MHGanainy/gpt-J-lora-ecthr
Browse files- README.md +2 -0
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1540 -0
README.md
CHANGED
|
@@ -15,6 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
# gpt-J-lora-ecthr
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [EleutherAI/gpt-j-6b](https://huggingface.co/EleutherAI/gpt-j-6b) on an unknown dataset.
|
|
|
|
|
|
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 15 |
# gpt-J-lora-ecthr
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [EleutherAI/gpt-j-6b](https://huggingface.co/EleutherAI/gpt-j-6b) on an unknown dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 1.4851
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 1.485129952430725,
|
| 4 |
+
"eval_runtime": 355.9545,
|
| 5 |
+
"eval_samples_per_second": 11.81,
|
| 6 |
+
"eval_steps_per_second": 5.905,
|
| 7 |
+
"perplexity": 4.415539185546311,
|
| 8 |
+
"total_flos": 3.101747713212678e+18,
|
| 9 |
+
"train_loss": 1.5223615952344096,
|
| 10 |
+
"train_runtime": 9293.396,
|
| 11 |
+
"train_samples_per_second": 4.624,
|
| 12 |
+
"train_steps_per_second": 2.312
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 1.485129952430725,
|
| 4 |
+
"eval_runtime": 355.9545,
|
| 5 |
+
"eval_samples_per_second": 11.81,
|
| 6 |
+
"eval_steps_per_second": 5.905,
|
| 7 |
+
"perplexity": 4.415539185546311
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 3.101747713212678e+18,
|
| 4 |
+
"train_loss": 1.5223615952344096,
|
| 5 |
+
"train_runtime": 9293.396,
|
| 6 |
+
"train_samples_per_second": 4.624,
|
| 7 |
+
"train_steps_per_second": 2.312
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1540 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 21487,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0046539768231954205,
|
| 13 |
+
"grad_norm": 0.2191164344549179,
|
| 14 |
+
"learning_rate": 9.31098696461825e-07,
|
| 15 |
+
"loss": 1.8284,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.009307953646390841,
|
| 20 |
+
"grad_norm": 0.23232442140579224,
|
| 21 |
+
"learning_rate": 1.86219739292365e-06,
|
| 22 |
+
"loss": 1.8324,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.013961930469586261,
|
| 27 |
+
"grad_norm": 0.27771520614624023,
|
| 28 |
+
"learning_rate": 2.793296089385475e-06,
|
| 29 |
+
"loss": 1.8126,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.018615907292781682,
|
| 34 |
+
"grad_norm": 0.46955519914627075,
|
| 35 |
+
"learning_rate": 3.7243947858473e-06,
|
| 36 |
+
"loss": 1.7894,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.023269884115977104,
|
| 41 |
+
"grad_norm": 0.4999992847442627,
|
| 42 |
+
"learning_rate": 4.655493482309125e-06,
|
| 43 |
+
"loss": 1.7438,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.027923860939172523,
|
| 48 |
+
"grad_norm": 0.6398751139640808,
|
| 49 |
+
"learning_rate": 5.58659217877095e-06,
|
| 50 |
+
"loss": 1.6969,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.032577837762367945,
|
| 55 |
+
"grad_norm": 0.7858927249908447,
|
| 56 |
+
"learning_rate": 6.517690875232775e-06,
|
| 57 |
+
"loss": 1.7084,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.037231814585563364,
|
| 62 |
+
"grad_norm": 0.7117326259613037,
|
| 63 |
+
"learning_rate": 7.4487895716946e-06,
|
| 64 |
+
"loss": 1.709,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.04188579140875878,
|
| 69 |
+
"grad_norm": 0.7050628066062927,
|
| 70 |
+
"learning_rate": 8.379888268156426e-06,
|
| 71 |
+
"loss": 1.6655,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.04653976823195421,
|
| 76 |
+
"grad_norm": 0.9369891285896301,
|
| 77 |
+
"learning_rate": 9.31098696461825e-06,
|
| 78 |
+
"loss": 1.6765,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.05119374505514963,
|
| 83 |
+
"grad_norm": 0.7342795729637146,
|
| 84 |
+
"learning_rate": 1.0242085661080074e-05,
|
| 85 |
+
"loss": 1.6398,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.055847721878345045,
|
| 90 |
+
"grad_norm": 1.5446929931640625,
|
| 91 |
+
"learning_rate": 1.11731843575419e-05,
|
| 92 |
+
"loss": 1.6714,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.060501698701540464,
|
| 97 |
+
"grad_norm": 0.8926479816436768,
|
| 98 |
+
"learning_rate": 1.2104283054003724e-05,
|
| 99 |
+
"loss": 1.6355,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.06515567552473589,
|
| 104 |
+
"grad_norm": 0.9157938361167908,
|
| 105 |
+
"learning_rate": 1.303538175046555e-05,
|
| 106 |
+
"loss": 1.5984,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.06980965234793131,
|
| 111 |
+
"grad_norm": 0.9163572192192078,
|
| 112 |
+
"learning_rate": 1.3966480446927374e-05,
|
| 113 |
+
"loss": 1.6372,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.07446362917112673,
|
| 118 |
+
"grad_norm": 1.3793407678604126,
|
| 119 |
+
"learning_rate": 1.48975791433892e-05,
|
| 120 |
+
"loss": 1.6025,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.07911760599432215,
|
| 125 |
+
"grad_norm": 1.2203880548477173,
|
| 126 |
+
"learning_rate": 1.5828677839851026e-05,
|
| 127 |
+
"loss": 1.6359,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08377158281751756,
|
| 132 |
+
"grad_norm": 1.18509840965271,
|
| 133 |
+
"learning_rate": 1.6759776536312852e-05,
|
| 134 |
+
"loss": 1.6273,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.08842555964071298,
|
| 139 |
+
"grad_norm": 1.1019757986068726,
|
| 140 |
+
"learning_rate": 1.7690875232774675e-05,
|
| 141 |
+
"loss": 1.6286,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.09307953646390842,
|
| 146 |
+
"grad_norm": 0.6310858726501465,
|
| 147 |
+
"learning_rate": 1.86219739292365e-05,
|
| 148 |
+
"loss": 1.6132,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.09773351328710383,
|
| 153 |
+
"grad_norm": 1.2102677822113037,
|
| 154 |
+
"learning_rate": 1.9553072625698326e-05,
|
| 155 |
+
"loss": 1.6125,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.10238749011029925,
|
| 160 |
+
"grad_norm": 1.16068696975708,
|
| 161 |
+
"learning_rate": 1.999964321562415e-05,
|
| 162 |
+
"loss": 1.618,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.10704146693349467,
|
| 167 |
+
"grad_norm": 1.1368638277053833,
|
| 168 |
+
"learning_rate": 1.999695163593931e-05,
|
| 169 |
+
"loss": 1.5991,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.11169544375669009,
|
| 174 |
+
"grad_norm": 0.9747415781021118,
|
| 175 |
+
"learning_rate": 1.9991621965335746e-05,
|
| 176 |
+
"loss": 1.5878,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.11634942057988551,
|
| 181 |
+
"grad_norm": 0.9197840690612793,
|
| 182 |
+
"learning_rate": 1.998365561025775e-05,
|
| 183 |
+
"loss": 1.6189,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.12100339740308093,
|
| 188 |
+
"grad_norm": 0.9155735373497009,
|
| 189 |
+
"learning_rate": 1.997305467294307e-05,
|
| 190 |
+
"loss": 1.587,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.12565737422627635,
|
| 195 |
+
"grad_norm": 0.7722734212875366,
|
| 196 |
+
"learning_rate": 1.9959821950868117e-05,
|
| 197 |
+
"loss": 1.5681,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.13031135104947178,
|
| 202 |
+
"grad_norm": 1.044062852859497,
|
| 203 |
+
"learning_rate": 1.994396093600977e-05,
|
| 204 |
+
"loss": 1.5845,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.13496532787266718,
|
| 209 |
+
"grad_norm": 0.7750725150108337,
|
| 210 |
+
"learning_rate": 1.9925475813923858e-05,
|
| 211 |
+
"loss": 1.5936,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.13961930469586262,
|
| 216 |
+
"grad_norm": 0.9063745141029358,
|
| 217 |
+
"learning_rate": 1.990437146264066e-05,
|
| 218 |
+
"loss": 1.5788,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.14427328151905802,
|
| 223 |
+
"grad_norm": 0.655881404876709,
|
| 224 |
+
"learning_rate": 1.988065345137762e-05,
|
| 225 |
+
"loss": 1.5927,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.14892725834225345,
|
| 230 |
+
"grad_norm": 1.1087632179260254,
|
| 231 |
+
"learning_rate": 1.98543280390697e-05,
|
| 232 |
+
"loss": 1.5607,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.1535812351654489,
|
| 237 |
+
"grad_norm": 0.7348114848136902,
|
| 238 |
+
"learning_rate": 1.982540217271772e-05,
|
| 239 |
+
"loss": 1.5344,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.1582352119886443,
|
| 244 |
+
"grad_norm": 0.8793938755989075,
|
| 245 |
+
"learning_rate": 1.979388348555509e-05,
|
| 246 |
+
"loss": 1.5576,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.16288918881183972,
|
| 251 |
+
"grad_norm": 0.9496509432792664,
|
| 252 |
+
"learning_rate": 1.975978029503352e-05,
|
| 253 |
+
"loss": 1.5439,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.16754316563503513,
|
| 258 |
+
"grad_norm": 0.9296020269393921,
|
| 259 |
+
"learning_rate": 1.9723101600628095e-05,
|
| 260 |
+
"loss": 1.6044,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.17219714245823056,
|
| 265 |
+
"grad_norm": 0.7520443797111511,
|
| 266 |
+
"learning_rate": 1.9683857081462423e-05,
|
| 267 |
+
"loss": 1.5319,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.17685111928142597,
|
| 272 |
+
"grad_norm": 0.9139896631240845,
|
| 273 |
+
"learning_rate": 1.9642057093754426e-05,
|
| 274 |
+
"loss": 1.5457,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.1815050961046214,
|
| 279 |
+
"grad_norm": 0.781461775302887,
|
| 280 |
+
"learning_rate": 1.9597712668083413e-05,
|
| 281 |
+
"loss": 1.5604,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.18615907292781683,
|
| 286 |
+
"grad_norm": 0.9205677509307861,
|
| 287 |
+
"learning_rate": 1.9550835506479266e-05,
|
| 288 |
+
"loss": 1.5255,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.19081304975101224,
|
| 293 |
+
"grad_norm": 0.7341404557228088,
|
| 294 |
+
"learning_rate": 1.9501437979334358e-05,
|
| 295 |
+
"loss": 1.5876,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.19546702657420767,
|
| 300 |
+
"grad_norm": 0.8709468245506287,
|
| 301 |
+
"learning_rate": 1.9449533122139157e-05,
|
| 302 |
+
"loss": 1.5469,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.20012100339740307,
|
| 307 |
+
"grad_norm": 0.9768949151039124,
|
| 308 |
+
"learning_rate": 1.939513463204229e-05,
|
| 309 |
+
"loss": 1.5417,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.2047749802205985,
|
| 314 |
+
"grad_norm": 0.7858554720878601,
|
| 315 |
+
"learning_rate": 1.9338256864236024e-05,
|
| 316 |
+
"loss": 1.5672,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.2094289570437939,
|
| 321 |
+
"grad_norm": 0.7643204927444458,
|
| 322 |
+
"learning_rate": 1.927891482816805e-05,
|
| 323 |
+
"loss": 1.5032,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.21408293386698934,
|
| 328 |
+
"grad_norm": 1.2041771411895752,
|
| 329 |
+
"learning_rate": 1.9217124183580674e-05,
|
| 330 |
+
"loss": 1.51,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.21873691069018475,
|
| 335 |
+
"grad_norm": 0.9453679323196411,
|
| 336 |
+
"learning_rate": 1.9152901236378364e-05,
|
| 337 |
+
"loss": 1.5662,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.22339088751338018,
|
| 342 |
+
"grad_norm": 0.6478219628334045,
|
| 343 |
+
"learning_rate": 1.908626293432479e-05,
|
| 344 |
+
"loss": 1.5932,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.22804486433657561,
|
| 349 |
+
"grad_norm": 0.838474690914154,
|
| 350 |
+
"learning_rate": 1.901722686257051e-05,
|
| 351 |
+
"loss": 1.567,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.23269884115977102,
|
| 356 |
+
"grad_norm": 1.1010560989379883,
|
| 357 |
+
"learning_rate": 1.894581123901238e-05,
|
| 358 |
+
"loss": 1.4849,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.23735281798296645,
|
| 363 |
+
"grad_norm": 0.8271942734718323,
|
| 364 |
+
"learning_rate": 1.88720349094861e-05,
|
| 365 |
+
"loss": 1.56,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.24200679480616186,
|
| 370 |
+
"grad_norm": 0.9592469334602356,
|
| 371 |
+
"learning_rate": 1.8795917342792967e-05,
|
| 372 |
+
"loss": 1.6055,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.2466607716293573,
|
| 377 |
+
"grad_norm": 0.7423321008682251,
|
| 378 |
+
"learning_rate": 1.8717478625562256e-05,
|
| 379 |
+
"loss": 1.5268,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2513147484525527,
|
| 384 |
+
"grad_norm": 1.0170488357543945,
|
| 385 |
+
"learning_rate": 1.8636739456950593e-05,
|
| 386 |
+
"loss": 1.5765,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.2559687252757481,
|
| 391 |
+
"grad_norm": 0.7664517760276794,
|
| 392 |
+
"learning_rate": 1.8553721143179626e-05,
|
| 393 |
+
"loss": 1.5625,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.26062270209894356,
|
| 398 |
+
"grad_norm": 0.712568461894989,
|
| 399 |
+
"learning_rate": 1.8468445591913587e-05,
|
| 400 |
+
"loss": 1.5853,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.26527667892213896,
|
| 405 |
+
"grad_norm": 0.9137941002845764,
|
| 406 |
+
"learning_rate": 1.8380935306478036e-05,
|
| 407 |
+
"loss": 1.5628,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.26993065574533437,
|
| 412 |
+
"grad_norm": 0.9478745460510254,
|
| 413 |
+
"learning_rate": 1.829121337992151e-05,
|
| 414 |
+
"loss": 1.4895,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.27458463256852983,
|
| 419 |
+
"grad_norm": 0.5866535305976868,
|
| 420 |
+
"learning_rate": 1.8199303488921493e-05,
|
| 421 |
+
"loss": 1.5484,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.27923860939172523,
|
| 426 |
+
"grad_norm": 0.9393543004989624,
|
| 427 |
+
"learning_rate": 1.810522988753638e-05,
|
| 428 |
+
"loss": 1.5307,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.28389258621492064,
|
| 433 |
+
"grad_norm": 0.6128756403923035,
|
| 434 |
+
"learning_rate": 1.8009017400805094e-05,
|
| 435 |
+
"loss": 1.5821,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.28854656303811604,
|
| 440 |
+
"grad_norm": 0.8191483020782471,
|
| 441 |
+
"learning_rate": 1.7910691418196022e-05,
|
| 442 |
+
"loss": 1.5206,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.2932005398613115,
|
| 447 |
+
"grad_norm": 0.8590126633644104,
|
| 448 |
+
"learning_rate": 1.7810277886906998e-05,
|
| 449 |
+
"loss": 1.5488,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.2978545166845069,
|
| 454 |
+
"grad_norm": 0.6708762645721436,
|
| 455 |
+
"learning_rate": 1.7707803305018097e-05,
|
| 456 |
+
"loss": 1.5383,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.3025084935077023,
|
| 461 |
+
"grad_norm": 0.6897620558738708,
|
| 462 |
+
"learning_rate": 1.760329471449909e-05,
|
| 463 |
+
"loss": 1.5497,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.3071624703308978,
|
| 468 |
+
"grad_norm": 0.7875939607620239,
|
| 469 |
+
"learning_rate": 1.7496779694073333e-05,
|
| 470 |
+
"loss": 1.5251,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.3118164471540932,
|
| 475 |
+
"grad_norm": 0.9739288091659546,
|
| 476 |
+
"learning_rate": 1.738828635194003e-05,
|
| 477 |
+
"loss": 1.5235,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.3164704239772886,
|
| 482 |
+
"grad_norm": 0.9243332743644714,
|
| 483 |
+
"learning_rate": 1.7277843318356798e-05,
|
| 484 |
+
"loss": 1.5256,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.321124400800484,
|
| 489 |
+
"grad_norm": 1.0979468822479248,
|
| 490 |
+
"learning_rate": 1.7165479738084405e-05,
|
| 491 |
+
"loss": 1.5841,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.32577837762367945,
|
| 496 |
+
"grad_norm": 0.931398868560791,
|
| 497 |
+
"learning_rate": 1.7051225262695812e-05,
|
| 498 |
+
"loss": 1.5408,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.33043235444687485,
|
| 503 |
+
"grad_norm": 0.7612720131874084,
|
| 504 |
+
"learning_rate": 1.693511004275139e-05,
|
| 505 |
+
"loss": 1.4606,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.33508633127007026,
|
| 510 |
+
"grad_norm": 1.0624606609344482,
|
| 511 |
+
"learning_rate": 1.6817164719842548e-05,
|
| 512 |
+
"loss": 1.5322,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.3397403080932657,
|
| 517 |
+
"grad_norm": 0.6526899933815002,
|
| 518 |
+
"learning_rate": 1.66974204185057e-05,
|
| 519 |
+
"loss": 1.5801,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.3443942849164611,
|
| 524 |
+
"grad_norm": 0.7007707953453064,
|
| 525 |
+
"learning_rate": 1.6575908738008847e-05,
|
| 526 |
+
"loss": 1.4867,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.34904826173965653,
|
| 531 |
+
"grad_norm": 0.8433778882026672,
|
| 532 |
+
"learning_rate": 1.645266174401285e-05,
|
| 533 |
+
"loss": 1.5193,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.35370223856285193,
|
| 538 |
+
"grad_norm": 1.0169028043746948,
|
| 539 |
+
"learning_rate": 1.632771196010965e-05,
|
| 540 |
+
"loss": 1.53,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.3583562153860474,
|
| 545 |
+
"grad_norm": 0.9545369744300842,
|
| 546 |
+
"learning_rate": 1.6201092359239624e-05,
|
| 547 |
+
"loss": 1.4862,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.3630101922092428,
|
| 552 |
+
"grad_norm": 1.2505319118499756,
|
| 553 |
+
"learning_rate": 1.607283635499036e-05,
|
| 554 |
+
"loss": 1.5349,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.3676641690324382,
|
| 559 |
+
"grad_norm": 0.8437191843986511,
|
| 560 |
+
"learning_rate": 1.5942977792779193e-05,
|
| 561 |
+
"loss": 1.5245,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.37231814585563366,
|
| 566 |
+
"grad_norm": 0.8392757773399353,
|
| 567 |
+
"learning_rate": 1.581155094092171e-05,
|
| 568 |
+
"loss": 1.519,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.37697212267882907,
|
| 573 |
+
"grad_norm": 0.7851161956787109,
|
| 574 |
+
"learning_rate": 1.5678590481588743e-05,
|
| 575 |
+
"loss": 1.5301,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.3816260995020245,
|
| 580 |
+
"grad_norm": 0.8554267883300781,
|
| 581 |
+
"learning_rate": 1.5544131501654078e-05,
|
| 582 |
+
"loss": 1.5143,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.3862800763252199,
|
| 587 |
+
"grad_norm": 0.9319170117378235,
|
| 588 |
+
"learning_rate": 1.5408209483435402e-05,
|
| 589 |
+
"loss": 1.5079,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.39093405314841534,
|
| 594 |
+
"grad_norm": 0.9828823804855347,
|
| 595 |
+
"learning_rate": 1.5270860295330912e-05,
|
| 596 |
+
"loss": 1.5248,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.39558802997161074,
|
| 601 |
+
"grad_norm": 0.8951559662818909,
|
| 602 |
+
"learning_rate": 1.5132120182353984e-05,
|
| 603 |
+
"loss": 1.4728,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.40024200679480615,
|
| 608 |
+
"grad_norm": 0.8046073317527771,
|
| 609 |
+
"learning_rate": 1.499202575656852e-05,
|
| 610 |
+
"loss": 1.5096,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.4048959836180016,
|
| 615 |
+
"grad_norm": 0.7216867208480835,
|
| 616 |
+
"learning_rate": 1.4850613987427398e-05,
|
| 617 |
+
"loss": 1.538,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.409549960441197,
|
| 622 |
+
"grad_norm": 0.5751676559448242,
|
| 623 |
+
"learning_rate": 1.470792219201662e-05,
|
| 624 |
+
"loss": 1.508,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.4142039372643924,
|
| 629 |
+
"grad_norm": 0.7514901161193848,
|
| 630 |
+
"learning_rate": 1.4563988025207726e-05,
|
| 631 |
+
"loss": 1.5365,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.4188579140875878,
|
| 636 |
+
"grad_norm": 0.8872545957565308,
|
| 637 |
+
"learning_rate": 1.4418849469721073e-05,
|
| 638 |
+
"loss": 1.4705,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.4235118909107833,
|
| 643 |
+
"grad_norm": 0.772324800491333,
|
| 644 |
+
"learning_rate": 1.4272544826102576e-05,
|
| 645 |
+
"loss": 1.521,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.4281658677339787,
|
| 650 |
+
"grad_norm": 0.8108826279640198,
|
| 651 |
+
"learning_rate": 1.4125112702616607e-05,
|
| 652 |
+
"loss": 1.4958,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.4328198445571741,
|
| 657 |
+
"grad_norm": 0.8969531655311584,
|
| 658 |
+
"learning_rate": 1.3976592005057684e-05,
|
| 659 |
+
"loss": 1.4596,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.4374738213803695,
|
| 664 |
+
"grad_norm": 0.93816739320755,
|
| 665 |
+
"learning_rate": 1.3827021926483622e-05,
|
| 666 |
+
"loss": 1.5404,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.44212779820356496,
|
| 671 |
+
"grad_norm": 0.5850194096565247,
|
| 672 |
+
"learning_rate": 1.3676441936872904e-05,
|
| 673 |
+
"loss": 1.5342,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.44678177502676036,
|
| 678 |
+
"grad_norm": 0.903218686580658,
|
| 679 |
+
"learning_rate": 1.352489177270898e-05,
|
| 680 |
+
"loss": 1.5315,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.45143575184995577,
|
| 685 |
+
"grad_norm": 0.9102717041969299,
|
| 686 |
+
"learning_rate": 1.337241142649419e-05,
|
| 687 |
+
"loss": 1.5105,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.45608972867315123,
|
| 692 |
+
"grad_norm": 0.5675022006034851,
|
| 693 |
+
"learning_rate": 1.3219041136196202e-05,
|
| 694 |
+
"loss": 1.465,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.46074370549634663,
|
| 699 |
+
"grad_norm": 1.007855772972107,
|
| 700 |
+
"learning_rate": 1.3064821374629606e-05,
|
| 701 |
+
"loss": 1.5131,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.46539768231954204,
|
| 706 |
+
"grad_norm": 0.8553173542022705,
|
| 707 |
+
"learning_rate": 1.2909792838775571e-05,
|
| 708 |
+
"loss": 1.5054,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.47005165914273744,
|
| 713 |
+
"grad_norm": 0.9074007868766785,
|
| 714 |
+
"learning_rate": 1.2755558069133832e-05,
|
| 715 |
+
"loss": 1.5582,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.4747056359659329,
|
| 720 |
+
"grad_norm": 0.7873587608337402,
|
| 721 |
+
"learning_rate": 1.2599041981920994e-05,
|
| 722 |
+
"loss": 1.5109,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.4793596127891283,
|
| 727 |
+
"grad_norm": 0.9822585582733154,
|
| 728 |
+
"learning_rate": 1.2441840034728076e-05,
|
| 729 |
+
"loss": 1.5067,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.4840135896123237,
|
| 734 |
+
"grad_norm": 1.0133826732635498,
|
| 735 |
+
"learning_rate": 1.2283993711503824e-05,
|
| 736 |
+
"loss": 1.4886,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.4886675664355192,
|
| 741 |
+
"grad_norm": 1.1799577474594116,
|
| 742 |
+
"learning_rate": 1.2125544666241078e-05,
|
| 743 |
+
"loss": 1.4655,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.4933215432587146,
|
| 748 |
+
"grad_norm": 0.927507221698761,
|
| 749 |
+
"learning_rate": 1.1966534711984717e-05,
|
| 750 |
+
"loss": 1.4748,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.49797552008191,
|
| 755 |
+
"grad_norm": 0.8527470827102661,
|
| 756 |
+
"learning_rate": 1.1808603529490986e-05,
|
| 757 |
+
"loss": 1.4607,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.5026294969051054,
|
| 762 |
+
"grad_norm": 0.8991233706474304,
|
| 763 |
+
"learning_rate": 1.1648602337072668e-05,
|
| 764 |
+
"loss": 1.5432,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.5072834737283008,
|
| 769 |
+
"grad_norm": 0.7587746381759644,
|
| 770 |
+
"learning_rate": 1.1488166095750207e-05,
|
| 771 |
+
"loss": 1.5053,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.5119374505514962,
|
| 776 |
+
"grad_norm": 0.6698244214057922,
|
| 777 |
+
"learning_rate": 1.132733714296872e-05,
|
| 778 |
+
"loss": 1.4281,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5165914273746917,
|
| 783 |
+
"grad_norm": 0.848193883895874,
|
| 784 |
+
"learning_rate": 1.1166157919805772e-05,
|
| 785 |
+
"loss": 1.5043,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.5212454041978871,
|
| 790 |
+
"grad_norm": 1.0735975503921509,
|
| 791 |
+
"learning_rate": 1.1004670959771624e-05,
|
| 792 |
+
"loss": 1.5072,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.5258993810210825,
|
| 797 |
+
"grad_norm": 1.1171404123306274,
|
| 798 |
+
"learning_rate": 1.0842918877585058e-05,
|
| 799 |
+
"loss": 1.4864,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.5305533578442779,
|
| 804 |
+
"grad_norm": 0.8309624195098877,
|
| 805 |
+
"learning_rate": 1.0680944357927836e-05,
|
| 806 |
+
"loss": 1.5373,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.5352073346674734,
|
| 811 |
+
"grad_norm": 0.8913981318473816,
|
| 812 |
+
"learning_rate": 1.0518790144180635e-05,
|
| 813 |
+
"loss": 1.4432,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.5398613114906687,
|
| 818 |
+
"grad_norm": 0.8174418807029724,
|
| 819 |
+
"learning_rate": 1.0356499027143519e-05,
|
| 820 |
+
"loss": 1.4852,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.5445152883138642,
|
| 825 |
+
"grad_norm": 0.9098989367485046,
|
| 826 |
+
"learning_rate": 1.0194113833743897e-05,
|
| 827 |
+
"loss": 1.5405,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.5491692651370597,
|
| 832 |
+
"grad_norm": 0.983168363571167,
|
| 833 |
+
"learning_rate": 1.0031677415734946e-05,
|
| 834 |
+
"loss": 1.5014,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.553823241960255,
|
| 839 |
+
"grad_norm": 0.793428361415863,
|
| 840 |
+
"learning_rate": 9.869232638387492e-06,
|
| 841 |
+
"loss": 1.5047,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.5584772187834505,
|
| 846 |
+
"grad_norm": 0.9047326445579529,
|
| 847 |
+
"learning_rate": 9.706822369178289e-06,
|
| 848 |
+
"loss": 1.4991,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.5631311956066459,
|
| 853 |
+
"grad_norm": 0.825706422328949,
|
| 854 |
+
"learning_rate": 9.54448946647775e-06,
|
| 855 |
+
"loss": 1.5175,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.5677851724298413,
|
| 860 |
+
"grad_norm": 0.9917203187942505,
|
| 861 |
+
"learning_rate": 9.382276768240104e-06,
|
| 862 |
+
"loss": 1.4916,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.5724391492530367,
|
| 867 |
+
"grad_norm": 0.8516975045204163,
|
| 868 |
+
"learning_rate": 9.220227080698847e-06,
|
| 869 |
+
"loss": 1.5003,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.5770931260762321,
|
| 874 |
+
"grad_norm": 0.7885426878929138,
|
| 875 |
+
"learning_rate": 9.058383167070676e-06,
|
| 876 |
+
"loss": 1.5427,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.5817471028994275,
|
| 881 |
+
"grad_norm": 0.9626107215881348,
|
| 882 |
+
"learning_rate": 8.89678773627071e-06,
|
| 883 |
+
"loss": 1.4672,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.586401079722623,
|
| 888 |
+
"grad_norm": 1.023768663406372,
|
| 889 |
+
"learning_rate": 8.735483431642033e-06,
|
| 890 |
+
"loss": 1.5236,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.5910550565458184,
|
| 895 |
+
"grad_norm": 0.9998337626457214,
|
| 896 |
+
"learning_rate": 8.574512819702594e-06,
|
| 897 |
+
"loss": 1.4932,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.5957090333690138,
|
| 902 |
+
"grad_norm": 0.9876322746276855,
|
| 903 |
+
"learning_rate": 8.413918378912325e-06,
|
| 904 |
+
"loss": 1.4849,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.6003630101922093,
|
| 909 |
+
"grad_norm": 0.9000487923622131,
|
| 910 |
+
"learning_rate": 8.253742488463531e-06,
|
| 911 |
+
"loss": 1.4532,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.6050169870154046,
|
| 916 |
+
"grad_norm": 0.9810521006584167,
|
| 917 |
+
"learning_rate": 8.094027417097468e-06,
|
| 918 |
+
"loss": 1.4818,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.6096709638386001,
|
| 923 |
+
"grad_norm": 0.8786018490791321,
|
| 924 |
+
"learning_rate": 7.936404805248885e-06,
|
| 925 |
+
"loss": 1.524,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.6143249406617955,
|
| 930 |
+
"grad_norm": 0.8514164686203003,
|
| 931 |
+
"learning_rate": 7.777732023413952e-06,
|
| 932 |
+
"loss": 1.4763,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.6189789174849909,
|
| 937 |
+
"grad_norm": 0.603767991065979,
|
| 938 |
+
"learning_rate": 7.619645674842026e-06,
|
| 939 |
+
"loss": 1.4977,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.6236328943081864,
|
| 944 |
+
"grad_norm": 0.8721168637275696,
|
| 945 |
+
"learning_rate": 7.462187476866129e-06,
|
| 946 |
+
"loss": 1.4339,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.6282868711313818,
|
| 951 |
+
"grad_norm": 0.8393872380256653,
|
| 952 |
+
"learning_rate": 7.305398981056899e-06,
|
| 953 |
+
"loss": 1.4888,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.6329408479545772,
|
| 958 |
+
"grad_norm": 1.0900713205337524,
|
| 959 |
+
"learning_rate": 7.149321562257598e-06,
|
| 960 |
+
"loss": 1.4612,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.6375948247777726,
|
| 965 |
+
"grad_norm": 0.857639491558075,
|
| 966 |
+
"learning_rate": 6.9939964076657e-06,
|
| 967 |
+
"loss": 1.4866,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.642248801600968,
|
| 972 |
+
"grad_norm": 0.7077500820159912,
|
| 973 |
+
"learning_rate": 6.839464505964017e-06,
|
| 974 |
+
"loss": 1.4587,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.6469027784241634,
|
| 979 |
+
"grad_norm": 0.9254826307296753,
|
| 980 |
+
"learning_rate": 6.6857666365042354e-06,
|
| 981 |
+
"loss": 1.463,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.6515567552473589,
|
| 986 |
+
"grad_norm": 0.8199195265769958,
|
| 987 |
+
"learning_rate": 6.5329433585456625e-06,
|
| 988 |
+
"loss": 1.5347,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.6562107320705542,
|
| 993 |
+
"grad_norm": 1.0524635314941406,
|
| 994 |
+
"learning_rate": 6.381035000552058e-06,
|
| 995 |
+
"loss": 1.5253,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.6608647088937497,
|
| 1000 |
+
"grad_norm": 0.9598236083984375,
|
| 1001 |
+
"learning_rate": 6.230081649549371e-06,
|
| 1002 |
+
"loss": 1.5064,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.6655186857169452,
|
| 1007 |
+
"grad_norm": 0.8717284202575684,
|
| 1008 |
+
"learning_rate": 6.080123140547213e-06,
|
| 1009 |
+
"loss": 1.5284,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.6701726625401405,
|
| 1014 |
+
"grad_norm": 0.7019008994102478,
|
| 1015 |
+
"learning_rate": 5.931199046026791e-06,
|
| 1016 |
+
"loss": 1.5057,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.674826639363336,
|
| 1021 |
+
"grad_norm": 0.7632008790969849,
|
| 1022 |
+
"learning_rate": 5.78334866549816e-06,
|
| 1023 |
+
"loss": 1.4382,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.6794806161865314,
|
| 1028 |
+
"grad_norm": 0.9148026704788208,
|
| 1029 |
+
"learning_rate": 5.63661101512949e-06,
|
| 1030 |
+
"loss": 1.4656,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.6841345930097268,
|
| 1035 |
+
"grad_norm": 1.0867414474487305,
|
| 1036 |
+
"learning_rate": 5.491024817451108e-06,
|
| 1037 |
+
"loss": 1.4611,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.6887885698329222,
|
| 1042 |
+
"grad_norm": 0.9941554665565491,
|
| 1043 |
+
"learning_rate": 5.346628491137016e-06,
|
| 1044 |
+
"loss": 1.471,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.6934425466561176,
|
| 1049 |
+
"grad_norm": 0.756461501121521,
|
| 1050 |
+
"learning_rate": 5.203460140866625e-06,
|
| 1051 |
+
"loss": 1.4691,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.6980965234793131,
|
| 1056 |
+
"grad_norm": 0.9088025093078613,
|
| 1057 |
+
"learning_rate": 5.061557547269322e-06,
|
| 1058 |
+
"loss": 1.4712,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.7027505003025085,
|
| 1063 |
+
"grad_norm": 0.850143313407898,
|
| 1064 |
+
"learning_rate": 4.920958156954544e-06,
|
| 1065 |
+
"loss": 1.4434,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.7074044771257039,
|
| 1070 |
+
"grad_norm": 1.072135090827942,
|
| 1071 |
+
"learning_rate": 4.781699072630018e-06,
|
| 1072 |
+
"loss": 1.5095,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.7120584539488993,
|
| 1077 |
+
"grad_norm": 0.8175047039985657,
|
| 1078 |
+
"learning_rate": 4.643817043310759e-06,
|
| 1079 |
+
"loss": 1.455,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.7167124307720948,
|
| 1084 |
+
"grad_norm": 0.8658547401428223,
|
| 1085 |
+
"learning_rate": 4.5073484546213705e-06,
|
| 1086 |
+
"loss": 1.4545,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.7213664075952901,
|
| 1091 |
+
"grad_norm": 0.9122436046600342,
|
| 1092 |
+
"learning_rate": 4.372329319194262e-06,
|
| 1093 |
+
"loss": 1.5157,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.7260203844184856,
|
| 1098 |
+
"grad_norm": 1.0552470684051514,
|
| 1099 |
+
"learning_rate": 4.238795267166298e-06,
|
| 1100 |
+
"loss": 1.4664,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.7306743612416811,
|
| 1105 |
+
"grad_norm": 0.8493475914001465,
|
| 1106 |
+
"learning_rate": 4.106781536776383e-06,
|
| 1107 |
+
"loss": 1.5032,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.7353283380648764,
|
| 1112 |
+
"grad_norm": 0.9446743130683899,
|
| 1113 |
+
"learning_rate": 3.976322965066436e-06,
|
| 1114 |
+
"loss": 1.4747,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.7399823148880719,
|
| 1119 |
+
"grad_norm": 0.7051513195037842,
|
| 1120 |
+
"learning_rate": 3.847453978688287e-06,
|
| 1121 |
+
"loss": 1.4649,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.7446362917112673,
|
| 1126 |
+
"grad_norm": 0.941286563873291,
|
| 1127 |
+
"learning_rate": 3.720208584818833e-06,
|
| 1128 |
+
"loss": 1.4755,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.7492902685344627,
|
| 1133 |
+
"grad_norm": 0.9868441820144653,
|
| 1134 |
+
"learning_rate": 3.5946203621858867e-06,
|
| 1135 |
+
"loss": 1.5239,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.7539442453576581,
|
| 1140 |
+
"grad_norm": 1.0158665180206299,
|
| 1141 |
+
"learning_rate": 3.4707224522071294e-06,
|
| 1142 |
+
"loss": 1.4625,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.7585982221808535,
|
| 1147 |
+
"grad_norm": 1.0703004598617554,
|
| 1148 |
+
"learning_rate": 3.3485475502444055e-06,
|
| 1149 |
+
"loss": 1.4627,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.763252199004049,
|
| 1154 |
+
"grad_norm": 0.9671617746353149,
|
| 1155 |
+
"learning_rate": 3.2281278969757913e-06,
|
| 1156 |
+
"loss": 1.4777,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.7679061758272444,
|
| 1161 |
+
"grad_norm": 1.0113691091537476,
|
| 1162 |
+
"learning_rate": 3.109495269887587e-06,
|
| 1163 |
+
"loss": 1.4954,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.7725601526504398,
|
| 1168 |
+
"grad_norm": 1.0288777351379395,
|
| 1169 |
+
"learning_rate": 2.9926809748885964e-06,
|
| 1170 |
+
"loss": 1.5105,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.7772141294736352,
|
| 1175 |
+
"grad_norm": 0.851613461971283,
|
| 1176 |
+
"learning_rate": 2.877715838048811e-06,
|
| 1177 |
+
"loss": 1.4307,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.7818681062968307,
|
| 1182 |
+
"grad_norm": 0.7523232102394104,
|
| 1183 |
+
"learning_rate": 2.764630197464729e-06,
|
| 1184 |
+
"loss": 1.4429,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.786522083120026,
|
| 1189 |
+
"grad_norm": 1.0177912712097168,
|
| 1190 |
+
"learning_rate": 2.653453895253466e-06,
|
| 1191 |
+
"loss": 1.4585,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.7911760599432215,
|
| 1196 |
+
"grad_norm": 0.7985556721687317,
|
| 1197 |
+
"learning_rate": 2.5442162696777095e-06,
|
| 1198 |
+
"loss": 1.4912,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.795830036766417,
|
| 1203 |
+
"grad_norm": 1.0856115818023682,
|
| 1204 |
+
"learning_rate": 2.4369461474036825e-06,
|
| 1205 |
+
"loss": 1.4824,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.8004840135896123,
|
| 1210 |
+
"grad_norm": 1.0469813346862793,
|
| 1211 |
+
"learning_rate": 2.332714608087351e-06,
|
| 1212 |
+
"loss": 1.4442,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.8051379904128078,
|
| 1217 |
+
"grad_norm": 0.7078351378440857,
|
| 1218 |
+
"learning_rate": 2.2294435164482266e-06,
|
| 1219 |
+
"loss": 1.5291,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.8097919672360032,
|
| 1224 |
+
"grad_norm": 0.9295824766159058,
|
| 1225 |
+
"learning_rate": 2.1282229933463116e-06,
|
| 1226 |
+
"loss": 1.4083,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.8144459440591986,
|
| 1231 |
+
"grad_norm": 0.7632860541343689,
|
| 1232 |
+
"learning_rate": 2.0290797498183858e-06,
|
| 1233 |
+
"loss": 1.4367,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.819099920882394,
|
| 1238 |
+
"grad_norm": 0.9528563022613525,
|
| 1239 |
+
"learning_rate": 1.9320399487288832e-06,
|
| 1240 |
+
"loss": 1.445,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.8237538977055894,
|
| 1245 |
+
"grad_norm": 0.8745253682136536,
|
| 1246 |
+
"learning_rate": 1.8371291978657857e-06,
|
| 1247 |
+
"loss": 1.4694,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.8284078745287848,
|
| 1252 |
+
"grad_norm": 0.9276142120361328,
|
| 1253 |
+
"learning_rate": 1.7443725431830127e-06,
|
| 1254 |
+
"loss": 1.4682,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.8330618513519803,
|
| 1259 |
+
"grad_norm": 0.8589171171188354,
|
| 1260 |
+
"learning_rate": 1.6537944621910307e-06,
|
| 1261 |
+
"loss": 1.4999,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.8377158281751756,
|
| 1266 |
+
"grad_norm": 0.8758667707443237,
|
| 1267 |
+
"learning_rate": 1.5654188574975215e-06,
|
| 1268 |
+
"loss": 1.4621,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.8423698049983711,
|
| 1273 |
+
"grad_norm": 0.9397442936897278,
|
| 1274 |
+
"learning_rate": 1.4792690504997087e-06,
|
| 1275 |
+
"loss": 1.4598,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.8470237818215666,
|
| 1280 |
+
"grad_norm": 1.0151734352111816,
|
| 1281 |
+
"learning_rate": 1.3953677752301032e-06,
|
| 1282 |
+
"loss": 1.4698,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.8516777586447619,
|
| 1287 |
+
"grad_norm": 0.9862604737281799,
|
| 1288 |
+
"learning_rate": 1.3137371723572257e-06,
|
| 1289 |
+
"loss": 1.4478,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.8563317354679574,
|
| 1294 |
+
"grad_norm": 1.16762113571167,
|
| 1295 |
+
"learning_rate": 1.2343987833429062e-06,
|
| 1296 |
+
"loss": 1.4806,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.8609857122911528,
|
| 1301 |
+
"grad_norm": 1.06341552734375,
|
| 1302 |
+
"learning_rate": 1.1573735447577327e-06,
|
| 1303 |
+
"loss": 1.4555,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.8656396891143482,
|
| 1308 |
+
"grad_norm": 1.0032464265823364,
|
| 1309 |
+
"learning_rate": 1.082681782756102e-06,
|
| 1310 |
+
"loss": 1.4471,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.8702936659375436,
|
| 1315 |
+
"grad_norm": 0.9144641160964966,
|
| 1316 |
+
"learning_rate": 1.0103432077123532e-06,
|
| 1317 |
+
"loss": 1.4721,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.874947642760739,
|
| 1322 |
+
"grad_norm": 1.0318922996520996,
|
| 1323 |
+
"learning_rate": 9.403769090194192e-07,
|
| 1324 |
+
"loss": 1.4542,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.8796016195839345,
|
| 1329 |
+
"grad_norm": 1.0590611696243286,
|
| 1330 |
+
"learning_rate": 8.72801350051331e-07,
|
| 1331 |
+
"loss": 1.497,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.8842555964071299,
|
| 1336 |
+
"grad_norm": 0.5952686667442322,
|
| 1337 |
+
"learning_rate": 8.076343632909211e-07,
|
| 1338 |
+
"loss": 1.4613,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.8889095732303253,
|
| 1343 |
+
"grad_norm": 0.8993837237358093,
|
| 1344 |
+
"learning_rate": 7.448931456240205e-07,
|
| 1345 |
+
"loss": 1.5027,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.8935635500535207,
|
| 1350 |
+
"grad_norm": 0.8999748229980469,
|
| 1351 |
+
"learning_rate": 6.845942538013894e-07,
|
| 1352 |
+
"loss": 1.4677,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.8982175268767162,
|
| 1357 |
+
"grad_norm": 1.1852108240127563,
|
| 1358 |
+
"learning_rate": 6.273197876706216e-07,
|
| 1359 |
+
"loss": 1.4537,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.9028715036999115,
|
| 1364 |
+
"grad_norm": 0.8120535016059875,
|
| 1365 |
+
"learning_rate": 5.71927827141906e-07,
|
| 1366 |
+
"loss": 1.4965,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.907525480523107,
|
| 1371 |
+
"grad_norm": 0.8027576804161072,
|
| 1372 |
+
"learning_rate": 5.190238361949229e-07,
|
| 1373 |
+
"loss": 1.5109,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.9121794573463025,
|
| 1378 |
+
"grad_norm": 1.2424169778823853,
|
| 1379 |
+
"learning_rate": 4.686217756392464e-07,
|
| 1380 |
+
"loss": 1.4614,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.9168334341694978,
|
| 1385 |
+
"grad_norm": 1.248057246208191,
|
| 1386 |
+
"learning_rate": 4.207349460512022e-07,
|
| 1387 |
+
"loss": 1.4786,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.9214874109926933,
|
| 1392 |
+
"grad_norm": 0.8463609218597412,
|
| 1393 |
+
"learning_rate": 3.7537598426397925e-07,
|
| 1394 |
+
"loss": 1.4714,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.9261413878158887,
|
| 1399 |
+
"grad_norm": 0.6483246684074402,
|
| 1400 |
+
"learning_rate": 3.325568600329143e-07,
|
| 1401 |
+
"loss": 1.5209,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.9307953646390841,
|
| 1406 |
+
"grad_norm": 0.6988236904144287,
|
| 1407 |
+
"learning_rate": 2.922888728767903e-07,
|
| 1408 |
+
"loss": 1.485,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.9354493414622795,
|
| 1413 |
+
"grad_norm": 0.9654485583305359,
|
| 1414 |
+
"learning_rate": 2.545826490960146e-07,
|
| 1415 |
+
"loss": 1.4532,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.9401033182854749,
|
| 1420 |
+
"grad_norm": 1.0685368776321411,
|
| 1421 |
+
"learning_rate": 2.194481389684433e-07,
|
| 1422 |
+
"loss": 1.4925,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.9447572951086703,
|
| 1427 |
+
"grad_norm": 1.161594033241272,
|
| 1428 |
+
"learning_rate": 1.8689461412362164e-07,
|
| 1429 |
+
"loss": 1.4376,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.9494112719318658,
|
| 1434 |
+
"grad_norm": 1.1302499771118164,
|
| 1435 |
+
"learning_rate": 1.5693066509608333e-07,
|
| 1436 |
+
"loss": 1.4776,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.9540652487550612,
|
| 1441 |
+
"grad_norm": 1.0777382850646973,
|
| 1442 |
+
"learning_rate": 1.295641990584007e-07,
|
| 1443 |
+
"loss": 1.4702,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.9587192255782566,
|
| 1448 |
+
"grad_norm": 1.027956247329712,
|
| 1449 |
+
"learning_rate": 1.0503714003311738e-07,
|
| 1450 |
+
"loss": 1.4699,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.9633732024014521,
|
| 1455 |
+
"grad_norm": 0.7537710666656494,
|
| 1456 |
+
"learning_rate": 8.286047531811392e-08,
|
| 1457 |
+
"loss": 1.489,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.9680271792246474,
|
| 1462 |
+
"grad_norm": 0.7544810175895691,
|
| 1463 |
+
"learning_rate": 6.33008399409063e-08,
|
| 1464 |
+
"loss": 1.4563,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.9726811560478429,
|
| 1469 |
+
"grad_norm": 0.9680439829826355,
|
| 1470 |
+
"learning_rate": 4.636339548457991e-08,
|
| 1471 |
+
"loss": 1.4345,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.9773351328710383,
|
| 1476 |
+
"grad_norm": 0.773059606552124,
|
| 1477 |
+
"learning_rate": 3.205261156347783e-08,
|
| 1478 |
+
"loss": 1.4591,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.9819891096942337,
|
| 1483 |
+
"grad_norm": 0.8769505023956299,
|
| 1484 |
+
"learning_rate": 2.0372264643713223e-08,
|
| 1485 |
+
"loss": 1.4756,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.9866430865174292,
|
| 1490 |
+
"grad_norm": 0.9209620356559753,
|
| 1491 |
+
"learning_rate": 1.132543704660316e-08,
|
| 1492 |
+
"loss": 1.4957,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.9912970633406246,
|
| 1497 |
+
"grad_norm": 0.9970301389694214,
|
| 1498 |
+
"learning_rate": 4.914516135275937e-09,
|
| 1499 |
+
"loss": 1.4321,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.99595104016382,
|
| 1504 |
+
"grad_norm": 0.6997882127761841,
|
| 1505 |
+
"learning_rate": 1.1411936846705828e-09,
|
| 1506 |
+
"loss": 1.4656,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.0,
|
| 1511 |
+
"step": 21487,
|
| 1512 |
+
"total_flos": 3.101747713212678e+18,
|
| 1513 |
+
"train_loss": 1.5223615952344096,
|
| 1514 |
+
"train_runtime": 9293.396,
|
| 1515 |
+
"train_samples_per_second": 4.624,
|
| 1516 |
+
"train_steps_per_second": 2.312
|
| 1517 |
+
}
|
| 1518 |
+
],
|
| 1519 |
+
"logging_steps": 100,
|
| 1520 |
+
"max_steps": 21487,
|
| 1521 |
+
"num_input_tokens_seen": 0,
|
| 1522 |
+
"num_train_epochs": 1,
|
| 1523 |
+
"save_steps": 500,
|
| 1524 |
+
"stateful_callbacks": {
|
| 1525 |
+
"TrainerControl": {
|
| 1526 |
+
"args": {
|
| 1527 |
+
"should_epoch_stop": false,
|
| 1528 |
+
"should_evaluate": false,
|
| 1529 |
+
"should_log": false,
|
| 1530 |
+
"should_save": true,
|
| 1531 |
+
"should_training_stop": true
|
| 1532 |
+
},
|
| 1533 |
+
"attributes": {}
|
| 1534 |
+
}
|
| 1535 |
+
},
|
| 1536 |
+
"total_flos": 3.101747713212678e+18,
|
| 1537 |
+
"train_batch_size": 2,
|
| 1538 |
+
"trial_name": null,
|
| 1539 |
+
"trial_params": null
|
| 1540 |
+
}
|