,
|
| 2556 |
+
"loss": 0.1255,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 11.68,
|
| 2561 |
+
"grad_norm": 0.4841473931899991,
|
| 2562 |
+
"learning_rate": 1.3471633444983312e-07,
|
| 2563 |
+
"loss": 0.0869,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 11.712,
|
| 2568 |
+
"grad_norm": 0.8047756699403728,
|
| 2569 |
+
"learning_rate": 1.2777953555401678e-07,
|
| 2570 |
+
"loss": 0.1842,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 11.744,
|
| 2575 |
+
"grad_norm": 0.5876396480845678,
|
| 2576 |
+
"learning_rate": 1.2102139978349497e-07,
|
| 2577 |
+
"loss": 0.1149,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 11.776,
|
| 2582 |
+
"grad_norm": 1.5264192867329822,
|
| 2583 |
+
"learning_rate": 1.144424361253868e-07,
|
| 2584 |
+
"loss": 0.1334,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 11.808,
|
| 2589 |
+
"grad_norm": 0.5435041856064506,
|
| 2590 |
+
"learning_rate": 1.0804314007251182e-07,
|
| 2591 |
+
"loss": 0.1301,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 11.84,
|
| 2596 |
+
"grad_norm": 0.6661075409363989,
|
| 2597 |
+
"learning_rate": 1.0182399358607542e-07,
|
| 2598 |
+
"loss": 0.1125,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 11.872,
|
| 2603 |
+
"grad_norm": 0.5677175975716918,
|
| 2604 |
+
"learning_rate": 9.578546505936675e-08,
|
| 2605 |
+
"loss": 0.113,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 11.904,
|
| 2610 |
+
"grad_norm": 0.6306908847089054,
|
| 2611 |
+
"learning_rate": 8.992800928248474e-08,
|
| 2612 |
+
"loss": 0.1059,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 11.936,
|
| 2617 |
+
"grad_norm": 0.6411331969032786,
|
| 2618 |
+
"learning_rate": 8.425206740808434e-08,
|
| 2619 |
+
"loss": 0.1153,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 11.968,
|
| 2624 |
+
"grad_norm": 0.515297730497707,
|
| 2625 |
+
"learning_rate": 7.875806691815064e-08,
|
| 2626 |
+
"loss": 0.0842,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 12.0,
|
| 2631 |
+
"grad_norm": 1.2295931484777365,
|
| 2632 |
+
"learning_rate": 7.34464215918046e-08,
|
| 2633 |
+
"loss": 0.2336,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 12.032,
|
| 2638 |
+
"grad_norm": 0.6205714133455,
|
| 2639 |
+
"learning_rate": 6.831753147413828e-08,
|
| 2640 |
+
"loss": 0.1519,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 12.064,
|
| 2645 |
+
"grad_norm": 0.6575189486492709,
|
| 2646 |
+
"learning_rate": 6.337178284608641e-08,
|
| 2647 |
+
"loss": 0.0928,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 12.096,
|
| 2652 |
+
"grad_norm": 0.5471078967999192,
|
| 2653 |
+
"learning_rate": 5.860954819533249e-08,
|
| 2654 |
+
"loss": 0.1174,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 12.128,
|
| 2659 |
+
"grad_norm": 0.5784338830953344,
|
| 2660 |
+
"learning_rate": 5.40311861882567e-08,
|
| 2661 |
+
"loss": 0.1208,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 12.16,
|
| 2666 |
+
"grad_norm": 0.5588173858413318,
|
| 2667 |
+
"learning_rate": 4.963704164292205e-08,
|
| 2668 |
+
"loss": 0.1118,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 12.192,
|
| 2673 |
+
"grad_norm": 0.6051147792562381,
|
| 2674 |
+
"learning_rate": 4.542744550310368e-08,
|
| 2675 |
+
"loss": 0.1107,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 12.224,
|
| 2680 |
+
"grad_norm": 0.6010092357005633,
|
| 2681 |
+
"learning_rate": 4.140271481336583e-08,
|
| 2682 |
+
"loss": 0.1171,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 12.256,
|
| 2687 |
+
"grad_norm": 0.6159762972953509,
|
| 2688 |
+
"learning_rate": 3.756315269518251e-08,
|
| 2689 |
+
"loss": 0.109,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 12.288,
|
| 2694 |
+
"grad_norm": 0.5921221747202706,
|
| 2695 |
+
"learning_rate": 3.39090483241078e-08,
|
| 2696 |
+
"loss": 0.125,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 12.32,
|
| 2701 |
+
"grad_norm": 0.5986695019052699,
|
| 2702 |
+
"learning_rate": 3.044067690799707e-08,
|
| 2703 |
+
"loss": 0.1158,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 12.352,
|
| 2708 |
+
"grad_norm": 0.5822260769678665,
|
| 2709 |
+
"learning_rate": 2.7158299666280864e-08,
|
| 2710 |
+
"loss": 0.1119,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 12.384,
|
| 2715 |
+
"grad_norm": 0.6184814451515146,
|
| 2716 |
+
"learning_rate": 2.4062163810288363e-08,
|
| 2717 |
+
"loss": 0.138,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 12.416,
|
| 2722 |
+
"grad_norm": 0.6384272384322683,
|
| 2723 |
+
"learning_rate": 2.1152502524632025e-08,
|
| 2724 |
+
"loss": 0.1261,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 12.448,
|
| 2729 |
+
"grad_norm": 0.5797511002078215,
|
| 2730 |
+
"learning_rate": 1.842953494964328e-08,
|
| 2731 |
+
"loss": 0.1273,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 12.48,
|
| 2736 |
+
"grad_norm": 0.6220225661738714,
|
| 2737 |
+
"learning_rate": 1.5893466164868255e-08,
|
| 2738 |
+
"loss": 0.131,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 12.512,
|
| 2743 |
+
"grad_norm": 0.6201657333450589,
|
| 2744 |
+
"learning_rate": 1.3544487173623444e-08,
|
| 2745 |
+
"loss": 0.1173,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 12.544,
|
| 2750 |
+
"grad_norm": 0.5888172014490931,
|
| 2751 |
+
"learning_rate": 1.1382774888608894e-08,
|
| 2752 |
+
"loss": 0.1272,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 12.576,
|
| 2757 |
+
"grad_norm": 0.5683202518706545,
|
| 2758 |
+
"learning_rate": 9.408492118584689e-09,
|
| 2759 |
+
"loss": 0.1049,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 12.608,
|
| 2764 |
+
"grad_norm": 0.5817766356891416,
|
| 2765 |
+
"learning_rate": 7.621787556109372e-09,
|
| 2766 |
+
"loss": 0.1095,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 12.64,
|
| 2771 |
+
"grad_norm": 0.6733083864881899,
|
| 2772 |
+
"learning_rate": 6.022795766340572e-09,
|
| 2773 |
+
"loss": 0.1072,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 12.672,
|
| 2778 |
+
"grad_norm": 0.5680755002498761,
|
| 2779 |
+
"learning_rate": 4.611637176901163e-09,
|
| 2780 |
+
"loss": 0.1067,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 12.704,
|
| 2785 |
+
"grad_norm": 0.598862935991944,
|
| 2786 |
+
"learning_rate": 3.3884180688076308e-09,
|
| 2787 |
+
"loss": 0.0976,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 12.736,
|
| 2792 |
+
"grad_norm": 0.605781750949151,
|
| 2793 |
+
"learning_rate": 2.353230568467868e-09,
|
| 2794 |
+
"loss": 0.1124,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 12.768,
|
| 2799 |
+
"grad_norm": 0.595819965756556,
|
| 2800 |
+
"learning_rate": 1.5061526407406102e-09,
|
| 2801 |
+
"loss": 0.0926,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 12.8,
|
| 2806 |
+
"grad_norm": 0.6003908503433764,
|
| 2807 |
+
"learning_rate": 8.472480830648555e-10,
|
| 2808 |
+
"loss": 0.1247,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 12.832,
|
| 2813 |
+
"grad_norm": 0.6390456145859789,
|
| 2814 |
+
"learning_rate": 3.76566520653987e-10,
|
| 2815 |
+
"loss": 0.1381,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 12.864,
|
| 2820 |
+
"grad_norm": 0.5388200277229742,
|
| 2821 |
+
"learning_rate": 9.414340275959443e-11,
|
| 2822 |
+
"loss": 0.1071,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 12.896,
|
| 2827 |
+
"grad_norm": 0.561710962149391,
|
| 2828 |
+
"learning_rate": 0.0,
|
| 2829 |
+
"loss": 0.0858,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 12.896,
|
| 2834 |
+
"step": 403,
|
| 2835 |
+
"total_flos": 2.2087303787769037e+17,
|
| 2836 |
+
"train_loss": 0.3277998325731559,
|
| 2837 |
+
"train_runtime": 13876.437,
|
| 2838 |
+
"train_samples_per_second": 0.933,
|
| 2839 |
+
"train_steps_per_second": 0.029
|
| 2840 |
+
}
|
| 2841 |
+
],
|
| 2842 |
+
"logging_steps": 1.0,
|
| 2843 |
+
"max_steps": 403,
|
| 2844 |
+
"num_input_tokens_seen": 0,
|
| 2845 |
+
"num_train_epochs": 13,
|
| 2846 |
+
"save_steps": 500,
|
| 2847 |
+
"stateful_callbacks": {
|
| 2848 |
+
"TrainerControl": {
|
| 2849 |
+
"args": {
|
| 2850 |
+
"should_epoch_stop": false,
|
| 2851 |
+
"should_evaluate": false,
|
| 2852 |
+
"should_log": false,
|
| 2853 |
+
"should_save": true,
|
| 2854 |
+
"should_training_stop": true
|
| 2855 |
+
},
|
| 2856 |
+
"attributes": {}
|
| 2857 |
+
}
|
| 2858 |
+
},
|
| 2859 |
+
"total_flos": 2.2087303787769037e+17,
|
| 2860 |
+
"train_batch_size": 1,
|
| 2861 |
+
"trial_name": null,
|
| 2862 |
+
"trial_params": null
|
| 2863 |
+
}
|
training_loss.png
ADDED
|