End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1617 -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_globalbatchsize64_lr1e5_epochs15
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# 1k_globalbatchsize64_lr1e5_epochs15
|
| 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_globalbatchsize64_lr1e5_epochs15
|
|
|
|
| 16 |
|
| 17 |
# 1k_globalbatchsize64_lr1e5_epochs15
|
| 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": 14.512,
|
| 3 |
+
"total_flos": 2.8529054747459584e+17,
|
| 4 |
+
"train_loss": 0.2718963330735763,
|
| 5 |
+
"train_runtime": 15518.1123,
|
| 6 |
+
"train_samples_per_second": 0.963,
|
| 7 |
+
"train_steps_per_second": 0.014
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 14.512,
|
| 3 |
+
"total_flos": 2.8529054747459584e+17,
|
| 4 |
+
"train_loss": 0.2718963330735763,
|
| 5 |
+
"train_runtime": 15518.1123,
|
| 6 |
+
"train_samples_per_second": 0.963,
|
| 7 |
+
"train_steps_per_second": 0.014
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1617 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 14.512,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 225,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.064,
|
| 13 |
+
"grad_norm": 5.8525362996043855,
|
| 14 |
+
"learning_rate": 4.347826086956522e-07,
|
| 15 |
+
"loss": 0.8383,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.128,
|
| 20 |
+
"grad_norm": 5.78756104300265,
|
| 21 |
+
"learning_rate": 8.695652173913044e-07,
|
| 22 |
+
"loss": 0.8521,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.192,
|
| 27 |
+
"grad_norm": 6.106904385373379,
|
| 28 |
+
"learning_rate": 1.3043478260869566e-06,
|
| 29 |
+
"loss": 0.8881,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.256,
|
| 34 |
+
"grad_norm": 5.5965900291075785,
|
| 35 |
+
"learning_rate": 1.7391304347826088e-06,
|
| 36 |
+
"loss": 0.8679,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.32,
|
| 41 |
+
"grad_norm": 5.526905298693686,
|
| 42 |
+
"learning_rate": 2.173913043478261e-06,
|
| 43 |
+
"loss": 0.8416,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.384,
|
| 48 |
+
"grad_norm": 4.500679022660513,
|
| 49 |
+
"learning_rate": 2.6086956521739132e-06,
|
| 50 |
+
"loss": 0.8584,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.448,
|
| 55 |
+
"grad_norm": 4.159198216640518,
|
| 56 |
+
"learning_rate": 3.043478260869566e-06,
|
| 57 |
+
"loss": 0.8066,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.512,
|
| 62 |
+
"grad_norm": 2.4769232699018717,
|
| 63 |
+
"learning_rate": 3.4782608695652175e-06,
|
| 64 |
+
"loss": 0.7736,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.576,
|
| 69 |
+
"grad_norm": 2.116014463402348,
|
| 70 |
+
"learning_rate": 3.91304347826087e-06,
|
| 71 |
+
"loss": 0.7852,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.64,
|
| 76 |
+
"grad_norm": 2.214216511445746,
|
| 77 |
+
"learning_rate": 4.347826086956522e-06,
|
| 78 |
+
"loss": 0.7497,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.704,
|
| 83 |
+
"grad_norm": 3.790997320488787,
|
| 84 |
+
"learning_rate": 4.782608695652174e-06,
|
| 85 |
+
"loss": 0.7672,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.768,
|
| 90 |
+
"grad_norm": 3.936669321722588,
|
| 91 |
+
"learning_rate": 5.2173913043478265e-06,
|
| 92 |
+
"loss": 0.7865,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.832,
|
| 97 |
+
"grad_norm": 3.7396699815921246,
|
| 98 |
+
"learning_rate": 5.652173913043479e-06,
|
| 99 |
+
"loss": 0.7996,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.896,
|
| 104 |
+
"grad_norm": 3.036836408025432,
|
| 105 |
+
"learning_rate": 6.086956521739132e-06,
|
| 106 |
+
"loss": 0.7315,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.96,
|
| 111 |
+
"grad_norm": 2.8367167768931303,
|
| 112 |
+
"learning_rate": 6.521739130434783e-06,
|
| 113 |
+
"loss": 0.7506,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.032,
|
| 118 |
+
"grad_norm": 3.2349565861328564,
|
| 119 |
+
"learning_rate": 6.956521739130435e-06,
|
| 120 |
+
"loss": 1.0599,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.096,
|
| 125 |
+
"grad_norm": 1.8549782722583077,
|
| 126 |
+
"learning_rate": 7.391304347826087e-06,
|
| 127 |
+
"loss": 0.8191,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.16,
|
| 132 |
+
"grad_norm": 1.2762952780873533,
|
| 133 |
+
"learning_rate": 7.82608695652174e-06,
|
| 134 |
+
"loss": 0.5772,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.224,
|
| 139 |
+
"grad_norm": 1.6737161766597488,
|
| 140 |
+
"learning_rate": 8.260869565217392e-06,
|
| 141 |
+
"loss": 0.6765,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.288,
|
| 146 |
+
"grad_norm": 1.5425559295299904,
|
| 147 |
+
"learning_rate": 8.695652173913044e-06,
|
| 148 |
+
"loss": 0.5882,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3519999999999999,
|
| 153 |
+
"grad_norm": 1.2701117211476693,
|
| 154 |
+
"learning_rate": 9.130434782608697e-06,
|
| 155 |
+
"loss": 0.6686,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.416,
|
| 160 |
+
"grad_norm": 1.2765644424519609,
|
| 161 |
+
"learning_rate": 9.565217391304349e-06,
|
| 162 |
+
"loss": 0.6173,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.48,
|
| 167 |
+
"grad_norm": 1.2248520686566564,
|
| 168 |
+
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 0.638,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.544,
|
| 174 |
+
"grad_norm": 1.0819719666916685,
|
| 175 |
+
"learning_rate": 9.999395316300748e-06,
|
| 176 |
+
"loss": 0.6091,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.608,
|
| 181 |
+
"grad_norm": 0.840984819919341,
|
| 182 |
+
"learning_rate": 9.99758141145994e-06,
|
| 183 |
+
"loss": 0.6371,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.6720000000000002,
|
| 188 |
+
"grad_norm": 0.9667721691842549,
|
| 189 |
+
"learning_rate": 9.994558724213056e-06,
|
| 190 |
+
"loss": 0.6428,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.736,
|
| 195 |
+
"grad_norm": 0.9129372264025077,
|
| 196 |
+
"learning_rate": 9.990327985667972e-06,
|
| 197 |
+
"loss": 0.5747,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.8,
|
| 202 |
+
"grad_norm": 0.791251397704499,
|
| 203 |
+
"learning_rate": 9.984890219128148e-06,
|
| 204 |
+
"loss": 0.5929,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.8639999999999999,
|
| 209 |
+
"grad_norm": 0.725342437147292,
|
| 210 |
+
"learning_rate": 9.978246739845095e-06,
|
| 211 |
+
"loss": 0.56,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.928,
|
| 216 |
+
"grad_norm": 0.9118360709457779,
|
| 217 |
+
"learning_rate": 9.970399154700264e-06,
|
| 218 |
+
"loss": 0.639,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.992,
|
| 223 |
+
"grad_norm": 1.0960461815714346,
|
| 224 |
+
"learning_rate": 9.961349361816384e-06,
|
| 225 |
+
"loss": 0.8514,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 2.064,
|
| 230 |
+
"grad_norm": 0.6813317893924205,
|
| 231 |
+
"learning_rate": 9.951099550098349e-06,
|
| 232 |
+
"loss": 0.5712,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.128,
|
| 237 |
+
"grad_norm": 0.6992958835941918,
|
| 238 |
+
"learning_rate": 9.939652198703785e-06,
|
| 239 |
+
"loss": 0.5159,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.192,
|
| 244 |
+
"grad_norm": 0.6667297050374282,
|
| 245 |
+
"learning_rate": 9.927010076443408e-06,
|
| 246 |
+
"loss": 0.5421,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 2.2560000000000002,
|
| 251 |
+
"grad_norm": 0.6988276331072514,
|
| 252 |
+
"learning_rate": 9.91317624111132e-06,
|
| 253 |
+
"loss": 0.5384,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.32,
|
| 258 |
+
"grad_norm": 0.7291358835041466,
|
| 259 |
+
"learning_rate": 9.898154038745408e-06,
|
| 260 |
+
"loss": 0.5578,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 2.384,
|
| 265 |
+
"grad_norm": 0.598999165443441,
|
| 266 |
+
"learning_rate": 9.881947102818036e-06,
|
| 267 |
+
"loss": 0.5103,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 2.448,
|
| 272 |
+
"grad_norm": 0.6840103036845726,
|
| 273 |
+
"learning_rate": 9.864559353357189e-06,
|
| 274 |
+
"loss": 0.5089,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 2.512,
|
| 279 |
+
"grad_norm": 0.6218637615275643,
|
| 280 |
+
"learning_rate": 9.845994995998332e-06,
|
| 281 |
+
"loss": 0.528,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.576,
|
| 286 |
+
"grad_norm": 0.6324736562360972,
|
| 287 |
+
"learning_rate": 9.826258520967178e-06,
|
| 288 |
+
"loss": 0.5366,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 2.64,
|
| 293 |
+
"grad_norm": 0.5714469193284164,
|
| 294 |
+
"learning_rate": 9.805354701993624e-06,
|
| 295 |
+
"loss": 0.5052,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.7039999999999997,
|
| 300 |
+
"grad_norm": 0.65303671408835,
|
| 301 |
+
"learning_rate": 9.7832885951571e-06,
|
| 302 |
+
"loss": 0.5307,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.768,
|
| 307 |
+
"grad_norm": 0.6815553933974848,
|
| 308 |
+
"learning_rate": 9.76006553766365e-06,
|
| 309 |
+
"loss": 0.5241,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 2.832,
|
| 314 |
+
"grad_norm": 0.5097125654552822,
|
| 315 |
+
"learning_rate": 9.735691146555002e-06,
|
| 316 |
+
"loss": 0.4761,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 2.896,
|
| 321 |
+
"grad_norm": 0.6045426300352323,
|
| 322 |
+
"learning_rate": 9.710171317349946e-06,
|
| 323 |
+
"loss": 0.5201,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.96,
|
| 328 |
+
"grad_norm": 0.5810731655280414,
|
| 329 |
+
"learning_rate": 9.683512222618376e-06,
|
| 330 |
+
"loss": 0.517,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 3.032,
|
| 335 |
+
"grad_norm": 0.941893597361931,
|
| 336 |
+
"learning_rate": 9.655720310488298e-06,
|
| 337 |
+
"loss": 0.7744,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.096,
|
| 342 |
+
"grad_norm": 0.5905258794913175,
|
| 343 |
+
"learning_rate": 9.62680230308621e-06,
|
| 344 |
+
"loss": 0.4769,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.16,
|
| 349 |
+
"grad_norm": 0.560120831840573,
|
| 350 |
+
"learning_rate": 9.596765194911182e-06,
|
| 351 |
+
"loss": 0.4663,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.224,
|
| 356 |
+
"grad_norm": 0.5011109468684818,
|
| 357 |
+
"learning_rate": 9.565616251143094e-06,
|
| 358 |
+
"loss": 0.3826,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.288,
|
| 363 |
+
"grad_norm": 0.6169125540763694,
|
| 364 |
+
"learning_rate": 9.533363005885362e-06,
|
| 365 |
+
"loss": 0.461,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 3.352,
|
| 370 |
+
"grad_norm": 0.6458751940569347,
|
| 371 |
+
"learning_rate": 9.50001326034265e-06,
|
| 372 |
+
"loss": 0.5085,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 3.416,
|
| 377 |
+
"grad_norm": 0.5236545564010824,
|
| 378 |
+
"learning_rate": 9.465575080933959e-06,
|
| 379 |
+
"loss": 0.3958,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.48,
|
| 384 |
+
"grad_norm": 0.6553956855287305,
|
| 385 |
+
"learning_rate": 9.430056797341574e-06,
|
| 386 |
+
"loss": 0.4642,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.544,
|
| 391 |
+
"grad_norm": 0.5306401782781227,
|
| 392 |
+
"learning_rate": 9.393467000496345e-06,
|
| 393 |
+
"loss": 0.3736,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 3.608,
|
| 398 |
+
"grad_norm": 0.6647419471112458,
|
| 399 |
+
"learning_rate": 9.355814540499753e-06,
|
| 400 |
+
"loss": 0.4966,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 3.672,
|
| 405 |
+
"grad_norm": 0.5420260014280127,
|
| 406 |
+
"learning_rate": 9.317108524483319e-06,
|
| 407 |
+
"loss": 0.4403,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 3.7359999999999998,
|
| 412 |
+
"grad_norm": 0.5668006984594732,
|
| 413 |
+
"learning_rate": 9.27735831440582e-06,
|
| 414 |
+
"loss": 0.4455,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 3.8,
|
| 419 |
+
"grad_norm": 0.470404315418926,
|
| 420 |
+
"learning_rate": 9.236573524788888e-06,
|
| 421 |
+
"loss": 0.3822,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 3.864,
|
| 426 |
+
"grad_norm": 0.5606512875533857,
|
| 427 |
+
"learning_rate": 9.194764020391507e-06,
|
| 428 |
+
"loss": 0.4383,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 3.928,
|
| 433 |
+
"grad_norm": 0.5389880440196645,
|
| 434 |
+
"learning_rate": 9.151939913823988e-06,
|
| 435 |
+
"loss": 0.4323,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 3.992,
|
| 440 |
+
"grad_norm": 0.9073757207521366,
|
| 441 |
+
"learning_rate": 9.108111563102005e-06,
|
| 442 |
+
"loss": 0.6958,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 4.064,
|
| 447 |
+
"grad_norm": 0.5602168732153348,
|
| 448 |
+
"learning_rate": 9.063289569141251e-06,
|
| 449 |
+
"loss": 0.388,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 4.128,
|
| 454 |
+
"grad_norm": 0.5107919391688228,
|
| 455 |
+
"learning_rate": 9.01748477319338e-06,
|
| 456 |
+
"loss": 0.3468,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 4.192,
|
| 461 |
+
"grad_norm": 0.6908446690586275,
|
| 462 |
+
"learning_rate": 8.970708254223768e-06,
|
| 463 |
+
"loss": 0.3901,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.256,
|
| 468 |
+
"grad_norm": 0.6466553589233165,
|
| 469 |
+
"learning_rate": 8.92297132623183e-06,
|
| 470 |
+
"loss": 0.3976,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 4.32,
|
| 475 |
+
"grad_norm": 0.5384514065787759,
|
| 476 |
+
"learning_rate": 8.87428553551445e-06,
|
| 477 |
+
"loss": 0.3453,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 4.384,
|
| 482 |
+
"grad_norm": 0.7285530566357619,
|
| 483 |
+
"learning_rate": 8.82466265787324e-06,
|
| 484 |
+
"loss": 0.3806,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 4.448,
|
| 489 |
+
"grad_norm": 0.5989140388661336,
|
| 490 |
+
"learning_rate": 8.774114695766286e-06,
|
| 491 |
+
"loss": 0.3495,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 4.5120000000000005,
|
| 496 |
+
"grad_norm": 0.518052599813886,
|
| 497 |
+
"learning_rate": 8.722653875405077e-06,
|
| 498 |
+
"loss": 0.3834,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 4.576,
|
| 503 |
+
"grad_norm": 0.5911105791052353,
|
| 504 |
+
"learning_rate": 8.670292643797302e-06,
|
| 505 |
+
"loss": 0.3399,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.64,
|
| 510 |
+
"grad_norm": 0.523965041751479,
|
| 511 |
+
"learning_rate": 8.61704366573625e-06,
|
| 512 |
+
"loss": 0.336,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 4.704,
|
| 517 |
+
"grad_norm": 0.5073392919848528,
|
| 518 |
+
"learning_rate": 8.562919820737537e-06,
|
| 519 |
+
"loss": 0.3711,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 4.768,
|
| 524 |
+
"grad_norm": 0.5212206834988787,
|
| 525 |
+
"learning_rate": 8.507934199923884e-06,
|
| 526 |
+
"loss": 0.3845,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 4.832,
|
| 531 |
+
"grad_norm": 0.5358602297829164,
|
| 532 |
+
"learning_rate": 8.452100102858734e-06,
|
| 533 |
+
"loss": 0.3671,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 4.896,
|
| 538 |
+
"grad_norm": 0.5277619872937459,
|
| 539 |
+
"learning_rate": 8.395431034329431e-06,
|
| 540 |
+
"loss": 0.3497,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 4.96,
|
| 545 |
+
"grad_norm": 0.4893041623137342,
|
| 546 |
+
"learning_rate": 8.33794070108077e-06,
|
| 547 |
+
"loss": 0.3533,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 5.032,
|
| 552 |
+
"grad_norm": 0.8475136266880684,
|
| 553 |
+
"learning_rate": 8.2796430084997e-06,
|
| 554 |
+
"loss": 0.5205,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 5.096,
|
| 559 |
+
"grad_norm": 0.6640736066588103,
|
| 560 |
+
"learning_rate": 8.22055205725199e-06,
|
| 561 |
+
"loss": 0.3104,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.16,
|
| 566 |
+
"grad_norm": 0.6602837679700819,
|
| 567 |
+
"learning_rate": 8.160682139871634e-06,
|
| 568 |
+
"loss": 0.3119,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.224,
|
| 573 |
+
"grad_norm": 0.6993492447341442,
|
| 574 |
+
"learning_rate": 8.100047737303877e-06,
|
| 575 |
+
"loss": 0.2667,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.288,
|
| 580 |
+
"grad_norm": 0.563086970013438,
|
| 581 |
+
"learning_rate": 8.038663515402659e-06,
|
| 582 |
+
"loss": 0.3049,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.352,
|
| 587 |
+
"grad_norm": 0.5876716648914958,
|
| 588 |
+
"learning_rate": 7.97654432138333e-06,
|
| 589 |
+
"loss": 0.2798,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.416,
|
| 594 |
+
"grad_norm": 0.6946128016542766,
|
| 595 |
+
"learning_rate": 7.913705180231505e-06,
|
| 596 |
+
"loss": 0.3216,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.48,
|
| 601 |
+
"grad_norm": 0.5436967919059682,
|
| 602 |
+
"learning_rate": 7.850161291068915e-06,
|
| 603 |
+
"loss": 0.313,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.5440000000000005,
|
| 608 |
+
"grad_norm": 0.5761779159798073,
|
| 609 |
+
"learning_rate": 7.785928023477142e-06,
|
| 610 |
+
"loss": 0.2787,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.608,
|
| 615 |
+
"grad_norm": 0.6548088172674036,
|
| 616 |
+
"learning_rate": 7.721020913780137e-06,
|
| 617 |
+
"loss": 0.3001,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.672,
|
| 622 |
+
"grad_norm": 0.5481775493910312,
|
| 623 |
+
"learning_rate": 7.655455661286376e-06,
|
| 624 |
+
"loss": 0.2774,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.736,
|
| 629 |
+
"grad_norm": 0.6453759418273927,
|
| 630 |
+
"learning_rate": 7.589248124491627e-06,
|
| 631 |
+
"loss": 0.2924,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.8,
|
| 636 |
+
"grad_norm": 0.46855533387707315,
|
| 637 |
+
"learning_rate": 7.5224143172432e-06,
|
| 638 |
+
"loss": 0.2459,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.864,
|
| 643 |
+
"grad_norm": 0.6886598300393304,
|
| 644 |
+
"learning_rate": 7.454970404866612e-06,
|
| 645 |
+
"loss": 0.3223,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.928,
|
| 650 |
+
"grad_norm": 0.46350588734415665,
|
| 651 |
+
"learning_rate": 7.386932700255635e-06,
|
| 652 |
+
"loss": 0.2405,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.992,
|
| 657 |
+
"grad_norm": 0.8906679221661152,
|
| 658 |
+
"learning_rate": 7.318317659926637e-06,
|
| 659 |
+
"loss": 0.4174,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.064,
|
| 664 |
+
"grad_norm": 0.8228780536908358,
|
| 665 |
+
"learning_rate": 7.249141880038181e-06,
|
| 666 |
+
"loss": 0.2376,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 6.128,
|
| 671 |
+
"grad_norm": 0.5670056091692153,
|
| 672 |
+
"learning_rate": 7.179422092376856e-06,
|
| 673 |
+
"loss": 0.2362,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.192,
|
| 678 |
+
"grad_norm": 1.082946843826278,
|
| 679 |
+
"learning_rate": 7.109175160310312e-06,
|
| 680 |
+
"loss": 0.2366,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 6.256,
|
| 685 |
+
"grad_norm": 0.7628771926745052,
|
| 686 |
+
"learning_rate": 7.038418074708444e-06,
|
| 687 |
+
"loss": 0.2617,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 6.32,
|
| 692 |
+
"grad_norm": 0.7308783027032895,
|
| 693 |
+
"learning_rate": 6.967167949833763e-06,
|
| 694 |
+
"loss": 0.2445,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 6.384,
|
| 699 |
+
"grad_norm": 0.667213630413206,
|
| 700 |
+
"learning_rate": 6.895442019201898e-06,
|
| 701 |
+
"loss": 0.2108,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 6.448,
|
| 706 |
+
"grad_norm": 0.5972918376979972,
|
| 707 |
+
"learning_rate": 6.8232576314132755e-06,
|
| 708 |
+
"loss": 0.1987,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 6.5120000000000005,
|
| 713 |
+
"grad_norm": 0.6989597587495724,
|
| 714 |
+
"learning_rate": 6.750632245956954e-06,
|
| 715 |
+
"loss": 0.1992,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 6.576,
|
| 720 |
+
"grad_norm": 0.7218139781195827,
|
| 721 |
+
"learning_rate": 6.677583428987625e-06,
|
| 722 |
+
"loss": 0.2202,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 6.64,
|
| 727 |
+
"grad_norm": 0.5623074622362645,
|
| 728 |
+
"learning_rate": 6.6041288490768385e-06,
|
| 729 |
+
"loss": 0.197,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 6.704,
|
| 734 |
+
"grad_norm": 0.6361221979816045,
|
| 735 |
+
"learning_rate": 6.530286272939438e-06,
|
| 736 |
+
"loss": 0.2125,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 6.768,
|
| 741 |
+
"grad_norm": 0.540298382564512,
|
| 742 |
+
"learning_rate": 6.456073561136261e-06,
|
| 743 |
+
"loss": 0.2131,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 6.832,
|
| 748 |
+
"grad_norm": 0.5267380906237015,
|
| 749 |
+
"learning_rate": 6.381508663754152e-06,
|
| 750 |
+
"loss": 0.2321,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 6.896,
|
| 755 |
+
"grad_norm": 0.5163319950443708,
|
| 756 |
+
"learning_rate": 6.306609616064304e-06,
|
| 757 |
+
"loss": 0.1836,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 6.96,
|
| 762 |
+
"grad_norm": 0.5552917485167181,
|
| 763 |
+
"learning_rate": 6.231394534160008e-06,
|
| 764 |
+
"loss": 0.2183,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 7.032,
|
| 769 |
+
"grad_norm": 0.8411413436796561,
|
| 770 |
+
"learning_rate": 6.15588161057485e-06,
|
| 771 |
+
"loss": 0.3007,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 7.096,
|
| 776 |
+
"grad_norm": 0.6224198336880072,
|
| 777 |
+
"learning_rate": 6.080089109882419e-06,
|
| 778 |
+
"loss": 0.1916,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 7.16,
|
| 783 |
+
"grad_norm": 0.5613992960207623,
|
| 784 |
+
"learning_rate": 6.004035364278593e-06,
|
| 785 |
+
"loss": 0.1552,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 7.224,
|
| 790 |
+
"grad_norm": 0.7257342928831365,
|
| 791 |
+
"learning_rate": 5.927738769147467e-06,
|
| 792 |
+
"loss": 0.1973,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 7.288,
|
| 797 |
+
"grad_norm": 0.634891501220631,
|
| 798 |
+
"learning_rate": 5.851217778611994e-06,
|
| 799 |
+
"loss": 0.1648,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 7.352,
|
| 804 |
+
"grad_norm": 0.5586971330331016,
|
| 805 |
+
"learning_rate": 5.774490901070424e-06,
|
| 806 |
+
"loss": 0.1677,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 7.416,
|
| 811 |
+
"grad_norm": 0.5697854712479786,
|
| 812 |
+
"learning_rate": 5.697576694719616e-06,
|
| 813 |
+
"loss": 0.1485,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 7.48,
|
| 818 |
+
"grad_norm": 0.6789664781071677,
|
| 819 |
+
"learning_rate": 5.6204937630662974e-06,
|
| 820 |
+
"loss": 0.18,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 7.5440000000000005,
|
| 825 |
+
"grad_norm": 0.6314613526064877,
|
| 826 |
+
"learning_rate": 5.543260750427373e-06,
|
| 827 |
+
"loss": 0.1633,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 7.608,
|
| 832 |
+
"grad_norm": 0.5852921557795249,
|
| 833 |
+
"learning_rate": 5.465896337420359e-06,
|
| 834 |
+
"loss": 0.1229,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 7.672,
|
| 839 |
+
"grad_norm": 0.7293278515447829,
|
| 840 |
+
"learning_rate": 5.388419236445033e-06,
|
| 841 |
+
"loss": 0.1638,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 7.736,
|
| 846 |
+
"grad_norm": 0.5691211961859317,
|
| 847 |
+
"learning_rate": 5.310848187157404e-06,
|
| 848 |
+
"loss": 0.1446,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 7.8,
|
| 853 |
+
"grad_norm": 0.6779883283192614,
|
| 854 |
+
"learning_rate": 5.233201951937088e-06,
|
| 855 |
+
"loss": 0.198,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 7.864,
|
| 860 |
+
"grad_norm": 0.5169377802565881,
|
| 861 |
+
"learning_rate": 5.155499311349185e-06,
|
| 862 |
+
"loss": 0.1108,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 7.928,
|
| 867 |
+
"grad_norm": 0.617304671774631,
|
| 868 |
+
"learning_rate": 5.077759059601756e-06,
|
| 869 |
+
"loss": 0.1765,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 7.992,
|
| 874 |
+
"grad_norm": 0.8749830751036564,
|
| 875 |
+
"learning_rate": 5e-06,
|
| 876 |
+
"loss": 0.2347,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 8.064,
|
| 881 |
+
"grad_norm": 0.6470842908256255,
|
| 882 |
+
"learning_rate": 4.922240940398246e-06,
|
| 883 |
+
"loss": 0.1311,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 8.128,
|
| 888 |
+
"grad_norm": 0.5493020353400846,
|
| 889 |
+
"learning_rate": 4.844500688650817e-06,
|
| 890 |
+
"loss": 0.1264,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 8.192,
|
| 895 |
+
"grad_norm": 0.6640315711273018,
|
| 896 |
+
"learning_rate": 4.766798048062913e-06,
|
| 897 |
+
"loss": 0.119,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 8.256,
|
| 902 |
+
"grad_norm": 0.5669054561741467,
|
| 903 |
+
"learning_rate": 4.689151812842598e-06,
|
| 904 |
+
"loss": 0.1007,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 8.32,
|
| 909 |
+
"grad_norm": 0.6391216031772399,
|
| 910 |
+
"learning_rate": 4.611580763554969e-06,
|
| 911 |
+
"loss": 0.1269,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 8.384,
|
| 916 |
+
"grad_norm": 0.554949112829902,
|
| 917 |
+
"learning_rate": 4.534103662579643e-06,
|
| 918 |
+
"loss": 0.1308,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 8.448,
|
| 923 |
+
"grad_norm": 0.540198400212999,
|
| 924 |
+
"learning_rate": 4.456739249572628e-06,
|
| 925 |
+
"loss": 0.103,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 8.512,
|
| 930 |
+
"grad_norm": 0.5257662879891519,
|
| 931 |
+
"learning_rate": 4.379506236933703e-06,
|
| 932 |
+
"loss": 0.1138,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 8.576,
|
| 937 |
+
"grad_norm": 0.5103315863694035,
|
| 938 |
+
"learning_rate": 4.3024233052803855e-06,
|
| 939 |
+
"loss": 0.0991,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 8.64,
|
| 944 |
+
"grad_norm": 0.5603180545470561,
|
| 945 |
+
"learning_rate": 4.2255090989295765e-06,
|
| 946 |
+
"loss": 0.1157,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 8.704,
|
| 951 |
+
"grad_norm": 0.5670621886774232,
|
| 952 |
+
"learning_rate": 4.148782221388007e-06,
|
| 953 |
+
"loss": 0.1138,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 8.768,
|
| 958 |
+
"grad_norm": 0.6672700892340953,
|
| 959 |
+
"learning_rate": 4.072261230852534e-06,
|
| 960 |
+
"loss": 0.1052,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 8.832,
|
| 965 |
+
"grad_norm": 0.5388583060063759,
|
| 966 |
+
"learning_rate": 3.995964635721409e-06,
|
| 967 |
+
"loss": 0.1128,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 8.896,
|
| 972 |
+
"grad_norm": 0.5124359276521201,
|
| 973 |
+
"learning_rate": 3.919910890117584e-06,
|
| 974 |
+
"loss": 0.1088,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 8.96,
|
| 979 |
+
"grad_norm": 0.5990552366765427,
|
| 980 |
+
"learning_rate": 3.844118389425154e-06,
|
| 981 |
+
"loss": 0.1135,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 9.032,
|
| 986 |
+
"grad_norm": 0.9034863036538634,
|
| 987 |
+
"learning_rate": 3.768605465839994e-06,
|
| 988 |
+
"loss": 0.1564,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 9.096,
|
| 993 |
+
"grad_norm": 0.5681816752160384,
|
| 994 |
+
"learning_rate": 3.6933903839356983e-06,
|
| 995 |
+
"loss": 0.0872,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 9.16,
|
| 1000 |
+
"grad_norm": 0.5156238573430424,
|
| 1001 |
+
"learning_rate": 3.6184913362458497e-06,
|
| 1002 |
+
"loss": 0.0872,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 9.224,
|
| 1007 |
+
"grad_norm": 0.7141899022135747,
|
| 1008 |
+
"learning_rate": 3.5439264388637407e-06,
|
| 1009 |
+
"loss": 0.0913,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 9.288,
|
| 1014 |
+
"grad_norm": 0.6753480096268889,
|
| 1015 |
+
"learning_rate": 3.469713727060564e-06,
|
| 1016 |
+
"loss": 0.0769,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 9.352,
|
| 1021 |
+
"grad_norm": 0.5142326331753825,
|
| 1022 |
+
"learning_rate": 3.3958711509231627e-06,
|
| 1023 |
+
"loss": 0.0742,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 9.416,
|
| 1028 |
+
"grad_norm": 0.5054090282507924,
|
| 1029 |
+
"learning_rate": 3.322416571012376e-06,
|
| 1030 |
+
"loss": 0.0718,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 9.48,
|
| 1035 |
+
"grad_norm": 0.5380442047155478,
|
| 1036 |
+
"learning_rate": 3.249367754043047e-06,
|
| 1037 |
+
"loss": 0.0732,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 9.544,
|
| 1042 |
+
"grad_norm": 0.5721458944985353,
|
| 1043 |
+
"learning_rate": 3.176742368586725e-06,
|
| 1044 |
+
"loss": 0.0958,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 9.608,
|
| 1049 |
+
"grad_norm": 0.5073658378412854,
|
| 1050 |
+
"learning_rate": 3.104557980798104e-06,
|
| 1051 |
+
"loss": 0.0879,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 9.672,
|
| 1056 |
+
"grad_norm": 0.4740748654264217,
|
| 1057 |
+
"learning_rate": 3.032832050166239e-06,
|
| 1058 |
+
"loss": 0.0611,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 9.736,
|
| 1063 |
+
"grad_norm": 0.6353397407742756,
|
| 1064 |
+
"learning_rate": 2.961581925291557e-06,
|
| 1065 |
+
"loss": 0.0813,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 9.8,
|
| 1070 |
+
"grad_norm": 0.6739552173743931,
|
| 1071 |
+
"learning_rate": 2.8908248396896893e-06,
|
| 1072 |
+
"loss": 0.0993,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 9.864,
|
| 1077 |
+
"grad_norm": 0.44633419135584906,
|
| 1078 |
+
"learning_rate": 2.820577907623145e-06,
|
| 1079 |
+
"loss": 0.076,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 9.928,
|
| 1084 |
+
"grad_norm": 0.5870447232019427,
|
| 1085 |
+
"learning_rate": 2.750858119961821e-06,
|
| 1086 |
+
"loss": 0.0898,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 9.992,
|
| 1091 |
+
"grad_norm": 0.7666025635059504,
|
| 1092 |
+
"learning_rate": 2.6816823400733628e-06,
|
| 1093 |
+
"loss": 0.0884,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 10.064,
|
| 1098 |
+
"grad_norm": 0.5934067797312592,
|
| 1099 |
+
"learning_rate": 2.613067299744364e-06,
|
| 1100 |
+
"loss": 0.0648,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 10.128,
|
| 1105 |
+
"grad_norm": 0.4607975710978408,
|
| 1106 |
+
"learning_rate": 2.5450295951333896e-06,
|
| 1107 |
+
"loss": 0.0604,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 10.192,
|
| 1112 |
+
"grad_norm": 0.5326136669464842,
|
| 1113 |
+
"learning_rate": 2.4775856827568016e-06,
|
| 1114 |
+
"loss": 0.0706,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 10.256,
|
| 1119 |
+
"grad_norm": 0.6026902430383296,
|
| 1120 |
+
"learning_rate": 2.410751875508373e-06,
|
| 1121 |
+
"loss": 0.0583,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 10.32,
|
| 1126 |
+
"grad_norm": 0.6620510675539595,
|
| 1127 |
+
"learning_rate": 2.3445443387136247e-06,
|
| 1128 |
+
"loss": 0.0518,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 10.384,
|
| 1133 |
+
"grad_norm": 0.5790104347262993,
|
| 1134 |
+
"learning_rate": 2.278979086219863e-06,
|
| 1135 |
+
"loss": 0.0722,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 10.448,
|
| 1140 |
+
"grad_norm": 0.44985565104923625,
|
| 1141 |
+
"learning_rate": 2.2140719765228587e-06,
|
| 1142 |
+
"loss": 0.0602,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 10.512,
|
| 1147 |
+
"grad_norm": 0.44282499345716836,
|
| 1148 |
+
"learning_rate": 2.149838708931087e-06,
|
| 1149 |
+
"loss": 0.0652,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 10.576,
|
| 1154 |
+
"grad_norm": 0.5063255073234909,
|
| 1155 |
+
"learning_rate": 2.086294819768496e-06,
|
| 1156 |
+
"loss": 0.0625,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 10.64,
|
| 1161 |
+
"grad_norm": 0.49048633159771426,
|
| 1162 |
+
"learning_rate": 2.0234556786166715e-06,
|
| 1163 |
+
"loss": 0.0538,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 10.704,
|
| 1168 |
+
"grad_norm": 0.44979760143578396,
|
| 1169 |
+
"learning_rate": 1.9613364845973433e-06,
|
| 1170 |
+
"loss": 0.0508,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 10.768,
|
| 1175 |
+
"grad_norm": 0.4652912131686526,
|
| 1176 |
+
"learning_rate": 1.8999522626961254e-06,
|
| 1177 |
+
"loss": 0.0595,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 10.832,
|
| 1182 |
+
"grad_norm": 0.5263420548962754,
|
| 1183 |
+
"learning_rate": 1.8393178601283684e-06,
|
| 1184 |
+
"loss": 0.0501,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 10.896,
|
| 1189 |
+
"grad_norm": 1.0307756225124574,
|
| 1190 |
+
"learning_rate": 1.7794479427480115e-06,
|
| 1191 |
+
"loss": 0.0544,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 10.96,
|
| 1196 |
+
"grad_norm": 0.5363489902168008,
|
| 1197 |
+
"learning_rate": 1.7203569915003005e-06,
|
| 1198 |
+
"loss": 0.059,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 11.032,
|
| 1203 |
+
"grad_norm": 0.7007495479012421,
|
| 1204 |
+
"learning_rate": 1.6620592989192318e-06,
|
| 1205 |
+
"loss": 0.074,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 11.096,
|
| 1210 |
+
"grad_norm": 0.3988528071157963,
|
| 1211 |
+
"learning_rate": 1.6045689656705715e-06,
|
| 1212 |
+
"loss": 0.0463,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 11.16,
|
| 1217 |
+
"grad_norm": 0.4091928665550615,
|
| 1218 |
+
"learning_rate": 1.5478998971412669e-06,
|
| 1219 |
+
"loss": 0.0417,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 11.224,
|
| 1224 |
+
"grad_norm": 0.425702955264231,
|
| 1225 |
+
"learning_rate": 1.4920658000761172e-06,
|
| 1226 |
+
"loss": 0.0417,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 11.288,
|
| 1231 |
+
"grad_norm": 0.38109643866642195,
|
| 1232 |
+
"learning_rate": 1.4370801792624656e-06,
|
| 1233 |
+
"loss": 0.0426,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 11.352,
|
| 1238 |
+
"grad_norm": 0.3686563339080507,
|
| 1239 |
+
"learning_rate": 1.3829563342637514e-06,
|
| 1240 |
+
"loss": 0.0478,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 11.416,
|
| 1245 |
+
"grad_norm": 0.45764182190591657,
|
| 1246 |
+
"learning_rate": 1.3297073562026992e-06,
|
| 1247 |
+
"loss": 0.045,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 11.48,
|
| 1252 |
+
"grad_norm": 0.4199082525098172,
|
| 1253 |
+
"learning_rate": 1.2773461245949249e-06,
|
| 1254 |
+
"loss": 0.039,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 11.544,
|
| 1259 |
+
"grad_norm": 0.45366152342194455,
|
| 1260 |
+
"learning_rate": 1.225885304233716e-06,
|
| 1261 |
+
"loss": 0.0431,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 11.608,
|
| 1266 |
+
"grad_norm": 0.41429138709830393,
|
| 1267 |
+
"learning_rate": 1.1753373421267622e-06,
|
| 1268 |
+
"loss": 0.0404,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 11.672,
|
| 1273 |
+
"grad_norm": 0.40101443835964756,
|
| 1274 |
+
"learning_rate": 1.125714464485551e-06,
|
| 1275 |
+
"loss": 0.0388,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 11.736,
|
| 1280 |
+
"grad_norm": 0.42230918519889976,
|
| 1281 |
+
"learning_rate": 1.0770286737681701e-06,
|
| 1282 |
+
"loss": 0.054,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 11.8,
|
| 1287 |
+
"grad_norm": 0.3955850162131075,
|
| 1288 |
+
"learning_rate": 1.0292917457762325e-06,
|
| 1289 |
+
"loss": 0.0519,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 11.864,
|
| 1294 |
+
"grad_norm": 0.40978210661533565,
|
| 1295 |
+
"learning_rate": 9.825152268066213e-07,
|
| 1296 |
+
"loss": 0.0529,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 11.928,
|
| 1301 |
+
"grad_norm": 0.3224405712930817,
|
| 1302 |
+
"learning_rate": 9.367104308587493e-07,
|
| 1303 |
+
"loss": 0.0343,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 11.992,
|
| 1308 |
+
"grad_norm": 0.6175724512403094,
|
| 1309 |
+
"learning_rate": 8.918884368979969e-07,
|
| 1310 |
+
"loss": 0.0761,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 12.064,
|
| 1315 |
+
"grad_norm": 0.3711252656480644,
|
| 1316 |
+
"learning_rate": 8.480600861760124e-07,
|
| 1317 |
+
"loss": 0.0395,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 12.128,
|
| 1322 |
+
"grad_norm": 0.3434399208970678,
|
| 1323 |
+
"learning_rate": 8.052359796084952e-07,
|
| 1324 |
+
"loss": 0.0384,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 12.192,
|
| 1329 |
+
"grad_norm": 0.3292608708161041,
|
| 1330 |
+
"learning_rate": 7.634264752111131e-07,
|
| 1331 |
+
"loss": 0.0385,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 12.256,
|
| 1336 |
+
"grad_norm": 0.32211799742825886,
|
| 1337 |
+
"learning_rate": 7.226416855941814e-07,
|
| 1338 |
+
"loss": 0.0357,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 12.32,
|
| 1343 |
+
"grad_norm": 0.32874209805701876,
|
| 1344 |
+
"learning_rate": 6.828914755166826e-07,
|
| 1345 |
+
"loss": 0.04,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 12.384,
|
| 1350 |
+
"grad_norm": 0.33126665641384606,
|
| 1351 |
+
"learning_rate": 6.441854595002478e-07,
|
| 1352 |
+
"loss": 0.0409,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 12.448,
|
| 1357 |
+
"grad_norm": 0.39660571210862894,
|
| 1358 |
+
"learning_rate": 6.065329995036573e-07,
|
| 1359 |
+
"loss": 0.0497,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 12.512,
|
| 1364 |
+
"grad_norm": 0.3840128969876996,
|
| 1365 |
+
"learning_rate": 5.699432026584267e-07,
|
| 1366 |
+
"loss": 0.0427,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 12.576,
|
| 1371 |
+
"grad_norm": 0.31928908520778054,
|
| 1372 |
+
"learning_rate": 5.344249190660427e-07,
|
| 1373 |
+
"loss": 0.0357,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 12.64,
|
| 1378 |
+
"grad_norm": 0.3944602562734707,
|
| 1379 |
+
"learning_rate": 4.999867396573499e-07,
|
| 1380 |
+
"loss": 0.0393,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 12.704,
|
| 1385 |
+
"grad_norm": 0.28793860178813374,
|
| 1386 |
+
"learning_rate": 4.666369941146376e-07,
|
| 1387 |
+
"loss": 0.0298,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 12.768,
|
| 1392 |
+
"grad_norm": 0.31930718863037527,
|
| 1393 |
+
"learning_rate": 4.343837488569058e-07,
|
| 1394 |
+
"loss": 0.0314,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 12.832,
|
| 1399 |
+
"grad_norm": 0.31395657161087953,
|
| 1400 |
+
"learning_rate": 4.03234805088818e-07,
|
| 1401 |
+
"loss": 0.0414,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 12.896,
|
| 1406 |
+
"grad_norm": 0.31679679068395816,
|
| 1407 |
+
"learning_rate": 3.7319769691379295e-07,
|
| 1408 |
+
"loss": 0.037,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 12.96,
|
| 1413 |
+
"grad_norm": 0.3174012256341073,
|
| 1414 |
+
"learning_rate": 3.4427968951170287e-07,
|
| 1415 |
+
"loss": 0.0302,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 13.032,
|
| 1420 |
+
"grad_norm": 0.4021334925947107,
|
| 1421 |
+
"learning_rate": 3.1648777738162496e-07,
|
| 1422 |
+
"loss": 0.0436,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 13.096,
|
| 1427 |
+
"grad_norm": 0.3276274321874788,
|
| 1428 |
+
"learning_rate": 2.8982868265005457e-07,
|
| 1429 |
+
"loss": 0.0432,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 13.16,
|
| 1434 |
+
"grad_norm": 0.2577333590211676,
|
| 1435 |
+
"learning_rate": 2.6430885344499944e-07,
|
| 1436 |
+
"loss": 0.0305,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 13.224,
|
| 1441 |
+
"grad_norm": 0.30312172258469733,
|
| 1442 |
+
"learning_rate": 2.399344623363503e-07,
|
| 1443 |
+
"loss": 0.038,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 13.288,
|
| 1448 |
+
"grad_norm": 0.22933906016596675,
|
| 1449 |
+
"learning_rate": 2.1671140484290144e-07,
|
| 1450 |
+
"loss": 0.0267,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 13.352,
|
| 1455 |
+
"grad_norm": 0.3417291551593501,
|
| 1456 |
+
"learning_rate": 1.9464529800637731e-07,
|
| 1457 |
+
"loss": 0.0457,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 13.416,
|
| 1462 |
+
"grad_norm": 0.22979896331556754,
|
| 1463 |
+
"learning_rate": 1.737414790328218e-07,
|
| 1464 |
+
"loss": 0.0234,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 13.48,
|
| 1469 |
+
"grad_norm": 0.23858046783478587,
|
| 1470 |
+
"learning_rate": 1.540050040016694e-07,
|
| 1471 |
+
"loss": 0.0276,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 13.544,
|
| 1476 |
+
"grad_norm": 0.3057108346535968,
|
| 1477 |
+
"learning_rate": 1.3544064664281266e-07,
|
| 1478 |
+
"loss": 0.0428,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 13.608,
|
| 1483 |
+
"grad_norm": 0.26201506253974,
|
| 1484 |
+
"learning_rate": 1.1805289718196499e-07,
|
| 1485 |
+
"loss": 0.03,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 13.672,
|
| 1490 |
+
"grad_norm": 0.26571881324755525,
|
| 1491 |
+
"learning_rate": 1.0184596125459134e-07,
|
| 1492 |
+
"loss": 0.0332,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 13.736,
|
| 1497 |
+
"grad_norm": 0.29404129035702864,
|
| 1498 |
+
"learning_rate": 8.682375888868167e-08,
|
| 1499 |
+
"loss": 0.0366,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 13.8,
|
| 1504 |
+
"grad_norm": 0.2798615470284527,
|
| 1505 |
+
"learning_rate": 7.29899235565934e-08,
|
| 1506 |
+
"loss": 0.0332,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 13.864,
|
| 1511 |
+
"grad_norm": 0.29092215438733693,
|
| 1512 |
+
"learning_rate": 6.034780129621664e-08,
|
| 1513 |
+
"loss": 0.0378,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 13.928,
|
| 1518 |
+
"grad_norm": 0.2915112874691917,
|
| 1519 |
+
"learning_rate": 4.8900449901653214e-08,
|
| 1520 |
+
"loss": 0.0371,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 13.992,
|
| 1525 |
+
"grad_norm": 0.433301617483451,
|
| 1526 |
+
"learning_rate": 3.8650638183617695e-08,
|
| 1527 |
+
"loss": 0.0411,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 14.064,
|
| 1532 |
+
"grad_norm": 0.2619925470967437,
|
| 1533 |
+
"learning_rate": 2.960084529973706e-08,
|
| 1534 |
+
"loss": 0.0338,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 14.128,
|
| 1539 |
+
"grad_norm": 0.26602808197291455,
|
| 1540 |
+
"learning_rate": 2.1753260154906973e-08,
|
| 1541 |
+
"loss": 0.0365,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 14.192,
|
| 1546 |
+
"grad_norm": 0.2669352376491801,
|
| 1547 |
+
"learning_rate": 1.5109780871853663e-08,
|
| 1548 |
+
"loss": 0.0362,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 14.256,
|
| 1553 |
+
"grad_norm": 0.27612825414292785,
|
| 1554 |
+
"learning_rate": 9.672014332028357e-09,
|
| 1555 |
+
"loss": 0.0336,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 14.32,
|
| 1560 |
+
"grad_norm": 0.2912032340081594,
|
| 1561 |
+
"learning_rate": 5.4412757869459765e-09,
|
| 1562 |
+
"loss": 0.0393,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 14.384,
|
| 1567 |
+
"grad_norm": 0.2705430971735193,
|
| 1568 |
+
"learning_rate": 2.4185885400596076e-09,
|
| 1569 |
+
"loss": 0.0366,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 14.448,
|
| 1574 |
+
"grad_norm": 0.36779741246292924,
|
| 1575 |
+
"learning_rate": 6.04683699252373e-10,
|
| 1576 |
+
"loss": 0.0282,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 14.512,
|
| 1581 |
+
"grad_norm": 0.25743877314168934,
|
| 1582 |
+
"learning_rate": 0.0,
|
| 1583 |
+
"loss": 0.0353,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 14.512,
|
| 1588 |
+
"step": 225,
|
| 1589 |
+
"total_flos": 2.8529054747459584e+17,
|
| 1590 |
+
"train_loss": 0.2718963330735763,
|
| 1591 |
+
"train_runtime": 15518.1123,
|
| 1592 |
+
"train_samples_per_second": 0.963,
|
| 1593 |
+
"train_steps_per_second": 0.014
|
| 1594 |
+
}
|
| 1595 |
+
],
|
| 1596 |
+
"logging_steps": 1.0,
|
| 1597 |
+
"max_steps": 225,
|
| 1598 |
+
"num_input_tokens_seen": 0,
|
| 1599 |
+
"num_train_epochs": 15,
|
| 1600 |
+
"save_steps": 500,
|
| 1601 |
+
"stateful_callbacks": {
|
| 1602 |
+
"TrainerControl": {
|
| 1603 |
+
"args": {
|
| 1604 |
+
"should_epoch_stop": false,
|
| 1605 |
+
"should_evaluate": false,
|
| 1606 |
+
"should_log": false,
|
| 1607 |
+
"should_save": true,
|
| 1608 |
+
"should_training_stop": true
|
| 1609 |
+
},
|
| 1610 |
+
"attributes": {}
|
| 1611 |
+
}
|
| 1612 |
+
},
|
| 1613 |
+
"total_flos": 2.8529054747459584e+17,
|
| 1614 |
+
"train_batch_size": 1,
|
| 1615 |
+
"trial_name": null,
|
| 1616 |
+
"trial_params": null
|
| 1617 |
+
}
|
training_loss.png
ADDED
|