Model save
Browse files- README.md +1 -1
- all_results.json +12 -0
- generation_config.json +1 -3
- train_results.json +12 -0
- trainer_state.json +2586 -0
README.md
CHANGED
|
@@ -4,9 +4,9 @@ library_name: transformers
|
|
| 4 |
model_name: qwen2.5-1.5b-sft-iter2
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
-
- hf_jobs
|
| 8 |
- trl
|
| 9 |
- sft
|
|
|
|
| 10 |
licence: license
|
| 11 |
---
|
| 12 |
|
|
|
|
| 4 |
model_name: qwen2.5-1.5b-sft-iter2
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
| 7 |
- trl
|
| 8 |
- sft
|
| 9 |
+
- hf_jobs
|
| 10 |
licence: license
|
| 11 |
---
|
| 12 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"entropy": 0.8685899376869202,
|
| 3 |
+
"epoch": 2.0,
|
| 4 |
+
"mean_token_accuracy": 0.7839328944683075,
|
| 5 |
+
"num_tokens": 42794273.0,
|
| 6 |
+
"total_flos": 79316336967680.0,
|
| 7 |
+
"train_loss": 0.8601785458425525,
|
| 8 |
+
"train_runtime": 2840.349,
|
| 9 |
+
"train_samples": 40756,
|
| 10 |
+
"train_samples_per_second": 28.698,
|
| 11 |
+
"train_steps_per_second": 0.449
|
| 12 |
+
}
|
generation_config.json
CHANGED
|
@@ -1,8 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"do_sample": true,
|
| 3 |
-
"eos_token_id":
|
| 4 |
-
151645
|
| 5 |
-
],
|
| 6 |
"pad_token_id": 151643,
|
| 7 |
"repetition_penalty": 1.1,
|
| 8 |
"temperature": 0.7,
|
|
|
|
| 1 |
{
|
| 2 |
"do_sample": true,
|
| 3 |
+
"eos_token_id": 151645,
|
|
|
|
|
|
|
| 4 |
"pad_token_id": 151643,
|
| 5 |
"repetition_penalty": 1.1,
|
| 6 |
"temperature": 0.7,
|
train_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"entropy": 0.8685899376869202,
|
| 3 |
+
"epoch": 2.0,
|
| 4 |
+
"mean_token_accuracy": 0.7839328944683075,
|
| 5 |
+
"num_tokens": 42794273.0,
|
| 6 |
+
"total_flos": 79316336967680.0,
|
| 7 |
+
"train_loss": 0.8601785458425525,
|
| 8 |
+
"train_runtime": 2840.349,
|
| 9 |
+
"train_samples": 40756,
|
| 10 |
+
"train_samples_per_second": 28.698,
|
| 11 |
+
"train_steps_per_second": 0.449
|
| 12 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2586 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1274,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 0.8923108279705048,
|
| 14 |
+
"epoch": 0.007849293563579277,
|
| 15 |
+
"grad_norm": 1.0846652671644206,
|
| 16 |
+
"learning_rate": 6.25e-07,
|
| 17 |
+
"loss": 0.9037,
|
| 18 |
+
"mean_token_accuracy": 0.7743591070175171,
|
| 19 |
+
"num_tokens": 178317.0,
|
| 20 |
+
"step": 5
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 0.920203971862793,
|
| 24 |
+
"epoch": 0.015698587127158554,
|
| 25 |
+
"grad_norm": 1.054834647195203,
|
| 26 |
+
"learning_rate": 1.40625e-06,
|
| 27 |
+
"loss": 0.9418,
|
| 28 |
+
"mean_token_accuracy": 0.7692570209503173,
|
| 29 |
+
"num_tokens": 357289.0,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 0.8834264755249024,
|
| 34 |
+
"epoch": 0.023547880690737835,
|
| 35 |
+
"grad_norm": 1.074626954566983,
|
| 36 |
+
"learning_rate": 2.1875000000000002e-06,
|
| 37 |
+
"loss": 0.8905,
|
| 38 |
+
"mean_token_accuracy": 0.77989462018013,
|
| 39 |
+
"num_tokens": 533821.0,
|
| 40 |
+
"step": 15
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 0.9269816219806671,
|
| 44 |
+
"epoch": 0.03139717425431711,
|
| 45 |
+
"grad_norm": 1.275445940747145,
|
| 46 |
+
"learning_rate": 2.96875e-06,
|
| 47 |
+
"loss": 0.948,
|
| 48 |
+
"mean_token_accuracy": 0.7666654944419861,
|
| 49 |
+
"num_tokens": 698178.0,
|
| 50 |
+
"step": 20
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"entropy": 0.9225781202316284,
|
| 54 |
+
"epoch": 0.03924646781789639,
|
| 55 |
+
"grad_norm": 1.2339414545355725,
|
| 56 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 57 |
+
"loss": 0.8796,
|
| 58 |
+
"mean_token_accuracy": 0.7792899131774902,
|
| 59 |
+
"num_tokens": 859082.0,
|
| 60 |
+
"step": 25
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"entropy": 0.9789364576339722,
|
| 64 |
+
"epoch": 0.04709576138147567,
|
| 65 |
+
"grad_norm": 1.1769021864963694,
|
| 66 |
+
"learning_rate": 4.53125e-06,
|
| 67 |
+
"loss": 0.9546,
|
| 68 |
+
"mean_token_accuracy": 0.7612079977989197,
|
| 69 |
+
"num_tokens": 1034131.0,
|
| 70 |
+
"step": 30
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"entropy": 0.9308826148509979,
|
| 74 |
+
"epoch": 0.054945054945054944,
|
| 75 |
+
"grad_norm": 1.012055593605149,
|
| 76 |
+
"learning_rate": 5.3125e-06,
|
| 77 |
+
"loss": 0.9083,
|
| 78 |
+
"mean_token_accuracy": 0.7734895884990692,
|
| 79 |
+
"num_tokens": 1205811.0,
|
| 80 |
+
"step": 35
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"entropy": 0.9731091439723969,
|
| 84 |
+
"epoch": 0.06279434850863422,
|
| 85 |
+
"grad_norm": 1.2998416722317658,
|
| 86 |
+
"learning_rate": 6.093750000000001e-06,
|
| 87 |
+
"loss": 0.9675,
|
| 88 |
+
"mean_token_accuracy": 0.7617174565792084,
|
| 89 |
+
"num_tokens": 1360770.0,
|
| 90 |
+
"step": 40
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"entropy": 0.8953827381134033,
|
| 94 |
+
"epoch": 0.0706436420722135,
|
| 95 |
+
"grad_norm": 0.9856112544308491,
|
| 96 |
+
"learning_rate": 6.875e-06,
|
| 97 |
+
"loss": 0.8916,
|
| 98 |
+
"mean_token_accuracy": 0.7822651624679565,
|
| 99 |
+
"num_tokens": 1533946.0,
|
| 100 |
+
"step": 45
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"entropy": 0.9609230756759644,
|
| 104 |
+
"epoch": 0.07849293563579278,
|
| 105 |
+
"grad_norm": 1.1151960140633406,
|
| 106 |
+
"learning_rate": 7.656250000000001e-06,
|
| 107 |
+
"loss": 0.952,
|
| 108 |
+
"mean_token_accuracy": 0.7624362647533417,
|
| 109 |
+
"num_tokens": 1701883.0,
|
| 110 |
+
"step": 50
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"entropy": 0.9360766351222992,
|
| 114 |
+
"epoch": 0.08634222919937205,
|
| 115 |
+
"grad_norm": 0.9956541686878148,
|
| 116 |
+
"learning_rate": 8.4375e-06,
|
| 117 |
+
"loss": 0.9278,
|
| 118 |
+
"mean_token_accuracy": 0.7703767657279968,
|
| 119 |
+
"num_tokens": 1874081.0,
|
| 120 |
+
"step": 55
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"entropy": 0.9806451737880707,
|
| 124 |
+
"epoch": 0.09419152276295134,
|
| 125 |
+
"grad_norm": 1.159466881871733,
|
| 126 |
+
"learning_rate": 9.21875e-06,
|
| 127 |
+
"loss": 0.9592,
|
| 128 |
+
"mean_token_accuracy": 0.7591827690601349,
|
| 129 |
+
"num_tokens": 2041363.0,
|
| 130 |
+
"step": 60
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"entropy": 0.9667136311531067,
|
| 134 |
+
"epoch": 0.10204081632653061,
|
| 135 |
+
"grad_norm": 1.1293730384420282,
|
| 136 |
+
"learning_rate": 1e-05,
|
| 137 |
+
"loss": 0.9653,
|
| 138 |
+
"mean_token_accuracy": 0.7633151412010193,
|
| 139 |
+
"num_tokens": 2200578.0,
|
| 140 |
+
"step": 65
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"entropy": 0.991092300415039,
|
| 144 |
+
"epoch": 0.10989010989010989,
|
| 145 |
+
"grad_norm": 1.1315667226075667,
|
| 146 |
+
"learning_rate": 1.0781250000000001e-05,
|
| 147 |
+
"loss": 0.9597,
|
| 148 |
+
"mean_token_accuracy": 0.7614585816860199,
|
| 149 |
+
"num_tokens": 2356944.0,
|
| 150 |
+
"step": 70
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"entropy": 0.9186211705207825,
|
| 154 |
+
"epoch": 0.11773940345368916,
|
| 155 |
+
"grad_norm": 0.9750633684496991,
|
| 156 |
+
"learning_rate": 1.1562500000000002e-05,
|
| 157 |
+
"loss": 0.8825,
|
| 158 |
+
"mean_token_accuracy": 0.7805460155010223,
|
| 159 |
+
"num_tokens": 2544833.0,
|
| 160 |
+
"step": 75
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"entropy": 1.0243119478225708,
|
| 164 |
+
"epoch": 0.12558869701726844,
|
| 165 |
+
"grad_norm": 1.1378562830275354,
|
| 166 |
+
"learning_rate": 1.234375e-05,
|
| 167 |
+
"loss": 0.9967,
|
| 168 |
+
"mean_token_accuracy": 0.7553607821464539,
|
| 169 |
+
"num_tokens": 2713694.0,
|
| 170 |
+
"step": 80
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"entropy": 0.9048753321170807,
|
| 174 |
+
"epoch": 0.13343799058084774,
|
| 175 |
+
"grad_norm": 0.9899557134460932,
|
| 176 |
+
"learning_rate": 1.3125e-05,
|
| 177 |
+
"loss": 0.8766,
|
| 178 |
+
"mean_token_accuracy": 0.7808430552482605,
|
| 179 |
+
"num_tokens": 2887769.0,
|
| 180 |
+
"step": 85
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"entropy": 0.9815359652042389,
|
| 184 |
+
"epoch": 0.141287284144427,
|
| 185 |
+
"grad_norm": 1.1366400203542357,
|
| 186 |
+
"learning_rate": 1.3906250000000001e-05,
|
| 187 |
+
"loss": 0.9827,
|
| 188 |
+
"mean_token_accuracy": 0.7579309821128846,
|
| 189 |
+
"num_tokens": 3059822.0,
|
| 190 |
+
"step": 90
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"entropy": 0.9039746880531311,
|
| 194 |
+
"epoch": 0.14913657770800628,
|
| 195 |
+
"grad_norm": 1.5861385589334023,
|
| 196 |
+
"learning_rate": 1.4687500000000001e-05,
|
| 197 |
+
"loss": 0.8752,
|
| 198 |
+
"mean_token_accuracy": 0.7799261450767517,
|
| 199 |
+
"num_tokens": 3232425.0,
|
| 200 |
+
"step": 95
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"entropy": 0.9870649099349975,
|
| 204 |
+
"epoch": 0.15698587127158556,
|
| 205 |
+
"grad_norm": 1.2371253838104694,
|
| 206 |
+
"learning_rate": 1.546875e-05,
|
| 207 |
+
"loss": 0.9574,
|
| 208 |
+
"mean_token_accuracy": 0.7566253483295441,
|
| 209 |
+
"num_tokens": 3385232.0,
|
| 210 |
+
"step": 100
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"entropy": 0.9329905152320862,
|
| 214 |
+
"epoch": 0.16483516483516483,
|
| 215 |
+
"grad_norm": 1.0837180728086673,
|
| 216 |
+
"learning_rate": 1.6250000000000002e-05,
|
| 217 |
+
"loss": 0.9102,
|
| 218 |
+
"mean_token_accuracy": 0.77213174700737,
|
| 219 |
+
"num_tokens": 3558956.0,
|
| 220 |
+
"step": 105
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"entropy": 0.9576933801174163,
|
| 224 |
+
"epoch": 0.1726844583987441,
|
| 225 |
+
"grad_norm": 1.0184373297120872,
|
| 226 |
+
"learning_rate": 1.703125e-05,
|
| 227 |
+
"loss": 0.9446,
|
| 228 |
+
"mean_token_accuracy": 0.7666807472705841,
|
| 229 |
+
"num_tokens": 3741267.0,
|
| 230 |
+
"step": 110
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"entropy": 0.9833621561527253,
|
| 234 |
+
"epoch": 0.18053375196232338,
|
| 235 |
+
"grad_norm": 1.081495146597845,
|
| 236 |
+
"learning_rate": 1.7812500000000003e-05,
|
| 237 |
+
"loss": 0.9751,
|
| 238 |
+
"mean_token_accuracy": 0.760509479045868,
|
| 239 |
+
"num_tokens": 3901752.0,
|
| 240 |
+
"step": 115
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"entropy": 1.09641934633255,
|
| 244 |
+
"epoch": 0.18838304552590268,
|
| 245 |
+
"grad_norm": 1.2012234349363566,
|
| 246 |
+
"learning_rate": 1.859375e-05,
|
| 247 |
+
"loss": 1.0963,
|
| 248 |
+
"mean_token_accuracy": 0.7360260009765625,
|
| 249 |
+
"num_tokens": 4072999.0,
|
| 250 |
+
"step": 120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"entropy": 0.9882058680057526,
|
| 254 |
+
"epoch": 0.19623233908948196,
|
| 255 |
+
"grad_norm": 0.9937322506697319,
|
| 256 |
+
"learning_rate": 1.9375e-05,
|
| 257 |
+
"loss": 0.9853,
|
| 258 |
+
"mean_token_accuracy": 0.7576613247394561,
|
| 259 |
+
"num_tokens": 4246177.0,
|
| 260 |
+
"step": 125
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"entropy": 1.0119176387786866,
|
| 264 |
+
"epoch": 0.20408163265306123,
|
| 265 |
+
"grad_norm": 1.0694914434524772,
|
| 266 |
+
"learning_rate": 1.999996242489157e-05,
|
| 267 |
+
"loss": 0.9631,
|
| 268 |
+
"mean_token_accuracy": 0.758573442697525,
|
| 269 |
+
"num_tokens": 4410505.0,
|
| 270 |
+
"step": 130
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"entropy": 0.9016034841537476,
|
| 274 |
+
"epoch": 0.2119309262166405,
|
| 275 |
+
"grad_norm": 1.3380668664631206,
|
| 276 |
+
"learning_rate": 1.9998647325745995e-05,
|
| 277 |
+
"loss": 0.9024,
|
| 278 |
+
"mean_token_accuracy": 0.7715110659599305,
|
| 279 |
+
"num_tokens": 4585906.0,
|
| 280 |
+
"step": 135
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"entropy": 0.9409748911857605,
|
| 284 |
+
"epoch": 0.21978021978021978,
|
| 285 |
+
"grad_norm": 1.2699863477033326,
|
| 286 |
+
"learning_rate": 1.9995453753547198e-05,
|
| 287 |
+
"loss": 0.9172,
|
| 288 |
+
"mean_token_accuracy": 0.769565087556839,
|
| 289 |
+
"num_tokens": 4752246.0,
|
| 290 |
+
"step": 140
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"entropy": 0.9414529740810395,
|
| 294 |
+
"epoch": 0.22762951334379905,
|
| 295 |
+
"grad_norm": 1.015452830363856,
|
| 296 |
+
"learning_rate": 1.9990382308280272e-05,
|
| 297 |
+
"loss": 0.931,
|
| 298 |
+
"mean_token_accuracy": 0.7670335710048676,
|
| 299 |
+
"num_tokens": 4911595.0,
|
| 300 |
+
"step": 145
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"entropy": 1.0376124560832978,
|
| 304 |
+
"epoch": 0.23547880690737832,
|
| 305 |
+
"grad_norm": 1.1058546158019706,
|
| 306 |
+
"learning_rate": 1.9983433942731427e-05,
|
| 307 |
+
"loss": 1.0206,
|
| 308 |
+
"mean_token_accuracy": 0.749090451002121,
|
| 309 |
+
"num_tokens": 5086433.0,
|
| 310 |
+
"step": 150
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"entropy": 0.9030189216136932,
|
| 314 |
+
"epoch": 0.24332810047095763,
|
| 315 |
+
"grad_norm": 1.0525541372324732,
|
| 316 |
+
"learning_rate": 1.9974609962308986e-05,
|
| 317 |
+
"loss": 0.9007,
|
| 318 |
+
"mean_token_accuracy": 0.7748336374759675,
|
| 319 |
+
"num_tokens": 5253732.0,
|
| 320 |
+
"step": 155
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"entropy": 0.9299406290054322,
|
| 324 |
+
"epoch": 0.25117739403453687,
|
| 325 |
+
"grad_norm": 1.2018132967943806,
|
| 326 |
+
"learning_rate": 1.9963912024798136e-05,
|
| 327 |
+
"loss": 0.9244,
|
| 328 |
+
"mean_token_accuracy": 0.7697316825389862,
|
| 329 |
+
"num_tokens": 5417674.0,
|
| 330 |
+
"step": 160
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"entropy": 1.0168113708496094,
|
| 334 |
+
"epoch": 0.25902668759811615,
|
| 335 |
+
"grad_norm": 1.9151888617227308,
|
| 336 |
+
"learning_rate": 1.9951342140049483e-05,
|
| 337 |
+
"loss": 0.9865,
|
| 338 |
+
"mean_token_accuracy": 0.7549250602722168,
|
| 339 |
+
"num_tokens": 5583642.0,
|
| 340 |
+
"step": 165
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"entropy": 1.0209120154380797,
|
| 344 |
+
"epoch": 0.2668759811616955,
|
| 345 |
+
"grad_norm": 1.1470899446647818,
|
| 346 |
+
"learning_rate": 1.9936902669601436e-05,
|
| 347 |
+
"loss": 1.0185,
|
| 348 |
+
"mean_token_accuracy": 0.7463882029056549,
|
| 349 |
+
"num_tokens": 5749362.0,
|
| 350 |
+
"step": 170
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"entropy": 1.000859671831131,
|
| 354 |
+
"epoch": 0.27472527472527475,
|
| 355 |
+
"grad_norm": 1.0314754852223484,
|
| 356 |
+
"learning_rate": 1.992059632623657e-05,
|
| 357 |
+
"loss": 0.9795,
|
| 358 |
+
"mean_token_accuracy": 0.7567110538482666,
|
| 359 |
+
"num_tokens": 5931525.0,
|
| 360 |
+
"step": 175
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"entropy": 0.9324127018451691,
|
| 364 |
+
"epoch": 0.282574568288854,
|
| 365 |
+
"grad_norm": 1.1948728155140573,
|
| 366 |
+
"learning_rate": 1.9902426173471933e-05,
|
| 367 |
+
"loss": 0.9432,
|
| 368 |
+
"mean_token_accuracy": 0.7646886527538299,
|
| 369 |
+
"num_tokens": 6107352.0,
|
| 370 |
+
"step": 180
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"entropy": 0.8979306757450104,
|
| 374 |
+
"epoch": 0.2904238618524333,
|
| 375 |
+
"grad_norm": 1.149225401693029,
|
| 376 |
+
"learning_rate": 1.9882395624983522e-05,
|
| 377 |
+
"loss": 0.8723,
|
| 378 |
+
"mean_token_accuracy": 0.7772454380989074,
|
| 379 |
+
"num_tokens": 6284190.0,
|
| 380 |
+
"step": 185
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"entropy": 0.9475451648235321,
|
| 384 |
+
"epoch": 0.29827315541601257,
|
| 385 |
+
"grad_norm": 1.0633396188858635,
|
| 386 |
+
"learning_rate": 1.986050844396493e-05,
|
| 387 |
+
"loss": 0.9477,
|
| 388 |
+
"mean_token_accuracy": 0.7611142456531524,
|
| 389 |
+
"num_tokens": 6448910.0,
|
| 390 |
+
"step": 190
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"entropy": 0.9558171093463897,
|
| 394 |
+
"epoch": 0.30612244897959184,
|
| 395 |
+
"grad_norm": 1.2123998006858923,
|
| 396 |
+
"learning_rate": 1.9836768742420355e-05,
|
| 397 |
+
"loss": 0.9448,
|
| 398 |
+
"mean_token_accuracy": 0.7647868692874908,
|
| 399 |
+
"num_tokens": 6605130.0,
|
| 400 |
+
"step": 195
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"entropy": 0.957121080160141,
|
| 404 |
+
"epoch": 0.3139717425431711,
|
| 405 |
+
"grad_norm": 1.060560663378554,
|
| 406 |
+
"learning_rate": 1.9811180980392054e-05,
|
| 407 |
+
"loss": 0.9441,
|
| 408 |
+
"mean_token_accuracy": 0.763433963060379,
|
| 409 |
+
"num_tokens": 6774784.0,
|
| 410 |
+
"step": 200
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"entropy": 0.9790769934654235,
|
| 414 |
+
"epoch": 0.3218210361067504,
|
| 415 |
+
"grad_norm": 1.0269436730850807,
|
| 416 |
+
"learning_rate": 1.9783749965122444e-05,
|
| 417 |
+
"loss": 0.9694,
|
| 418 |
+
"mean_token_accuracy": 0.7610109508037567,
|
| 419 |
+
"num_tokens": 6951192.0,
|
| 420 |
+
"step": 205
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"entropy": 0.944574511051178,
|
| 424 |
+
"epoch": 0.32967032967032966,
|
| 425 |
+
"grad_norm": 1.207945428006486,
|
| 426 |
+
"learning_rate": 1.975448085015093e-05,
|
| 427 |
+
"loss": 0.921,
|
| 428 |
+
"mean_token_accuracy": 0.7690154552459717,
|
| 429 |
+
"num_tokens": 7126342.0,
|
| 430 |
+
"step": 210
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"entropy": 0.941559374332428,
|
| 434 |
+
"epoch": 0.33751962323390894,
|
| 435 |
+
"grad_norm": 1.3645198482716716,
|
| 436 |
+
"learning_rate": 1.9723379134345698e-05,
|
| 437 |
+
"loss": 0.9277,
|
| 438 |
+
"mean_token_accuracy": 0.7681373357772827,
|
| 439 |
+
"num_tokens": 7291928.0,
|
| 440 |
+
"step": 215
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"entropy": 0.8783131539821625,
|
| 444 |
+
"epoch": 0.3453689167974882,
|
| 445 |
+
"grad_norm": 1.1324121467643162,
|
| 446 |
+
"learning_rate": 1.9690450660870657e-05,
|
| 447 |
+
"loss": 0.8823,
|
| 448 |
+
"mean_token_accuracy": 0.7756971180438995,
|
| 449 |
+
"num_tokens": 7455891.0,
|
| 450 |
+
"step": 220
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"entropy": 0.9457615196704865,
|
| 454 |
+
"epoch": 0.3532182103610675,
|
| 455 |
+
"grad_norm": 1.4176003588865003,
|
| 456 |
+
"learning_rate": 1.965570161608762e-05,
|
| 457 |
+
"loss": 0.9054,
|
| 458 |
+
"mean_token_accuracy": 0.7736207127571106,
|
| 459 |
+
"num_tokens": 7614914.0,
|
| 460 |
+
"step": 225
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"entropy": 0.9694023311138154,
|
| 464 |
+
"epoch": 0.36106750392464676,
|
| 465 |
+
"grad_norm": 1.0480976699894644,
|
| 466 |
+
"learning_rate": 1.961913852839409e-05,
|
| 467 |
+
"loss": 0.9443,
|
| 468 |
+
"mean_token_accuracy": 0.7673780143260955,
|
| 469 |
+
"num_tokens": 7779157.0,
|
| 470 |
+
"step": 230
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"entropy": 1.0053590416908265,
|
| 474 |
+
"epoch": 0.36891679748822603,
|
| 475 |
+
"grad_norm": 1.1458460377251418,
|
| 476 |
+
"learning_rate": 1.958076826699676e-05,
|
| 477 |
+
"loss": 1.0173,
|
| 478 |
+
"mean_token_accuracy": 0.7478827059268951,
|
| 479 |
+
"num_tokens": 7944245.0,
|
| 480 |
+
"step": 235
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"entropy": 0.9745344460010529,
|
| 484 |
+
"epoch": 0.37676609105180536,
|
| 485 |
+
"grad_norm": 1.0262121083072202,
|
| 486 |
+
"learning_rate": 1.954059804062092e-05,
|
| 487 |
+
"loss": 0.9545,
|
| 488 |
+
"mean_token_accuracy": 0.7606645584106445,
|
| 489 |
+
"num_tokens": 8108061.0,
|
| 490 |
+
"step": 240
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"entropy": 0.9760924875736237,
|
| 494 |
+
"epoch": 0.38461538461538464,
|
| 495 |
+
"grad_norm": 1.102383611635918,
|
| 496 |
+
"learning_rate": 1.9498635396156217e-05,
|
| 497 |
+
"loss": 0.9593,
|
| 498 |
+
"mean_token_accuracy": 0.7614710688591003,
|
| 499 |
+
"num_tokens": 8276341.0,
|
| 500 |
+
"step": 245
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"entropy": 1.015894317626953,
|
| 504 |
+
"epoch": 0.3924646781789639,
|
| 505 |
+
"grad_norm": 1.0921383863500178,
|
| 506 |
+
"learning_rate": 1.945488821723873e-05,
|
| 507 |
+
"loss": 1.008,
|
| 508 |
+
"mean_token_accuracy": 0.7551724553108216,
|
| 509 |
+
"num_tokens": 8451051.0,
|
| 510 |
+
"step": 250
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"entropy": 0.9748592793941497,
|
| 514 |
+
"epoch": 0.4003139717425432,
|
| 515 |
+
"grad_norm": 1.2778893065222514,
|
| 516 |
+
"learning_rate": 1.9409364722769882e-05,
|
| 517 |
+
"loss": 0.9637,
|
| 518 |
+
"mean_token_accuracy": 0.7573827385902405,
|
| 519 |
+
"num_tokens": 8618022.0,
|
| 520 |
+
"step": 255
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"entropy": 0.9644428014755249,
|
| 524 |
+
"epoch": 0.40816326530612246,
|
| 525 |
+
"grad_norm": 1.1784396768189709,
|
| 526 |
+
"learning_rate": 1.936207346537233e-05,
|
| 527 |
+
"loss": 0.9545,
|
| 528 |
+
"mean_token_accuracy": 0.7592711687088013,
|
| 529 |
+
"num_tokens": 8787213.0,
|
| 530 |
+
"step": 260
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"entropy": 0.9041996359825134,
|
| 534 |
+
"epoch": 0.41601255886970173,
|
| 535 |
+
"grad_norm": 1.0883429235361104,
|
| 536 |
+
"learning_rate": 1.931302332978316e-05,
|
| 537 |
+
"loss": 0.8941,
|
| 538 |
+
"mean_token_accuracy": 0.7768190681934357,
|
| 539 |
+
"num_tokens": 8958532.0,
|
| 540 |
+
"step": 265
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"entropy": 1.011337196826935,
|
| 544 |
+
"epoch": 0.423861852433281,
|
| 545 |
+
"grad_norm": 1.0113459071796085,
|
| 546 |
+
"learning_rate": 1.9262223531184678e-05,
|
| 547 |
+
"loss": 1.0015,
|
| 548 |
+
"mean_token_accuracy": 0.7520352244377136,
|
| 549 |
+
"num_tokens": 9121824.0,
|
| 550 |
+
"step": 270
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"entropy": 1.041982936859131,
|
| 554 |
+
"epoch": 0.4317111459968603,
|
| 555 |
+
"grad_norm": 1.0776344102623627,
|
| 556 |
+
"learning_rate": 1.9209683613473143e-05,
|
| 557 |
+
"loss": 1.0368,
|
| 558 |
+
"mean_token_accuracy": 0.7477185308933259,
|
| 559 |
+
"num_tokens": 9283293.0,
|
| 560 |
+
"step": 275
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"entropy": 0.9901135206222534,
|
| 564 |
+
"epoch": 0.43956043956043955,
|
| 565 |
+
"grad_norm": 1.1557566124905563,
|
| 566 |
+
"learning_rate": 1.9155413447465715e-05,
|
| 567 |
+
"loss": 0.9556,
|
| 568 |
+
"mean_token_accuracy": 0.7589726448059082,
|
| 569 |
+
"num_tokens": 9444880.0,
|
| 570 |
+
"step": 280
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"entropy": 0.9939525187015533,
|
| 574 |
+
"epoch": 0.4474097331240188,
|
| 575 |
+
"grad_norm": 1.2709530119251884,
|
| 576 |
+
"learning_rate": 1.9099423229046015e-05,
|
| 577 |
+
"loss": 0.9939,
|
| 578 |
+
"mean_token_accuracy": 0.7533400356769562,
|
| 579 |
+
"num_tokens": 9603203.0,
|
| 580 |
+
"step": 285
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"entropy": 1.0107034623622895,
|
| 584 |
+
"epoch": 0.4552590266875981,
|
| 585 |
+
"grad_norm": 1.024836086226542,
|
| 586 |
+
"learning_rate": 1.9041723477248575e-05,
|
| 587 |
+
"loss": 0.9986,
|
| 588 |
+
"mean_token_accuracy": 0.7503722310066223,
|
| 589 |
+
"num_tokens": 9773987.0,
|
| 590 |
+
"step": 290
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"entropy": 0.9630116701126099,
|
| 594 |
+
"epoch": 0.4631083202511774,
|
| 595 |
+
"grad_norm": 1.1004965555660209,
|
| 596 |
+
"learning_rate": 1.8982325032282616e-05,
|
| 597 |
+
"loss": 0.945,
|
| 598 |
+
"mean_token_accuracy": 0.7651254773139954,
|
| 599 |
+
"num_tokens": 9946995.0,
|
| 600 |
+
"step": 295
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"entropy": 0.9473274886608124,
|
| 604 |
+
"epoch": 0.47095761381475665,
|
| 605 |
+
"grad_norm": 1.0501422807822132,
|
| 606 |
+
"learning_rate": 1.8921239053495465e-05,
|
| 607 |
+
"loss": 0.9258,
|
| 608 |
+
"mean_token_accuracy": 0.7670532703399658,
|
| 609 |
+
"num_tokens": 10120753.0,
|
| 610 |
+
"step": 300
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"entropy": 1.024631142616272,
|
| 614 |
+
"epoch": 0.478806907378336,
|
| 615 |
+
"grad_norm": 1.154364710157582,
|
| 616 |
+
"learning_rate": 1.8858477017276002e-05,
|
| 617 |
+
"loss": 1.0069,
|
| 618 |
+
"mean_token_accuracy": 0.750711727142334,
|
| 619 |
+
"num_tokens": 10282985.0,
|
| 620 |
+
"step": 305
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"entropy": 0.9688675343990326,
|
| 624 |
+
"epoch": 0.48665620094191525,
|
| 625 |
+
"grad_norm": 1.1885072584895293,
|
| 626 |
+
"learning_rate": 1.8794050714898596e-05,
|
| 627 |
+
"loss": 0.9633,
|
| 628 |
+
"mean_token_accuracy": 0.7601287066936493,
|
| 629 |
+
"num_tokens": 10448522.0,
|
| 630 |
+
"step": 310
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"entropy": 1.006821882724762,
|
| 634 |
+
"epoch": 0.4945054945054945,
|
| 635 |
+
"grad_norm": 1.11770193555216,
|
| 636 |
+
"learning_rate": 1.87279722503078e-05,
|
| 637 |
+
"loss": 0.9778,
|
| 638 |
+
"mean_token_accuracy": 0.7547464072704315,
|
| 639 |
+
"num_tokens": 10609140.0,
|
| 640 |
+
"step": 315
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"entropy": 0.9243999183177948,
|
| 644 |
+
"epoch": 0.5023547880690737,
|
| 645 |
+
"grad_norm": 1.0034673037415462,
|
| 646 |
+
"learning_rate": 1.866025403784439e-05,
|
| 647 |
+
"loss": 0.9233,
|
| 648 |
+
"mean_token_accuracy": 0.7648786187171936,
|
| 649 |
+
"num_tokens": 10788613.0,
|
| 650 |
+
"step": 320
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"entropy": 1.0077704966068268,
|
| 654 |
+
"epoch": 0.5102040816326531,
|
| 655 |
+
"grad_norm": 1.1068829395312032,
|
| 656 |
+
"learning_rate": 1.859090879991302e-05,
|
| 657 |
+
"loss": 0.9831,
|
| 658 |
+
"mean_token_accuracy": 0.7548430681228637,
|
| 659 |
+
"num_tokens": 10959077.0,
|
| 660 |
+
"step": 325
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"entropy": 0.9248578608036041,
|
| 664 |
+
"epoch": 0.5180533751962323,
|
| 665 |
+
"grad_norm": 1.3266089337457716,
|
| 666 |
+
"learning_rate": 1.8519949564592047e-05,
|
| 667 |
+
"loss": 0.9237,
|
| 668 |
+
"mean_token_accuracy": 0.770110958814621,
|
| 669 |
+
"num_tokens": 11125485.0,
|
| 670 |
+
"step": 330
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"entropy": 0.9888102948665619,
|
| 674 |
+
"epoch": 0.5259026687598116,
|
| 675 |
+
"grad_norm": 0.983419114939915,
|
| 676 |
+
"learning_rate": 1.8447389663185905e-05,
|
| 677 |
+
"loss": 0.9709,
|
| 678 |
+
"mean_token_accuracy": 0.7550422787666321,
|
| 679 |
+
"num_tokens": 11303037.0,
|
| 680 |
+
"step": 335
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"entropy": 0.9655410647392273,
|
| 684 |
+
"epoch": 0.533751962323391,
|
| 685 |
+
"grad_norm": 1.4182083064120892,
|
| 686 |
+
"learning_rate": 1.837324272772052e-05,
|
| 687 |
+
"loss": 0.9662,
|
| 688 |
+
"mean_token_accuracy": 0.7561042129993438,
|
| 689 |
+
"num_tokens": 11476471.0,
|
| 690 |
+
"step": 340
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"entropy": 0.9178342401981354,
|
| 694 |
+
"epoch": 0.5416012558869702,
|
| 695 |
+
"grad_norm": 1.0185913242694713,
|
| 696 |
+
"learning_rate": 1.829752268838222e-05,
|
| 697 |
+
"loss": 0.8979,
|
| 698 |
+
"mean_token_accuracy": 0.770701014995575,
|
| 699 |
+
"num_tokens": 11650594.0,
|
| 700 |
+
"step": 345
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"entropy": 0.961567485332489,
|
| 704 |
+
"epoch": 0.5494505494505495,
|
| 705 |
+
"grad_norm": 1.0035892038018919,
|
| 706 |
+
"learning_rate": 1.8220243770900623e-05,
|
| 707 |
+
"loss": 0.9243,
|
| 708 |
+
"mean_token_accuracy": 0.7712348163127899,
|
| 709 |
+
"num_tokens": 11825780.0,
|
| 710 |
+
"step": 350
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"entropy": 0.9918673753738403,
|
| 714 |
+
"epoch": 0.5572998430141287,
|
| 715 |
+
"grad_norm": 1.105682266658899,
|
| 716 |
+
"learning_rate": 1.8141420493876035e-05,
|
| 717 |
+
"loss": 1.003,
|
| 718 |
+
"mean_token_accuracy": 0.7465662598609925,
|
| 719 |
+
"num_tokens": 11988358.0,
|
| 720 |
+
"step": 355
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"entropy": 0.9163014113903045,
|
| 724 |
+
"epoch": 0.565149136577708,
|
| 725 |
+
"grad_norm": 1.049630919575967,
|
| 726 |
+
"learning_rate": 1.806106766605178e-05,
|
| 727 |
+
"loss": 0.8911,
|
| 728 |
+
"mean_token_accuracy": 0.7753135740756989,
|
| 729 |
+
"num_tokens": 12163426.0,
|
| 730 |
+
"step": 360
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"entropy": 0.9849283695220947,
|
| 734 |
+
"epoch": 0.5729984301412873,
|
| 735 |
+
"grad_norm": 1.170395657785239,
|
| 736 |
+
"learning_rate": 1.7979200383532055e-05,
|
| 737 |
+
"loss": 0.968,
|
| 738 |
+
"mean_token_accuracy": 0.7599095821380615,
|
| 739 |
+
"num_tokens": 12332951.0,
|
| 740 |
+
"step": 365
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"entropy": 0.9746610879898071,
|
| 744 |
+
"epoch": 0.5808477237048666,
|
| 745 |
+
"grad_norm": 1.1325528608232394,
|
| 746 |
+
"learning_rate": 1.789583402694577e-05,
|
| 747 |
+
"loss": 0.937,
|
| 748 |
+
"mean_token_accuracy": 0.7656144857406616,
|
| 749 |
+
"num_tokens": 12502149.0,
|
| 750 |
+
"step": 370
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"entropy": 0.9254151880741119,
|
| 754 |
+
"epoch": 0.5886970172684458,
|
| 755 |
+
"grad_norm": 1.1231042460652556,
|
| 756 |
+
"learning_rate": 1.7810984258556955e-05,
|
| 757 |
+
"loss": 0.9305,
|
| 758 |
+
"mean_token_accuracy": 0.7643348395824432,
|
| 759 |
+
"num_tokens": 12663339.0,
|
| 760 |
+
"step": 375
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"entropy": 0.9249713122844696,
|
| 764 |
+
"epoch": 0.5965463108320251,
|
| 765 |
+
"grad_norm": 1.0211861003034832,
|
| 766 |
+
"learning_rate": 1.7724667019322258e-05,
|
| 767 |
+
"loss": 0.9084,
|
| 768 |
+
"mean_token_accuracy": 0.7695456087589264,
|
| 769 |
+
"num_tokens": 12836294.0,
|
| 770 |
+
"step": 380
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"entropy": 0.9364533722400665,
|
| 774 |
+
"epoch": 0.6043956043956044,
|
| 775 |
+
"grad_norm": 1.1150645843653726,
|
| 776 |
+
"learning_rate": 1.7636898525896057e-05,
|
| 777 |
+
"loss": 0.9429,
|
| 778 |
+
"mean_token_accuracy": 0.7654342174530029,
|
| 779 |
+
"num_tokens": 13006335.0,
|
| 780 |
+
"step": 385
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"entropy": 0.9740090370178223,
|
| 784 |
+
"epoch": 0.6122448979591837,
|
| 785 |
+
"grad_norm": 0.9949354603131415,
|
| 786 |
+
"learning_rate": 1.7547695267583794e-05,
|
| 787 |
+
"loss": 0.9343,
|
| 788 |
+
"mean_token_accuracy": 0.7641371607780456,
|
| 789 |
+
"num_tokens": 13174822.0,
|
| 790 |
+
"step": 390
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"entropy": 1.0029440701007843,
|
| 794 |
+
"epoch": 0.6200941915227629,
|
| 795 |
+
"grad_norm": 1.065256944897286,
|
| 796 |
+
"learning_rate": 1.74570740032441e-05,
|
| 797 |
+
"loss": 1.0214,
|
| 798 |
+
"mean_token_accuracy": 0.7473598957061768,
|
| 799 |
+
"num_tokens": 13345589.0,
|
| 800 |
+
"step": 395
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"entropy": 1.03370258808136,
|
| 804 |
+
"epoch": 0.6279434850863422,
|
| 805 |
+
"grad_norm": 1.063443882967833,
|
| 806 |
+
"learning_rate": 1.736505175814025e-05,
|
| 807 |
+
"loss": 1.0078,
|
| 808 |
+
"mean_token_accuracy": 0.7507629573345185,
|
| 809 |
+
"num_tokens": 13521755.0,
|
| 810 |
+
"step": 400
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"entropy": 0.9608456194400787,
|
| 814 |
+
"epoch": 0.6357927786499215,
|
| 815 |
+
"grad_norm": 1.1864725278261707,
|
| 816 |
+
"learning_rate": 1.7271645820741586e-05,
|
| 817 |
+
"loss": 0.972,
|
| 818 |
+
"mean_token_accuracy": 0.758358484506607,
|
| 819 |
+
"num_tokens": 13688708.0,
|
| 820 |
+
"step": 405
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"entropy": 0.9463632822036743,
|
| 824 |
+
"epoch": 0.6436420722135008,
|
| 825 |
+
"grad_norm": 1.0890362990595408,
|
| 826 |
+
"learning_rate": 1.7176873739475475e-05,
|
| 827 |
+
"loss": 0.9391,
|
| 828 |
+
"mean_token_accuracy": 0.7674518883228302,
|
| 829 |
+
"num_tokens": 13860224.0,
|
| 830 |
+
"step": 410
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"entropy": 0.9748165845870972,
|
| 834 |
+
"epoch": 0.6514913657770801,
|
| 835 |
+
"grad_norm": 0.9520771806522914,
|
| 836 |
+
"learning_rate": 1.7080753319430452e-05,
|
| 837 |
+
"loss": 0.9696,
|
| 838 |
+
"mean_token_accuracy": 0.7622897684574127,
|
| 839 |
+
"num_tokens": 14023392.0,
|
| 840 |
+
"step": 415
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"entropy": 0.9707380533218384,
|
| 844 |
+
"epoch": 0.6593406593406593,
|
| 845 |
+
"grad_norm": 1.2204279651479377,
|
| 846 |
+
"learning_rate": 1.6983302619011125e-05,
|
| 847 |
+
"loss": 0.966,
|
| 848 |
+
"mean_token_accuracy": 0.7559765994548797,
|
| 849 |
+
"num_tokens": 14191662.0,
|
| 850 |
+
"step": 420
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"entropy": 0.9302991569042206,
|
| 854 |
+
"epoch": 0.6671899529042387,
|
| 855 |
+
"grad_norm": 0.9961924695228186,
|
| 856 |
+
"learning_rate": 1.6884539946545486e-05,
|
| 857 |
+
"loss": 0.9207,
|
| 858 |
+
"mean_token_accuracy": 0.7661088109016418,
|
| 859 |
+
"num_tokens": 14349448.0,
|
| 860 |
+
"step": 425
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"entropy": 0.9614250659942627,
|
| 864 |
+
"epoch": 0.6750392464678179,
|
| 865 |
+
"grad_norm": 0.9998907901716252,
|
| 866 |
+
"learning_rate": 1.6784483856845287e-05,
|
| 867 |
+
"loss": 0.9574,
|
| 868 |
+
"mean_token_accuracy": 0.7567176103591919,
|
| 869 |
+
"num_tokens": 14521779.0,
|
| 870 |
+
"step": 430
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"entropy": 0.9261027753353119,
|
| 874 |
+
"epoch": 0.6828885400313972,
|
| 875 |
+
"grad_norm": 1.0424535740324896,
|
| 876 |
+
"learning_rate": 1.6683153147720098e-05,
|
| 877 |
+
"loss": 0.911,
|
| 878 |
+
"mean_token_accuracy": 0.7726584434509277,
|
| 879 |
+
"num_tokens": 14689226.0,
|
| 880 |
+
"step": 435
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"entropy": 0.9621528267860413,
|
| 884 |
+
"epoch": 0.6907378335949764,
|
| 885 |
+
"grad_norm": 0.9502816401280559,
|
| 886 |
+
"learning_rate": 1.6580566856445684e-05,
|
| 887 |
+
"loss": 0.9364,
|
| 888 |
+
"mean_token_accuracy": 0.7639203608036041,
|
| 889 |
+
"num_tokens": 14858582.0,
|
| 890 |
+
"step": 440
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"entropy": 1.0020318806171418,
|
| 894 |
+
"epoch": 0.6985871271585558,
|
| 895 |
+
"grad_norm": 1.0159981652474168,
|
| 896 |
+
"learning_rate": 1.647674425618747e-05,
|
| 897 |
+
"loss": 0.9902,
|
| 898 |
+
"mean_token_accuracy": 0.7512995481491089,
|
| 899 |
+
"num_tokens": 15023085.0,
|
| 900 |
+
"step": 445
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"entropy": 0.9520144641399384,
|
| 904 |
+
"epoch": 0.706436420722135,
|
| 905 |
+
"grad_norm": 1.1375644230651754,
|
| 906 |
+
"learning_rate": 1.6371704852379587e-05,
|
| 907 |
+
"loss": 0.9387,
|
| 908 |
+
"mean_token_accuracy": 0.764590722322464,
|
| 909 |
+
"num_tokens": 15188234.0,
|
| 910 |
+
"step": 450
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"entropy": 0.9459578275680542,
|
| 914 |
+
"epoch": 0.7142857142857143,
|
| 915 |
+
"grad_norm": 1.2071178598300847,
|
| 916 |
+
"learning_rate": 1.6265468379060364e-05,
|
| 917 |
+
"loss": 0.9294,
|
| 918 |
+
"mean_token_accuracy": 0.7656639575958252,
|
| 919 |
+
"num_tokens": 15350283.0,
|
| 920 |
+
"step": 455
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"entropy": 0.9672309875488281,
|
| 924 |
+
"epoch": 0.7221350078492935,
|
| 925 |
+
"grad_norm": 0.9739572209997952,
|
| 926 |
+
"learning_rate": 1.615805479516484e-05,
|
| 927 |
+
"loss": 0.944,
|
| 928 |
+
"mean_token_accuracy": 0.7605835914611816,
|
| 929 |
+
"num_tokens": 15520942.0,
|
| 930 |
+
"step": 460
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"entropy": 0.9020306468009949,
|
| 934 |
+
"epoch": 0.7299843014128728,
|
| 935 |
+
"grad_norm": 0.9384604260954476,
|
| 936 |
+
"learning_rate": 1.6049484280775012e-05,
|
| 937 |
+
"loss": 0.9007,
|
| 938 |
+
"mean_token_accuracy": 0.7723349690437317,
|
| 939 |
+
"num_tokens": 15696815.0,
|
| 940 |
+
"step": 465
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"entropy": 1.0272919476032256,
|
| 944 |
+
"epoch": 0.7378335949764521,
|
| 945 |
+
"grad_norm": 1.1057780899513212,
|
| 946 |
+
"learning_rate": 1.593977723332855e-05,
|
| 947 |
+
"loss": 1.0052,
|
| 948 |
+
"mean_token_accuracy": 0.7468396842479705,
|
| 949 |
+
"num_tokens": 15865973.0,
|
| 950 |
+
"step": 470
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"entropy": 0.9665712773799896,
|
| 954 |
+
"epoch": 0.7456828885400314,
|
| 955 |
+
"grad_norm": 1.1136016104932442,
|
| 956 |
+
"learning_rate": 1.5828954263786688e-05,
|
| 957 |
+
"loss": 0.9556,
|
| 958 |
+
"mean_token_accuracy": 0.7609355866909027,
|
| 959 |
+
"num_tokens": 16039387.0,
|
| 960 |
+
"step": 475
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"entropy": 0.933134937286377,
|
| 964 |
+
"epoch": 0.7535321821036107,
|
| 965 |
+
"grad_norm": 0.9548295867670389,
|
| 966 |
+
"learning_rate": 1.571703619276197e-05,
|
| 967 |
+
"loss": 0.9306,
|
| 968 |
+
"mean_token_accuracy": 0.7638529598712921,
|
| 969 |
+
"num_tokens": 16206056.0,
|
| 970 |
+
"step": 480
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"entropy": 0.9867853939533233,
|
| 974 |
+
"epoch": 0.7613814756671899,
|
| 975 |
+
"grad_norm": 1.0975317101204531,
|
| 976 |
+
"learning_rate": 1.5604044046606638e-05,
|
| 977 |
+
"loss": 0.9564,
|
| 978 |
+
"mean_token_accuracy": 0.7611208140850068,
|
| 979 |
+
"num_tokens": 16374570.0,
|
| 980 |
+
"step": 485
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"entropy": 0.9764430522918701,
|
| 984 |
+
"epoch": 0.7692307692307693,
|
| 985 |
+
"grad_norm": 1.0751088885823445,
|
| 986 |
+
"learning_rate": 1.548999905346234e-05,
|
| 987 |
+
"loss": 0.9784,
|
| 988 |
+
"mean_token_accuracy": 0.7565559089183808,
|
| 989 |
+
"num_tokens": 16534261.0,
|
| 990 |
+
"step": 490
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"entropy": 0.8999385595321655,
|
| 994 |
+
"epoch": 0.7770800627943485,
|
| 995 |
+
"grad_norm": 0.9254165712526795,
|
| 996 |
+
"learning_rate": 1.537492263927196e-05,
|
| 997 |
+
"loss": 0.8757,
|
| 998 |
+
"mean_token_accuracy": 0.7786536812782288,
|
| 999 |
+
"num_tokens": 16707250.0,
|
| 1000 |
+
"step": 495
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"entropy": 0.9715596318244935,
|
| 1004 |
+
"epoch": 0.7849293563579278,
|
| 1005 |
+
"grad_norm": 1.0969480560873561,
|
| 1006 |
+
"learning_rate": 1.5258836423754258e-05,
|
| 1007 |
+
"loss": 0.9645,
|
| 1008 |
+
"mean_token_accuracy": 0.7627908825874329,
|
| 1009 |
+
"num_tokens": 16876707.0,
|
| 1010 |
+
"step": 500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"entropy": 0.9281849145889283,
|
| 1014 |
+
"epoch": 0.792778649921507,
|
| 1015 |
+
"grad_norm": 1.0763995618308921,
|
| 1016 |
+
"learning_rate": 1.5141762216342107e-05,
|
| 1017 |
+
"loss": 0.9161,
|
| 1018 |
+
"mean_token_accuracy": 0.767313414812088,
|
| 1019 |
+
"num_tokens": 17034946.0,
|
| 1020 |
+
"step": 505
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"entropy": 0.9874213755130767,
|
| 1024 |
+
"epoch": 0.8006279434850864,
|
| 1025 |
+
"grad_norm": 1.0756471428565888,
|
| 1026 |
+
"learning_rate": 1.5023722012085098e-05,
|
| 1027 |
+
"loss": 0.9616,
|
| 1028 |
+
"mean_token_accuracy": 0.7579117834568023,
|
| 1029 |
+
"num_tokens": 17195368.0,
|
| 1030 |
+
"step": 510
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"entropy": 0.9650024175643921,
|
| 1034 |
+
"epoch": 0.8084772370486656,
|
| 1035 |
+
"grad_norm": 1.1457012639077697,
|
| 1036 |
+
"learning_rate": 1.4904737987517293e-05,
|
| 1037 |
+
"loss": 0.9521,
|
| 1038 |
+
"mean_token_accuracy": 0.7610667467117309,
|
| 1039 |
+
"num_tokens": 17349182.0,
|
| 1040 |
+
"step": 515
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"entropy": 0.9764435827732086,
|
| 1044 |
+
"epoch": 0.8163265306122449,
|
| 1045 |
+
"grad_norm": 1.0593991307798754,
|
| 1046 |
+
"learning_rate": 1.4784832496490824e-05,
|
| 1047 |
+
"loss": 0.9714,
|
| 1048 |
+
"mean_token_accuracy": 0.7524283647537231,
|
| 1049 |
+
"num_tokens": 17517371.0,
|
| 1050 |
+
"step": 520
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"entropy": 1.021035259962082,
|
| 1054 |
+
"epoch": 0.8241758241758241,
|
| 1055 |
+
"grad_norm": 1.1807579798750705,
|
| 1056 |
+
"learning_rate": 1.4664028065976245e-05,
|
| 1057 |
+
"loss": 0.9827,
|
| 1058 |
+
"mean_token_accuracy": 0.752406656742096,
|
| 1059 |
+
"num_tokens": 17691303.0,
|
| 1060 |
+
"step": 525
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"entropy": 0.9431538581848145,
|
| 1064 |
+
"epoch": 0.8320251177394035,
|
| 1065 |
+
"grad_norm": 1.1578643533684798,
|
| 1066 |
+
"learning_rate": 1.4542347391830308e-05,
|
| 1067 |
+
"loss": 0.9507,
|
| 1068 |
+
"mean_token_accuracy": 0.7616939425468445,
|
| 1069 |
+
"num_tokens": 17855630.0,
|
| 1070 |
+
"step": 530
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"entropy": 0.9709876954555512,
|
| 1074 |
+
"epoch": 0.8398744113029827,
|
| 1075 |
+
"grad_norm": 1.0499244127175924,
|
| 1076 |
+
"learning_rate": 1.4419813334532037e-05,
|
| 1077 |
+
"loss": 0.9543,
|
| 1078 |
+
"mean_token_accuracy": 0.7565864741802215,
|
| 1079 |
+
"num_tokens": 18013194.0,
|
| 1080 |
+
"step": 535
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"entropy": 0.9452216863632202,
|
| 1084 |
+
"epoch": 0.847723704866562,
|
| 1085 |
+
"grad_norm": 1.1743849862757485,
|
| 1086 |
+
"learning_rate": 1.4296448914887866e-05,
|
| 1087 |
+
"loss": 0.9157,
|
| 1088 |
+
"mean_token_accuracy": 0.7641405820846557,
|
| 1089 |
+
"num_tokens": 18176197.0,
|
| 1090 |
+
"step": 540
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"entropy": 0.9463288307189941,
|
| 1094 |
+
"epoch": 0.8555729984301413,
|
| 1095 |
+
"grad_norm": 1.0534691968313326,
|
| 1096 |
+
"learning_rate": 1.4172277309706677e-05,
|
| 1097 |
+
"loss": 0.9362,
|
| 1098 |
+
"mean_token_accuracy": 0.7646844327449799,
|
| 1099 |
+
"num_tokens": 18341936.0,
|
| 1100 |
+
"step": 545
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"entropy": 0.9249040186405182,
|
| 1104 |
+
"epoch": 0.8634222919937206,
|
| 1105 |
+
"grad_norm": 1.1519006839776653,
|
| 1106 |
+
"learning_rate": 1.4047321847445474e-05,
|
| 1107 |
+
"loss": 0.9273,
|
| 1108 |
+
"mean_token_accuracy": 0.7698991119861602,
|
| 1109 |
+
"num_tokens": 18506253.0,
|
| 1110 |
+
"step": 550
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"entropy": 0.9990038812160492,
|
| 1114 |
+
"epoch": 0.8712715855572999,
|
| 1115 |
+
"grad_norm": 1.0521190280976112,
|
| 1116 |
+
"learning_rate": 1.392160600382663e-05,
|
| 1117 |
+
"loss": 0.9652,
|
| 1118 |
+
"mean_token_accuracy": 0.7603400409221649,
|
| 1119 |
+
"num_tokens": 18669225.0,
|
| 1120 |
+
"step": 555
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"entropy": 0.976806640625,
|
| 1124 |
+
"epoch": 0.8791208791208791,
|
| 1125 |
+
"grad_norm": 0.9873791768090713,
|
| 1126 |
+
"learning_rate": 1.3795153397427426e-05,
|
| 1127 |
+
"loss": 0.9662,
|
| 1128 |
+
"mean_token_accuracy": 0.7616451740264892,
|
| 1129 |
+
"num_tokens": 18839983.0,
|
| 1130 |
+
"step": 560
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"entropy": 0.9923348844051361,
|
| 1134 |
+
"epoch": 0.8869701726844584,
|
| 1135 |
+
"grad_norm": 1.0387842421398645,
|
| 1136 |
+
"learning_rate": 1.3667987785242776e-05,
|
| 1137 |
+
"loss": 0.966,
|
| 1138 |
+
"mean_token_accuracy": 0.7592363655567169,
|
| 1139 |
+
"num_tokens": 19004847.0,
|
| 1140 |
+
"step": 565
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"entropy": 0.9111327111721039,
|
| 1144 |
+
"epoch": 0.8948194662480377,
|
| 1145 |
+
"grad_norm": 1.1091766330035904,
|
| 1146 |
+
"learning_rate": 1.3540133058221927e-05,
|
| 1147 |
+
"loss": 0.8894,
|
| 1148 |
+
"mean_token_accuracy": 0.7733675181865692,
|
| 1149 |
+
"num_tokens": 19164281.0,
|
| 1150 |
+
"step": 570
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"entropy": 0.953569757938385,
|
| 1154 |
+
"epoch": 0.902668759811617,
|
| 1155 |
+
"grad_norm": 1.1650697131207366,
|
| 1156 |
+
"learning_rate": 1.3411613236779996e-05,
|
| 1157 |
+
"loss": 0.9536,
|
| 1158 |
+
"mean_token_accuracy": 0.7601088881492615,
|
| 1159 |
+
"num_tokens": 19324666.0,
|
| 1160 |
+
"step": 575
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"entropy": 0.903479254245758,
|
| 1164 |
+
"epoch": 0.9105180533751962,
|
| 1165 |
+
"grad_norm": 1.0217432944115734,
|
| 1166 |
+
"learning_rate": 1.328245246628521e-05,
|
| 1167 |
+
"loss": 0.8938,
|
| 1168 |
+
"mean_token_accuracy": 0.7746713519096374,
|
| 1169 |
+
"num_tokens": 19487917.0,
|
| 1170 |
+
"step": 580
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"entropy": 0.9520645260810852,
|
| 1174 |
+
"epoch": 0.9183673469387755,
|
| 1175 |
+
"grad_norm": 1.0027838924807486,
|
| 1176 |
+
"learning_rate": 1.3152675012522629e-05,
|
| 1177 |
+
"loss": 0.9298,
|
| 1178 |
+
"mean_token_accuracy": 0.7667790412902832,
|
| 1179 |
+
"num_tokens": 19658140.0,
|
| 1180 |
+
"step": 585
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"entropy": 0.9800197124481201,
|
| 1184 |
+
"epoch": 0.9262166405023547,
|
| 1185 |
+
"grad_norm": 1.1007589371451882,
|
| 1186 |
+
"learning_rate": 1.302230525713527e-05,
|
| 1187 |
+
"loss": 0.9875,
|
| 1188 |
+
"mean_token_accuracy": 0.7534834921360016,
|
| 1189 |
+
"num_tokens": 19822714.0,
|
| 1190 |
+
"step": 590
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"entropy": 0.9933344900608063,
|
| 1194 |
+
"epoch": 0.9340659340659341,
|
| 1195 |
+
"grad_norm": 1.1374557236032408,
|
| 1196 |
+
"learning_rate": 1.2891367693043477e-05,
|
| 1197 |
+
"loss": 1.0055,
|
| 1198 |
+
"mean_token_accuracy": 0.7489987254142761,
|
| 1199 |
+
"num_tokens": 19978750.0,
|
| 1200 |
+
"step": 595
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"entropy": 0.8856159925460816,
|
| 1204 |
+
"epoch": 0.9419152276295133,
|
| 1205 |
+
"grad_norm": 1.0409967495386052,
|
| 1206 |
+
"learning_rate": 1.2759886919843354e-05,
|
| 1207 |
+
"loss": 0.8475,
|
| 1208 |
+
"mean_token_accuracy": 0.7862884163856506,
|
| 1209 |
+
"num_tokens": 20142475.0,
|
| 1210 |
+
"step": 600
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"entropy": 0.9482007265090943,
|
| 1214 |
+
"epoch": 0.9497645211930926,
|
| 1215 |
+
"grad_norm": 1.124699481009541,
|
| 1216 |
+
"learning_rate": 1.262788763918518e-05,
|
| 1217 |
+
"loss": 0.9616,
|
| 1218 |
+
"mean_token_accuracy": 0.7595623314380646,
|
| 1219 |
+
"num_tokens": 20304529.0,
|
| 1220 |
+
"step": 605
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"entropy": 1.0299462676048279,
|
| 1224 |
+
"epoch": 0.957613814756672,
|
| 1225 |
+
"grad_norm": 1.040395105371751,
|
| 1226 |
+
"learning_rate": 1.2495394650132628e-05,
|
| 1227 |
+
"loss": 0.9978,
|
| 1228 |
+
"mean_token_accuracy": 0.7538327634334564,
|
| 1229 |
+
"num_tokens": 20469501.0,
|
| 1230 |
+
"step": 610
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"entropy": 0.9115546882152558,
|
| 1234 |
+
"epoch": 0.9654631083202512,
|
| 1235 |
+
"grad_norm": 1.1694196529690937,
|
| 1236 |
+
"learning_rate": 1.2362432844503725e-05,
|
| 1237 |
+
"loss": 0.9081,
|
| 1238 |
+
"mean_token_accuracy": 0.7722381055355072,
|
| 1239 |
+
"num_tokens": 20640991.0,
|
| 1240 |
+
"step": 615
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"entropy": 0.9577880501747131,
|
| 1244 |
+
"epoch": 0.9733124018838305,
|
| 1245 |
+
"grad_norm": 1.103459451739408,
|
| 1246 |
+
"learning_rate": 1.222902720219433e-05,
|
| 1247 |
+
"loss": 0.9528,
|
| 1248 |
+
"mean_token_accuracy": 0.7636137962341308,
|
| 1249 |
+
"num_tokens": 20810696.0,
|
| 1250 |
+
"step": 620
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"entropy": 0.935895311832428,
|
| 1254 |
+
"epoch": 0.9811616954474097,
|
| 1255 |
+
"grad_norm": 0.9487913233732936,
|
| 1256 |
+
"learning_rate": 1.209520278648512e-05,
|
| 1257 |
+
"loss": 0.9021,
|
| 1258 |
+
"mean_token_accuracy": 0.7763786315917969,
|
| 1259 |
+
"num_tokens": 20990530.0,
|
| 1260 |
+
"step": 625
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"entropy": 0.8163468480110169,
|
| 1264 |
+
"epoch": 0.989010989010989,
|
| 1265 |
+
"grad_norm": 0.9330276032956102,
|
| 1266 |
+
"learning_rate": 1.1960984739332851e-05,
|
| 1267 |
+
"loss": 0.7945,
|
| 1268 |
+
"mean_token_accuracy": 0.7963767290115357,
|
| 1269 |
+
"num_tokens": 21168734.0,
|
| 1270 |
+
"step": 630
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"entropy": 0.9846008777618408,
|
| 1274 |
+
"epoch": 0.9968602825745683,
|
| 1275 |
+
"grad_norm": 1.0256914219636313,
|
| 1276 |
+
"learning_rate": 1.1826398276646897e-05,
|
| 1277 |
+
"loss": 0.9983,
|
| 1278 |
+
"mean_token_accuracy": 0.7496028244495392,
|
| 1279 |
+
"num_tokens": 21327447.0,
|
| 1280 |
+
"step": 635
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"entropy": 0.9138310849666595,
|
| 1284 |
+
"epoch": 1.0047095761381475,
|
| 1285 |
+
"grad_norm": 1.0812559144942584,
|
| 1286 |
+
"learning_rate": 1.1691468683551865e-05,
|
| 1287 |
+
"loss": 0.8344,
|
| 1288 |
+
"mean_token_accuracy": 0.7886788547039032,
|
| 1289 |
+
"num_tokens": 21499715.0,
|
| 1290 |
+
"step": 640
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"entropy": 0.8012929916381836,
|
| 1294 |
+
"epoch": 1.012558869701727,
|
| 1295 |
+
"grad_norm": 1.0532908535496344,
|
| 1296 |
+
"learning_rate": 1.1556221309637204e-05,
|
| 1297 |
+
"loss": 0.7737,
|
| 1298 |
+
"mean_token_accuracy": 0.7994522035121918,
|
| 1299 |
+
"num_tokens": 21664223.0,
|
| 1300 |
+
"step": 645
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"entropy": 0.8007775723934174,
|
| 1304 |
+
"epoch": 1.0204081632653061,
|
| 1305 |
+
"grad_norm": 1.2247973138776649,
|
| 1306 |
+
"learning_rate": 1.1420681564194694e-05,
|
| 1307 |
+
"loss": 0.778,
|
| 1308 |
+
"mean_token_accuracy": 0.7998305559158325,
|
| 1309 |
+
"num_tokens": 21826895.0,
|
| 1310 |
+
"step": 650
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"entropy": 0.8300498962402344,
|
| 1314 |
+
"epoch": 1.0282574568288854,
|
| 1315 |
+
"grad_norm": 0.9966548585427317,
|
| 1316 |
+
"learning_rate": 1.1284874911444763e-05,
|
| 1317 |
+
"loss": 0.7699,
|
| 1318 |
+
"mean_token_accuracy": 0.8034215152263642,
|
| 1319 |
+
"num_tokens": 21993866.0,
|
| 1320 |
+
"step": 655
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"entropy": 0.7994357228279114,
|
| 1324 |
+
"epoch": 1.0361067503924646,
|
| 1325 |
+
"grad_norm": 1.0501870540204665,
|
| 1326 |
+
"learning_rate": 1.1148826865752445e-05,
|
| 1327 |
+
"loss": 0.7898,
|
| 1328 |
+
"mean_token_accuracy": 0.7990375220775604,
|
| 1329 |
+
"num_tokens": 22163013.0,
|
| 1330 |
+
"step": 660
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"entropy": 0.801983118057251,
|
| 1334 |
+
"epoch": 1.043956043956044,
|
| 1335 |
+
"grad_norm": 0.9191498252681857,
|
| 1336 |
+
"learning_rate": 1.1012562986833909e-05,
|
| 1337 |
+
"loss": 0.7637,
|
| 1338 |
+
"mean_token_accuracy": 0.8037278771400451,
|
| 1339 |
+
"num_tokens": 22336561.0,
|
| 1340 |
+
"step": 665
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"entropy": 0.7570578813552856,
|
| 1344 |
+
"epoch": 1.0518053375196232,
|
| 1345 |
+
"grad_norm": 0.9832626897663099,
|
| 1346 |
+
"learning_rate": 1.0876108874954498e-05,
|
| 1347 |
+
"loss": 0.7129,
|
| 1348 |
+
"mean_token_accuracy": 0.8170925438404083,
|
| 1349 |
+
"num_tokens": 22503077.0,
|
| 1350 |
+
"step": 670
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"entropy": 0.8149272561073303,
|
| 1354 |
+
"epoch": 1.0596546310832025,
|
| 1355 |
+
"grad_norm": 1.1304521768941131,
|
| 1356 |
+
"learning_rate": 1.0739490166119155e-05,
|
| 1357 |
+
"loss": 0.7865,
|
| 1358 |
+
"mean_token_accuracy": 0.8010401368141175,
|
| 1359 |
+
"num_tokens": 22675010.0,
|
| 1360 |
+
"step": 675
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"entropy": 0.7653459012508392,
|
| 1364 |
+
"epoch": 1.0675039246467817,
|
| 1365 |
+
"grad_norm": 1.4761235017107492,
|
| 1366 |
+
"learning_rate": 1.060273252725609e-05,
|
| 1367 |
+
"loss": 0.7315,
|
| 1368 |
+
"mean_token_accuracy": 0.8124203622341156,
|
| 1369 |
+
"num_tokens": 22837858.0,
|
| 1370 |
+
"step": 680
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"entropy": 0.8208756923675538,
|
| 1374 |
+
"epoch": 1.0753532182103611,
|
| 1375 |
+
"grad_norm": 1.0741850631374394,
|
| 1376 |
+
"learning_rate": 1.0465861651394673e-05,
|
| 1377 |
+
"loss": 0.7885,
|
| 1378 |
+
"mean_token_accuracy": 0.7971214592456818,
|
| 1379 |
+
"num_tokens": 22998267.0,
|
| 1380 |
+
"step": 685
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"entropy": 0.8316072285175323,
|
| 1384 |
+
"epoch": 1.0832025117739403,
|
| 1385 |
+
"grad_norm": 1.05989384509782,
|
| 1386 |
+
"learning_rate": 1.0328903252838415e-05,
|
| 1387 |
+
"loss": 0.8138,
|
| 1388 |
+
"mean_token_accuracy": 0.7903102576732636,
|
| 1389 |
+
"num_tokens": 23163250.0,
|
| 1390 |
+
"step": 690
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"entropy": 0.8117113530635833,
|
| 1394 |
+
"epoch": 1.0910518053375196,
|
| 1395 |
+
"grad_norm": 0.9620563544580413,
|
| 1396 |
+
"learning_rate": 1.0191883062333964e-05,
|
| 1397 |
+
"loss": 0.7643,
|
| 1398 |
+
"mean_token_accuracy": 0.8044250011444092,
|
| 1399 |
+
"num_tokens": 23338134.0,
|
| 1400 |
+
"step": 695
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"entropy": 0.7617044448852539,
|
| 1404 |
+
"epoch": 1.098901098901099,
|
| 1405 |
+
"grad_norm": 0.9834039954530569,
|
| 1406 |
+
"learning_rate": 1.0054826822236983e-05,
|
| 1407 |
+
"loss": 0.7273,
|
| 1408 |
+
"mean_token_accuracy": 0.8108560085296631,
|
| 1409 |
+
"num_tokens": 23499035.0,
|
| 1410 |
+
"step": 700
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"entropy": 0.8119458794593811,
|
| 1414 |
+
"epoch": 1.1067503924646782,
|
| 1415 |
+
"grad_norm": 0.8922309018148304,
|
| 1416 |
+
"learning_rate": 9.917760281675867e-06,
|
| 1417 |
+
"loss": 0.8056,
|
| 1418 |
+
"mean_token_accuracy": 0.7917751908302307,
|
| 1419 |
+
"num_tokens": 23666069.0,
|
| 1420 |
+
"step": 705
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"entropy": 0.796909236907959,
|
| 1424 |
+
"epoch": 1.1145996860282574,
|
| 1425 |
+
"grad_norm": 1.0705197733359055,
|
| 1426 |
+
"learning_rate": 9.780709191714187e-06,
|
| 1427 |
+
"loss": 0.7719,
|
| 1428 |
+
"mean_token_accuracy": 0.8027232766151429,
|
| 1429 |
+
"num_tokens": 23835714.0,
|
| 1430 |
+
"step": 710
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"entropy": 0.8005717933177948,
|
| 1434 |
+
"epoch": 1.1224489795918366,
|
| 1435 |
+
"grad_norm": 1.0642859073819562,
|
| 1436 |
+
"learning_rate": 9.643699300512781e-06,
|
| 1437 |
+
"loss": 0.7706,
|
| 1438 |
+
"mean_token_accuracy": 0.8004081726074219,
|
| 1439 |
+
"num_tokens": 23995150.0,
|
| 1440 |
+
"step": 715
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"entropy": 0.855769693851471,
|
| 1444 |
+
"epoch": 1.130298273155416,
|
| 1445 |
+
"grad_norm": 1.0460745966572866,
|
| 1446 |
+
"learning_rate": 9.506756348492348e-06,
|
| 1447 |
+
"loss": 0.8374,
|
| 1448 |
+
"mean_token_accuracy": 0.7850587904453278,
|
| 1449 |
+
"num_tokens": 24158757.0,
|
| 1450 |
+
"step": 720
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"entropy": 0.7870798826217651,
|
| 1454 |
+
"epoch": 1.1381475667189953,
|
| 1455 |
+
"grad_norm": 0.9976832224332017,
|
| 1456 |
+
"learning_rate": 9.369906063497547e-06,
|
| 1457 |
+
"loss": 0.7445,
|
| 1458 |
+
"mean_token_accuracy": 0.8053227722644806,
|
| 1459 |
+
"num_tokens": 24326311.0,
|
| 1460 |
+
"step": 725
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"entropy": 0.8291393995285035,
|
| 1464 |
+
"epoch": 1.1459968602825745,
|
| 1465 |
+
"grad_norm": 0.996673672294385,
|
| 1466 |
+
"learning_rate": 9.233174155963432e-06,
|
| 1467 |
+
"loss": 0.8141,
|
| 1468 |
+
"mean_token_accuracy": 0.7964518368244171,
|
| 1469 |
+
"num_tokens": 24492165.0,
|
| 1470 |
+
"step": 730
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"entropy": 0.7422019600868225,
|
| 1474 |
+
"epoch": 1.1538461538461537,
|
| 1475 |
+
"grad_norm": 1.0756638757031713,
|
| 1476 |
+
"learning_rate": 9.096586314085162e-06,
|
| 1477 |
+
"loss": 0.7197,
|
| 1478 |
+
"mean_token_accuracy": 0.8132814466953278,
|
| 1479 |
+
"num_tokens": 24664507.0,
|
| 1480 |
+
"step": 735
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"entropy": 0.7962768793106079,
|
| 1484 |
+
"epoch": 1.1616954474097332,
|
| 1485 |
+
"grad_norm": 1.1377928625620208,
|
| 1486 |
+
"learning_rate": 8.960168198991885e-06,
|
| 1487 |
+
"loss": 0.7557,
|
| 1488 |
+
"mean_token_accuracy": 0.8038122892379761,
|
| 1489 |
+
"num_tokens": 24824172.0,
|
| 1490 |
+
"step": 740
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"entropy": 0.6977714002132416,
|
| 1494 |
+
"epoch": 1.1695447409733124,
|
| 1495 |
+
"grad_norm": 0.9708645265144701,
|
| 1496 |
+
"learning_rate": 8.823945439925725e-06,
|
| 1497 |
+
"loss": 0.6836,
|
| 1498 |
+
"mean_token_accuracy": 0.8212838768959045,
|
| 1499 |
+
"num_tokens": 24992172.0,
|
| 1500 |
+
"step": 745
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"entropy": 0.7833775043487549,
|
| 1504 |
+
"epoch": 1.1773940345368916,
|
| 1505 |
+
"grad_norm": 1.1739278827813637,
|
| 1506 |
+
"learning_rate": 8.687943629426725e-06,
|
| 1507 |
+
"loss": 0.758,
|
| 1508 |
+
"mean_token_accuracy": 0.8031582951545715,
|
| 1509 |
+
"num_tokens": 25159967.0,
|
| 1510 |
+
"step": 750
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"entropy": 0.8251801669597626,
|
| 1514 |
+
"epoch": 1.185243328100471,
|
| 1515 |
+
"grad_norm": 1.1814132581564163,
|
| 1516 |
+
"learning_rate": 8.552188318524737e-06,
|
| 1517 |
+
"loss": 0.8154,
|
| 1518 |
+
"mean_token_accuracy": 0.7922390282154084,
|
| 1519 |
+
"num_tokens": 25318992.0,
|
| 1520 |
+
"step": 755
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"entropy": 0.7560604989528656,
|
| 1524 |
+
"epoch": 1.1930926216640503,
|
| 1525 |
+
"grad_norm": 0.8951739538342747,
|
| 1526 |
+
"learning_rate": 8.416705011939052e-06,
|
| 1527 |
+
"loss": 0.7366,
|
| 1528 |
+
"mean_token_accuracy": 0.8119826078414917,
|
| 1529 |
+
"num_tokens": 25493572.0,
|
| 1530 |
+
"step": 760
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"entropy": 0.8065792441368103,
|
| 1534 |
+
"epoch": 1.2009419152276295,
|
| 1535 |
+
"grad_norm": 1.1315496393007614,
|
| 1536 |
+
"learning_rate": 8.281519163286772e-06,
|
| 1537 |
+
"loss": 0.7634,
|
| 1538 |
+
"mean_token_accuracy": 0.802753496170044,
|
| 1539 |
+
"num_tokens": 25665824.0,
|
| 1540 |
+
"step": 765
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"entropy": 0.8229409158229828,
|
| 1544 |
+
"epoch": 1.2087912087912087,
|
| 1545 |
+
"grad_norm": 1.0384454910962662,
|
| 1546 |
+
"learning_rate": 8.146656170300772e-06,
|
| 1547 |
+
"loss": 0.7831,
|
| 1548 |
+
"mean_token_accuracy": 0.8028485774993896,
|
| 1549 |
+
"num_tokens": 25826600.0,
|
| 1550 |
+
"step": 770
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"entropy": 0.8109376728534698,
|
| 1554 |
+
"epoch": 1.2166405023547882,
|
| 1555 |
+
"grad_norm": 1.0826233765035977,
|
| 1556 |
+
"learning_rate": 8.01214137005815e-06,
|
| 1557 |
+
"loss": 0.7867,
|
| 1558 |
+
"mean_token_accuracy": 0.797510153055191,
|
| 1559 |
+
"num_tokens": 25990358.0,
|
| 1560 |
+
"step": 775
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"entropy": 0.7986651659011841,
|
| 1564 |
+
"epoch": 1.2244897959183674,
|
| 1565 |
+
"grad_norm": 0.9510634195804456,
|
| 1566 |
+
"learning_rate": 7.878000034220092e-06,
|
| 1567 |
+
"loss": 0.7693,
|
| 1568 |
+
"mean_token_accuracy": 0.801447582244873,
|
| 1569 |
+
"num_tokens": 26166469.0,
|
| 1570 |
+
"step": 780
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"entropy": 0.7796730637550354,
|
| 1574 |
+
"epoch": 1.2323390894819466,
|
| 1575 |
+
"grad_norm": 0.9674095865305092,
|
| 1576 |
+
"learning_rate": 7.74425736428401e-06,
|
| 1577 |
+
"loss": 0.7547,
|
| 1578 |
+
"mean_token_accuracy": 0.808059710264206,
|
| 1579 |
+
"num_tokens": 26339746.0,
|
| 1580 |
+
"step": 785
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"entropy": 0.7995173811912537,
|
| 1584 |
+
"epoch": 1.2401883830455258,
|
| 1585 |
+
"grad_norm": 1.0518387365429118,
|
| 1586 |
+
"learning_rate": 7.6109384868488646e-06,
|
| 1587 |
+
"loss": 0.7883,
|
| 1588 |
+
"mean_token_accuracy": 0.8016106963157654,
|
| 1589 |
+
"num_tokens": 26505540.0,
|
| 1590 |
+
"step": 790
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"entropy": 0.8127647697925567,
|
| 1594 |
+
"epoch": 1.2480376766091053,
|
| 1595 |
+
"grad_norm": 0.9711975789074598,
|
| 1596 |
+
"learning_rate": 7.478068448894577e-06,
|
| 1597 |
+
"loss": 0.7656,
|
| 1598 |
+
"mean_token_accuracy": 0.801650446653366,
|
| 1599 |
+
"num_tokens": 26669521.0,
|
| 1600 |
+
"step": 795
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"entropy": 0.7887676239013672,
|
| 1604 |
+
"epoch": 1.2558869701726845,
|
| 1605 |
+
"grad_norm": 0.9327700469775977,
|
| 1606 |
+
"learning_rate": 7.3456722130763665e-06,
|
| 1607 |
+
"loss": 0.7799,
|
| 1608 |
+
"mean_token_accuracy": 0.8046572506427765,
|
| 1609 |
+
"num_tokens": 26848214.0,
|
| 1610 |
+
"step": 800
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"entropy": 0.8123725712299347,
|
| 1614 |
+
"epoch": 1.2637362637362637,
|
| 1615 |
+
"grad_norm": 1.0537567232285652,
|
| 1616 |
+
"learning_rate": 7.213774653034958e-06,
|
| 1617 |
+
"loss": 0.7953,
|
| 1618 |
+
"mean_token_accuracy": 0.7974987208843232,
|
| 1619 |
+
"num_tokens": 27015893.0,
|
| 1620 |
+
"step": 805
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"entropy": 0.8397608995437622,
|
| 1624 |
+
"epoch": 1.2715855572998431,
|
| 1625 |
+
"grad_norm": 1.104054891980114,
|
| 1626 |
+
"learning_rate": 7.082400548723505e-06,
|
| 1627 |
+
"loss": 0.8154,
|
| 1628 |
+
"mean_token_accuracy": 0.7920099139213562,
|
| 1629 |
+
"num_tokens": 27180368.0,
|
| 1630 |
+
"step": 810
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"entropy": 0.7759601533412933,
|
| 1634 |
+
"epoch": 1.2794348508634223,
|
| 1635 |
+
"grad_norm": 1.0490176058805003,
|
| 1636 |
+
"learning_rate": 6.951574581752111e-06,
|
| 1637 |
+
"loss": 0.7502,
|
| 1638 |
+
"mean_token_accuracy": 0.807463413476944,
|
| 1639 |
+
"num_tokens": 27343682.0,
|
| 1640 |
+
"step": 815
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"entropy": 0.8411307275295258,
|
| 1644 |
+
"epoch": 1.2872841444270016,
|
| 1645 |
+
"grad_norm": 1.166847601682725,
|
| 1646 |
+
"learning_rate": 6.8213213307508205e-06,
|
| 1647 |
+
"loss": 0.7991,
|
| 1648 |
+
"mean_token_accuracy": 0.7980861485004425,
|
| 1649 |
+
"num_tokens": 27516238.0,
|
| 1650 |
+
"step": 820
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"entropy": 0.7412167847156524,
|
| 1654 |
+
"epoch": 1.2951334379905808,
|
| 1655 |
+
"grad_norm": 1.0923464885518581,
|
| 1656 |
+
"learning_rate": 6.6916652667519855e-06,
|
| 1657 |
+
"loss": 0.7196,
|
| 1658 |
+
"mean_token_accuracy": 0.8138726353645325,
|
| 1659 |
+
"num_tokens": 27688076.0,
|
| 1660 |
+
"step": 825
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"entropy": 0.82211754322052,
|
| 1664 |
+
"epoch": 1.30298273155416,
|
| 1665 |
+
"grad_norm": 1.160628687991898,
|
| 1666 |
+
"learning_rate": 6.562630748592794e-06,
|
| 1667 |
+
"loss": 0.795,
|
| 1668 |
+
"mean_token_accuracy": 0.7955183386802673,
|
| 1669 |
+
"num_tokens": 27851949.0,
|
| 1670 |
+
"step": 830
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"entropy": 0.8261380314826965,
|
| 1674 |
+
"epoch": 1.3108320251177394,
|
| 1675 |
+
"grad_norm": 1.0586659878919926,
|
| 1676 |
+
"learning_rate": 6.434242018338948e-06,
|
| 1677 |
+
"loss": 0.8086,
|
| 1678 |
+
"mean_token_accuracy": 0.7943514943122864,
|
| 1679 |
+
"num_tokens": 28013184.0,
|
| 1680 |
+
"step": 835
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"entropy": 0.7742564082145691,
|
| 1684 |
+
"epoch": 1.3186813186813187,
|
| 1685 |
+
"grad_norm": 1.1470217200167852,
|
| 1686 |
+
"learning_rate": 6.3065231967302055e-06,
|
| 1687 |
+
"loss": 0.7459,
|
| 1688 |
+
"mean_token_accuracy": 0.8088241875171661,
|
| 1689 |
+
"num_tokens": 28180563.0,
|
| 1690 |
+
"step": 840
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"entropy": 0.8125581502914428,
|
| 1694 |
+
"epoch": 1.3265306122448979,
|
| 1695 |
+
"grad_norm": 1.180920424036568,
|
| 1696 |
+
"learning_rate": 6.179498278648766e-06,
|
| 1697 |
+
"loss": 0.7844,
|
| 1698 |
+
"mean_token_accuracy": 0.7975386023521424,
|
| 1699 |
+
"num_tokens": 28348933.0,
|
| 1700 |
+
"step": 845
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"entropy": 0.85399090051651,
|
| 1704 |
+
"epoch": 1.3343799058084773,
|
| 1705 |
+
"grad_norm": 1.017696845491611,
|
| 1706 |
+
"learning_rate": 6.053191128611298e-06,
|
| 1707 |
+
"loss": 0.8317,
|
| 1708 |
+
"mean_token_accuracy": 0.7885974526405335,
|
| 1709 |
+
"num_tokens": 28510060.0,
|
| 1710 |
+
"step": 850
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"entropy": 0.7950326681137085,
|
| 1714 |
+
"epoch": 1.3422291993720565,
|
| 1715 |
+
"grad_norm": 0.9519056545968589,
|
| 1716 |
+
"learning_rate": 5.927625476285426e-06,
|
| 1717 |
+
"loss": 0.7737,
|
| 1718 |
+
"mean_token_accuracy": 0.798821222782135,
|
| 1719 |
+
"num_tokens": 28673404.0,
|
| 1720 |
+
"step": 855
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"entropy": 0.7562436521053314,
|
| 1724 |
+
"epoch": 1.3500784929356358,
|
| 1725 |
+
"grad_norm": 1.1866713225060308,
|
| 1726 |
+
"learning_rate": 5.802824912031588e-06,
|
| 1727 |
+
"loss": 0.7237,
|
| 1728 |
+
"mean_token_accuracy": 0.8115784227848053,
|
| 1729 |
+
"num_tokens": 28846665.0,
|
| 1730 |
+
"step": 860
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"entropy": 0.8442178428173065,
|
| 1734 |
+
"epoch": 1.3579277864992152,
|
| 1735 |
+
"grad_norm": 1.0527704560360036,
|
| 1736 |
+
"learning_rate": 5.678812882471047e-06,
|
| 1737 |
+
"loss": 0.8088,
|
| 1738 |
+
"mean_token_accuracy": 0.7886203587055206,
|
| 1739 |
+
"num_tokens": 29016122.0,
|
| 1740 |
+
"step": 865
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"entropy": 0.7864629566669464,
|
| 1744 |
+
"epoch": 1.3657770800627944,
|
| 1745 |
+
"grad_norm": 1.0568094694328418,
|
| 1746 |
+
"learning_rate": 5.555612686080909e-06,
|
| 1747 |
+
"loss": 0.7514,
|
| 1748 |
+
"mean_token_accuracy": 0.8058109283447266,
|
| 1749 |
+
"num_tokens": 29180726.0,
|
| 1750 |
+
"step": 870
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"entropy": 0.8194212853908539,
|
| 1754 |
+
"epoch": 1.3736263736263736,
|
| 1755 |
+
"grad_norm": 1.1492348193285462,
|
| 1756 |
+
"learning_rate": 5.4332474688169766e-06,
|
| 1757 |
+
"loss": 0.7897,
|
| 1758 |
+
"mean_token_accuracy": 0.7988565325736999,
|
| 1759 |
+
"num_tokens": 29354903.0,
|
| 1760 |
+
"step": 875
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"entropy": 0.7987347841262817,
|
| 1764 |
+
"epoch": 1.3814756671899528,
|
| 1765 |
+
"grad_norm": 0.8999725861039704,
|
| 1766 |
+
"learning_rate": 5.311740219765247e-06,
|
| 1767 |
+
"loss": 0.7757,
|
| 1768 |
+
"mean_token_accuracy": 0.80171217918396,
|
| 1769 |
+
"num_tokens": 29537609.0,
|
| 1770 |
+
"step": 880
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"entropy": 0.8379440605640411,
|
| 1774 |
+
"epoch": 1.389324960753532,
|
| 1775 |
+
"grad_norm": 1.1068737293584308,
|
| 1776 |
+
"learning_rate": 5.191113766822905e-06,
|
| 1777 |
+
"loss": 0.8081,
|
| 1778 |
+
"mean_token_accuracy": 0.7938673734664917,
|
| 1779 |
+
"num_tokens": 29703501.0,
|
| 1780 |
+
"step": 885
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"entropy": 0.8395914256572723,
|
| 1784 |
+
"epoch": 1.3971742543171115,
|
| 1785 |
+
"grad_norm": 0.9692236763696376,
|
| 1786 |
+
"learning_rate": 5.071390772409579e-06,
|
| 1787 |
+
"loss": 0.8263,
|
| 1788 |
+
"mean_token_accuracy": 0.7933382868766785,
|
| 1789 |
+
"num_tokens": 29873409.0,
|
| 1790 |
+
"step": 890
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"entropy": 0.8258692562580109,
|
| 1794 |
+
"epoch": 1.4050235478806907,
|
| 1795 |
+
"grad_norm": 1.1584118626238182,
|
| 1796 |
+
"learning_rate": 4.952593729209671e-06,
|
| 1797 |
+
"loss": 0.8016,
|
| 1798 |
+
"mean_token_accuracy": 0.7933484017848969,
|
| 1799 |
+
"num_tokens": 30027021.0,
|
| 1800 |
+
"step": 895
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"entropy": 0.7889597296714783,
|
| 1804 |
+
"epoch": 1.41287284144427,
|
| 1805 |
+
"grad_norm": 1.072165602590816,
|
| 1806 |
+
"learning_rate": 4.834744955946631e-06,
|
| 1807 |
+
"loss": 0.7632,
|
| 1808 |
+
"mean_token_accuracy": 0.8002647578716278,
|
| 1809 |
+
"num_tokens": 30185182.0,
|
| 1810 |
+
"step": 900
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"entropy": 0.8507044613361359,
|
| 1814 |
+
"epoch": 1.4207221350078494,
|
| 1815 |
+
"grad_norm": 1.103567455594152,
|
| 1816 |
+
"learning_rate": 4.717866593189847e-06,
|
| 1817 |
+
"loss": 0.8193,
|
| 1818 |
+
"mean_token_accuracy": 0.7897421777248382,
|
| 1819 |
+
"num_tokens": 30339297.0,
|
| 1820 |
+
"step": 905
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"entropy": 0.8800127923488616,
|
| 1824 |
+
"epoch": 1.4285714285714286,
|
| 1825 |
+
"grad_norm": 1.1458812579100477,
|
| 1826 |
+
"learning_rate": 4.60198059919505e-06,
|
| 1827 |
+
"loss": 0.8706,
|
| 1828 |
+
"mean_token_accuracy": 0.7778860569000244,
|
| 1829 |
+
"num_tokens": 30493963.0,
|
| 1830 |
+
"step": 910
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"entropy": 0.7648905456066132,
|
| 1834 |
+
"epoch": 1.4364207221350078,
|
| 1835 |
+
"grad_norm": 1.0500741631200436,
|
| 1836 |
+
"learning_rate": 4.487108745778958e-06,
|
| 1837 |
+
"loss": 0.7182,
|
| 1838 |
+
"mean_token_accuracy": 0.8138745665550232,
|
| 1839 |
+
"num_tokens": 30666803.0,
|
| 1840 |
+
"step": 915
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"entropy": 0.7712486982345581,
|
| 1844 |
+
"epoch": 1.4442700156985873,
|
| 1845 |
+
"grad_norm": 1.0542089160734922,
|
| 1846 |
+
"learning_rate": 4.373272614228932e-06,
|
| 1847 |
+
"loss": 0.7553,
|
| 1848 |
+
"mean_token_accuracy": 0.8085428476333618,
|
| 1849 |
+
"num_tokens": 30836059.0,
|
| 1850 |
+
"step": 920
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"entropy": 0.7824968814849853,
|
| 1854 |
+
"epoch": 1.4521193092621665,
|
| 1855 |
+
"grad_norm": 1.1739948319837288,
|
| 1856 |
+
"learning_rate": 4.260493591248458e-06,
|
| 1857 |
+
"loss": 0.7488,
|
| 1858 |
+
"mean_token_accuracy": 0.8080036818981171,
|
| 1859 |
+
"num_tokens": 31002260.0,
|
| 1860 |
+
"step": 925
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"entropy": 0.8732044517993927,
|
| 1864 |
+
"epoch": 1.4599686028257457,
|
| 1865 |
+
"grad_norm": 1.218588367639695,
|
| 1866 |
+
"learning_rate": 4.148792864939164e-06,
|
| 1867 |
+
"loss": 0.8311,
|
| 1868 |
+
"mean_token_accuracy": 0.7848361492156982,
|
| 1869 |
+
"num_tokens": 31168586.0,
|
| 1870 |
+
"step": 930
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"entropy": 0.7919686675071717,
|
| 1874 |
+
"epoch": 1.467817896389325,
|
| 1875 |
+
"grad_norm": 1.0395587921432232,
|
| 1876 |
+
"learning_rate": 4.038191420820139e-06,
|
| 1877 |
+
"loss": 0.7583,
|
| 1878 |
+
"mean_token_accuracy": 0.8057900071144104,
|
| 1879 |
+
"num_tokens": 31329810.0,
|
| 1880 |
+
"step": 935
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"entropy": 0.857411801815033,
|
| 1884 |
+
"epoch": 1.4756671899529041,
|
| 1885 |
+
"grad_norm": 1.2129732997330736,
|
| 1886 |
+
"learning_rate": 3.92871003788535e-06,
|
| 1887 |
+
"loss": 0.8246,
|
| 1888 |
+
"mean_token_accuracy": 0.7912562787532806,
|
| 1889 |
+
"num_tokens": 31491966.0,
|
| 1890 |
+
"step": 940
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"entropy": 0.8164745211601258,
|
| 1894 |
+
"epoch": 1.4835164835164836,
|
| 1895 |
+
"grad_norm": 1.0962169271219249,
|
| 1896 |
+
"learning_rate": 3.820369284699823e-06,
|
| 1897 |
+
"loss": 0.7797,
|
| 1898 |
+
"mean_token_accuracy": 0.7972842991352082,
|
| 1899 |
+
"num_tokens": 31654493.0,
|
| 1900 |
+
"step": 945
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"entropy": 0.8125701963901519,
|
| 1904 |
+
"epoch": 1.4913657770800628,
|
| 1905 |
+
"grad_norm": 1.0866004282193915,
|
| 1906 |
+
"learning_rate": 3.713189515535368e-06,
|
| 1907 |
+
"loss": 0.7922,
|
| 1908 |
+
"mean_token_accuracy": 0.7976612865924835,
|
| 1909 |
+
"num_tokens": 31819877.0,
|
| 1910 |
+
"step": 950
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"entropy": 0.7080917954444885,
|
| 1914 |
+
"epoch": 1.499215070643642,
|
| 1915 |
+
"grad_norm": 0.9180013433354448,
|
| 1916 |
+
"learning_rate": 3.607190866546578e-06,
|
| 1917 |
+
"loss": 0.6795,
|
| 1918 |
+
"mean_token_accuracy": 0.8217414200305939,
|
| 1919 |
+
"num_tokens": 32000700.0,
|
| 1920 |
+
"step": 955
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"entropy": 0.7802863955497742,
|
| 1924 |
+
"epoch": 1.5070643642072215,
|
| 1925 |
+
"grad_norm": 1.0444040538958685,
|
| 1926 |
+
"learning_rate": 3.502393251987776e-06,
|
| 1927 |
+
"loss": 0.7541,
|
| 1928 |
+
"mean_token_accuracy": 0.8063533782958985,
|
| 1929 |
+
"num_tokens": 32162701.0,
|
| 1930 |
+
"step": 960
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"entropy": 0.7960240602493286,
|
| 1934 |
+
"epoch": 1.5149136577708007,
|
| 1935 |
+
"grad_norm": 1.0318984478840416,
|
| 1936 |
+
"learning_rate": 3.3988163604716928e-06,
|
| 1937 |
+
"loss": 0.7762,
|
| 1938 |
+
"mean_token_accuracy": 0.8041753470897675,
|
| 1939 |
+
"num_tokens": 32325569.0,
|
| 1940 |
+
"step": 965
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"entropy": 0.8061594307422638,
|
| 1944 |
+
"epoch": 1.5227629513343799,
|
| 1945 |
+
"grad_norm": 1.012431871323513,
|
| 1946 |
+
"learning_rate": 3.296479651270502e-06,
|
| 1947 |
+
"loss": 0.7674,
|
| 1948 |
+
"mean_token_accuracy": 0.8041636168956756,
|
| 1949 |
+
"num_tokens": 32488499.0,
|
| 1950 |
+
"step": 970
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"entropy": 0.7571884632110596,
|
| 1954 |
+
"epoch": 1.5306122448979593,
|
| 1955 |
+
"grad_norm": 1.1234093176199544,
|
| 1956 |
+
"learning_rate": 3.195402350659945e-06,
|
| 1957 |
+
"loss": 0.7382,
|
| 1958 |
+
"mean_token_accuracy": 0.8137228667736054,
|
| 1959 |
+
"num_tokens": 32655906.0,
|
| 1960 |
+
"step": 975
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"entropy": 0.8204948246479035,
|
| 1964 |
+
"epoch": 1.5384615384615383,
|
| 1965 |
+
"grad_norm": 1.08811573994006,
|
| 1966 |
+
"learning_rate": 3.0956034483072573e-06,
|
| 1967 |
+
"loss": 0.7989,
|
| 1968 |
+
"mean_token_accuracy": 0.7967490673065185,
|
| 1969 |
+
"num_tokens": 32822319.0,
|
| 1970 |
+
"step": 980
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"entropy": 0.8921033620834351,
|
| 1974 |
+
"epoch": 1.5463108320251178,
|
| 1975 |
+
"grad_norm": 1.1469019706898658,
|
| 1976 |
+
"learning_rate": 2.997101693703518e-06,
|
| 1977 |
+
"loss": 0.8717,
|
| 1978 |
+
"mean_token_accuracy": 0.7777008473873138,
|
| 1979 |
+
"num_tokens": 32979290.0,
|
| 1980 |
+
"step": 985
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"entropy": 0.8035219550132752,
|
| 1984 |
+
"epoch": 1.554160125588697,
|
| 1985 |
+
"grad_norm": 1.2978500966418507,
|
| 1986 |
+
"learning_rate": 2.8999155926411203e-06,
|
| 1987 |
+
"loss": 0.7669,
|
| 1988 |
+
"mean_token_accuracy": 0.8013963401317596,
|
| 1989 |
+
"num_tokens": 33147410.0,
|
| 1990 |
+
"step": 990
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"entropy": 0.7756295680999756,
|
| 1994 |
+
"epoch": 1.5620094191522762,
|
| 1995 |
+
"grad_norm": 0.9911883814709748,
|
| 1996 |
+
"learning_rate": 2.8040634037370727e-06,
|
| 1997 |
+
"loss": 0.7473,
|
| 1998 |
+
"mean_token_accuracy": 0.807004177570343,
|
| 1999 |
+
"num_tokens": 33319385.0,
|
| 2000 |
+
"step": 995
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"entropy": 0.7991841673851013,
|
| 2004 |
+
"epoch": 1.5698587127158556,
|
| 2005 |
+
"grad_norm": 1.0475227313712396,
|
| 2006 |
+
"learning_rate": 2.7095631350026585e-06,
|
| 2007 |
+
"loss": 0.7809,
|
| 2008 |
+
"mean_token_accuracy": 0.7999978542327881,
|
| 2009 |
+
"num_tokens": 33481523.0,
|
| 2010 |
+
"step": 1000
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"entropy": 0.7900123536586762,
|
| 2014 |
+
"epoch": 1.5777080062794349,
|
| 2015 |
+
"grad_norm": 1.1397172123077388,
|
| 2016 |
+
"learning_rate": 2.616432540460255e-06,
|
| 2017 |
+
"loss": 0.7702,
|
| 2018 |
+
"mean_token_accuracy": 0.8040990054607391,
|
| 2019 |
+
"num_tokens": 33652162.0,
|
| 2020 |
+
"step": 1005
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"entropy": 0.8678532779216767,
|
| 2024 |
+
"epoch": 1.585557299843014,
|
| 2025 |
+
"grad_norm": 1.108645276204879,
|
| 2026 |
+
"learning_rate": 2.524689116807826e-06,
|
| 2027 |
+
"loss": 0.8332,
|
| 2028 |
+
"mean_token_accuracy": 0.7851932346820831,
|
| 2029 |
+
"num_tokens": 33820982.0,
|
| 2030 |
+
"step": 1010
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"entropy": 0.754007887840271,
|
| 2034 |
+
"epoch": 1.5934065934065935,
|
| 2035 |
+
"grad_norm": 0.9962749875685968,
|
| 2036 |
+
"learning_rate": 2.4343501001317604e-06,
|
| 2037 |
+
"loss": 0.7015,
|
| 2038 |
+
"mean_token_accuracy": 0.8182987689971923,
|
| 2039 |
+
"num_tokens": 34001297.0,
|
| 2040 |
+
"step": 1015
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"entropy": 0.7921090185642242,
|
| 2044 |
+
"epoch": 1.6012558869701727,
|
| 2045 |
+
"grad_norm": 1.0446446137216074,
|
| 2046 |
+
"learning_rate": 2.345432462668702e-06,
|
| 2047 |
+
"loss": 0.7485,
|
| 2048 |
+
"mean_token_accuracy": 0.8051556289196015,
|
| 2049 |
+
"num_tokens": 34165590.0,
|
| 2050 |
+
"step": 1020
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"entropy": 0.774686723947525,
|
| 2054 |
+
"epoch": 1.609105180533752,
|
| 2055 |
+
"grad_norm": 1.015493262986232,
|
| 2056 |
+
"learning_rate": 2.257952909616914e-06,
|
| 2057 |
+
"loss": 0.7384,
|
| 2058 |
+
"mean_token_accuracy": 0.8051503002643585,
|
| 2059 |
+
"num_tokens": 34331619.0,
|
| 2060 |
+
"step": 1025
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"entropy": 0.7873682260513306,
|
| 2064 |
+
"epoch": 1.6169544740973314,
|
| 2065 |
+
"grad_norm": 0.9906531738438915,
|
| 2066 |
+
"learning_rate": 2.1719278759978225e-06,
|
| 2067 |
+
"loss": 0.7661,
|
| 2068 |
+
"mean_token_accuracy": 0.804135799407959,
|
| 2069 |
+
"num_tokens": 34496754.0,
|
| 2070 |
+
"step": 1030
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"entropy": 0.773938137292862,
|
| 2074 |
+
"epoch": 1.6248037676609104,
|
| 2075 |
+
"grad_norm": 1.109452175715515,
|
| 2076 |
+
"learning_rate": 2.0873735235683535e-06,
|
| 2077 |
+
"loss": 0.7551,
|
| 2078 |
+
"mean_token_accuracy": 0.8059284925460816,
|
| 2079 |
+
"num_tokens": 34662861.0,
|
| 2080 |
+
"step": 1035
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"entropy": 0.8283044755458832,
|
| 2084 |
+
"epoch": 1.6326530612244898,
|
| 2085 |
+
"grad_norm": 1.0014049035620205,
|
| 2086 |
+
"learning_rate": 2.004305737784541e-06,
|
| 2087 |
+
"loss": 0.8114,
|
| 2088 |
+
"mean_token_accuracy": 0.7896087825298309,
|
| 2089 |
+
"num_tokens": 34821766.0,
|
| 2090 |
+
"step": 1040
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"entropy": 0.8537304162979126,
|
| 2094 |
+
"epoch": 1.640502354788069,
|
| 2095 |
+
"grad_norm": 1.1516235651478994,
|
| 2096 |
+
"learning_rate": 1.922740124817113e-06,
|
| 2097 |
+
"loss": 0.8335,
|
| 2098 |
+
"mean_token_accuracy": 0.7951138377189636,
|
| 2099 |
+
"num_tokens": 34989416.0,
|
| 2100 |
+
"step": 1045
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"entropy": 0.7588850557804108,
|
| 2104 |
+
"epoch": 1.6483516483516483,
|
| 2105 |
+
"grad_norm": 1.093458446274152,
|
| 2106 |
+
"learning_rate": 1.8426920086195065e-06,
|
| 2107 |
+
"loss": 0.7212,
|
| 2108 |
+
"mean_token_accuracy": 0.8150915265083313,
|
| 2109 |
+
"num_tokens": 35166126.0,
|
| 2110 |
+
"step": 1050
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"entropy": 0.8021154761314392,
|
| 2114 |
+
"epoch": 1.6562009419152277,
|
| 2115 |
+
"grad_norm": 0.9205134757749072,
|
| 2116 |
+
"learning_rate": 1.7641764280489081e-06,
|
| 2117 |
+
"loss": 0.7775,
|
| 2118 |
+
"mean_token_accuracy": 0.8028938889503479,
|
| 2119 |
+
"num_tokens": 35335893.0,
|
| 2120 |
+
"step": 1055
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"entropy": 0.8320653736591339,
|
| 2124 |
+
"epoch": 1.664050235478807,
|
| 2125 |
+
"grad_norm": 1.0614802376501893,
|
| 2126 |
+
"learning_rate": 1.6872081340408763e-06,
|
| 2127 |
+
"loss": 0.8022,
|
| 2128 |
+
"mean_token_accuracy": 0.7986601531505585,
|
| 2129 |
+
"num_tokens": 35504349.0,
|
| 2130 |
+
"step": 1060
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"entropy": 0.7887833297252655,
|
| 2134 |
+
"epoch": 1.6718995290423861,
|
| 2135 |
+
"grad_norm": 0.9602064611345482,
|
| 2136 |
+
"learning_rate": 1.6118015868380387e-06,
|
| 2137 |
+
"loss": 0.7592,
|
| 2138 |
+
"mean_token_accuracy": 0.804213547706604,
|
| 2139 |
+
"num_tokens": 35672314.0,
|
| 2140 |
+
"step": 1065
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"entropy": 0.7736551225185394,
|
| 2144 |
+
"epoch": 1.6797488226059656,
|
| 2145 |
+
"grad_norm": 1.0970164822742665,
|
| 2146 |
+
"learning_rate": 1.5379709532733944e-06,
|
| 2147 |
+
"loss": 0.7438,
|
| 2148 |
+
"mean_token_accuracy": 0.8068382501602173,
|
| 2149 |
+
"num_tokens": 35842346.0,
|
| 2150 |
+
"step": 1070
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"entropy": 0.744723516702652,
|
| 2154 |
+
"epoch": 1.6875981161695446,
|
| 2155 |
+
"grad_norm": 1.1340196314083988,
|
| 2156 |
+
"learning_rate": 1.4657301041087812e-06,
|
| 2157 |
+
"loss": 0.7095,
|
| 2158 |
+
"mean_token_accuracy": 0.8160220980644226,
|
| 2159 |
+
"num_tokens": 36011555.0,
|
| 2160 |
+
"step": 1075
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"entropy": 0.7648721992969513,
|
| 2164 |
+
"epoch": 1.695447409733124,
|
| 2165 |
+
"grad_norm": 1.0060542106159092,
|
| 2166 |
+
"learning_rate": 1.395092611428902e-06,
|
| 2167 |
+
"loss": 0.7382,
|
| 2168 |
+
"mean_token_accuracy": 0.8110412538051606,
|
| 2169 |
+
"num_tokens": 36185171.0,
|
| 2170 |
+
"step": 1080
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"entropy": 0.7410736858844758,
|
| 2174 |
+
"epoch": 1.7032967032967035,
|
| 2175 |
+
"grad_norm": 1.0062154758791388,
|
| 2176 |
+
"learning_rate": 1.3260717460915296e-06,
|
| 2177 |
+
"loss": 0.7187,
|
| 2178 |
+
"mean_token_accuracy": 0.818413233757019,
|
| 2179 |
+
"num_tokens": 36358154.0,
|
| 2180 |
+
"step": 1085
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"entropy": 0.8113618612289428,
|
| 2184 |
+
"epoch": 1.7111459968602825,
|
| 2185 |
+
"grad_norm": 1.0517432840388885,
|
| 2186 |
+
"learning_rate": 1.2586804752342596e-06,
|
| 2187 |
+
"loss": 0.7874,
|
| 2188 |
+
"mean_token_accuracy": 0.7986697673797607,
|
| 2189 |
+
"num_tokens": 36521426.0,
|
| 2190 |
+
"step": 1090
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"entropy": 0.7561818957328796,
|
| 2194 |
+
"epoch": 1.718995290423862,
|
| 2195 |
+
"grad_norm": 1.0638794285921913,
|
| 2196 |
+
"learning_rate": 1.1929314598383423e-06,
|
| 2197 |
+
"loss": 0.7301,
|
| 2198 |
+
"mean_token_accuracy": 0.8118774890899658,
|
| 2199 |
+
"num_tokens": 36704082.0,
|
| 2200 |
+
"step": 1095
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"entropy": 0.7728309392929077,
|
| 2204 |
+
"epoch": 1.7268445839874411,
|
| 2205 |
+
"grad_norm": 1.0250744986378886,
|
| 2206 |
+
"learning_rate": 1.1288370523500303e-06,
|
| 2207 |
+
"loss": 0.7269,
|
| 2208 |
+
"mean_token_accuracy": 0.8104123413562775,
|
| 2209 |
+
"num_tokens": 36868416.0,
|
| 2210 |
+
"step": 1100
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"entropy": 0.7941875874996185,
|
| 2214 |
+
"epoch": 1.7346938775510203,
|
| 2215 |
+
"grad_norm": 1.1537121982654628,
|
| 2216 |
+
"learning_rate": 1.0664092943598936e-06,
|
| 2217 |
+
"loss": 0.7753,
|
| 2218 |
+
"mean_token_accuracy": 0.8011082530021667,
|
| 2219 |
+
"num_tokens": 37032626.0,
|
| 2220 |
+
"step": 1105
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"entropy": 0.7873058438301086,
|
| 2224 |
+
"epoch": 1.7425431711145998,
|
| 2225 |
+
"grad_norm": 1.0152506875692804,
|
| 2226 |
+
"learning_rate": 1.0056599143405244e-06,
|
| 2227 |
+
"loss": 0.7437,
|
| 2228 |
+
"mean_token_accuracy": 0.8103858172893524,
|
| 2229 |
+
"num_tokens": 37204340.0,
|
| 2230 |
+
"step": 1110
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"entropy": 0.7289246320724487,
|
| 2234 |
+
"epoch": 1.750392464678179,
|
| 2235 |
+
"grad_norm": 1.0340442042834421,
|
| 2236 |
+
"learning_rate": 9.466003254430933e-07,
|
| 2237 |
+
"loss": 0.7103,
|
| 2238 |
+
"mean_token_accuracy": 0.817768144607544,
|
| 2239 |
+
"num_tokens": 37367113.0,
|
| 2240 |
+
"step": 1115
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"entropy": 0.791058075428009,
|
| 2244 |
+
"epoch": 1.7582417582417582,
|
| 2245 |
+
"grad_norm": 1.146254149472898,
|
| 2246 |
+
"learning_rate": 8.892416233531064e-07,
|
| 2247 |
+
"loss": 0.7552,
|
| 2248 |
+
"mean_token_accuracy": 0.806511789560318,
|
| 2249 |
+
"num_tokens": 37531543.0,
|
| 2250 |
+
"step": 1120
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"entropy": 0.8607315957546234,
|
| 2254 |
+
"epoch": 1.7660910518053377,
|
| 2255 |
+
"grad_norm": 1.012987781402804,
|
| 2256 |
+
"learning_rate": 8.335945842058524e-07,
|
| 2257 |
+
"loss": 0.8219,
|
| 2258 |
+
"mean_token_accuracy": 0.7920519292354584,
|
| 2259 |
+
"num_tokens": 37709584.0,
|
| 2260 |
+
"step": 1125
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"entropy": 0.8378117561340332,
|
| 2264 |
+
"epoch": 1.7739403453689166,
|
| 2265 |
+
"grad_norm": 0.9858004782540654,
|
| 2266 |
+
"learning_rate": 7.79669662561845e-07,
|
| 2267 |
+
"loss": 0.8277,
|
| 2268 |
+
"mean_token_accuracy": 0.7862304747104645,
|
| 2269 |
+
"num_tokens": 37870467.0,
|
| 2270 |
+
"step": 1130
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"entropy": 0.7667446494102478,
|
| 2274 |
+
"epoch": 1.781789638932496,
|
| 2275 |
+
"grad_norm": 0.9733579910483114,
|
| 2276 |
+
"learning_rate": 7.274769894426992e-07,
|
| 2277 |
+
"loss": 0.7315,
|
| 2278 |
+
"mean_token_accuracy": 0.8114334285259247,
|
| 2279 |
+
"num_tokens": 38037411.0,
|
| 2280 |
+
"step": 1135
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"entropy": 0.7864938259124756,
|
| 2284 |
+
"epoch": 1.7896389324960753,
|
| 2285 |
+
"grad_norm": 1.0732339135329727,
|
| 2286 |
+
"learning_rate": 6.770263704277958e-07,
|
| 2287 |
+
"loss": 0.7532,
|
| 2288 |
+
"mean_token_accuracy": 0.807527220249176,
|
| 2289 |
+
"num_tokens": 38213798.0,
|
| 2290 |
+
"step": 1140
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"entropy": 0.815716129541397,
|
| 2294 |
+
"epoch": 1.7974882260596545,
|
| 2295 |
+
"grad_norm": 1.1582020185029405,
|
| 2296 |
+
"learning_rate": 6.283272838120747e-07,
|
| 2297 |
+
"loss": 0.7874,
|
| 2298 |
+
"mean_token_accuracy": 0.7992064177989959,
|
| 2299 |
+
"num_tokens": 38375263.0,
|
| 2300 |
+
"step": 1145
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"entropy": 0.8013858616352081,
|
| 2304 |
+
"epoch": 1.805337519623234,
|
| 2305 |
+
"grad_norm": 0.9712830331555026,
|
| 2306 |
+
"learning_rate": 5.813888788253153e-07,
|
| 2307 |
+
"loss": 0.7648,
|
| 2308 |
+
"mean_token_accuracy": 0.8033481001853943,
|
| 2309 |
+
"num_tokens": 38547445.0,
|
| 2310 |
+
"step": 1150
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"entropy": 0.756001740694046,
|
| 2314 |
+
"epoch": 1.8131868131868132,
|
| 2315 |
+
"grad_norm": 0.9858264012618121,
|
| 2316 |
+
"learning_rate": 5.362199739132656e-07,
|
| 2317 |
+
"loss": 0.7125,
|
| 2318 |
+
"mean_token_accuracy": 0.8135588228702545,
|
| 2319 |
+
"num_tokens": 38722268.0,
|
| 2320 |
+
"step": 1155
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"entropy": 0.871394807100296,
|
| 2324 |
+
"epoch": 1.8210361067503924,
|
| 2325 |
+
"grad_norm": 1.0744537954859468,
|
| 2326 |
+
"learning_rate": 4.928290550808734e-07,
|
| 2327 |
+
"loss": 0.8562,
|
| 2328 |
+
"mean_token_accuracy": 0.7852151751518249,
|
| 2329 |
+
"num_tokens": 38900320.0,
|
| 2330 |
+
"step": 1160
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"entropy": 0.8059657752513886,
|
| 2334 |
+
"epoch": 1.8288854003139718,
|
| 2335 |
+
"grad_norm": 1.0540449615504737,
|
| 2336 |
+
"learning_rate": 4.512242742980155e-07,
|
| 2337 |
+
"loss": 0.7846,
|
| 2338 |
+
"mean_token_accuracy": 0.7966613590717315,
|
| 2339 |
+
"num_tokens": 39062062.0,
|
| 2340 |
+
"step": 1165
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"entropy": 0.8244526743888855,
|
| 2344 |
+
"epoch": 1.836734693877551,
|
| 2345 |
+
"grad_norm": 0.9991737375881368,
|
| 2346 |
+
"learning_rate": 4.114134479679543e-07,
|
| 2347 |
+
"loss": 0.7972,
|
| 2348 |
+
"mean_token_accuracy": 0.7997494339942932,
|
| 2349 |
+
"num_tokens": 39240192.0,
|
| 2350 |
+
"step": 1170
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"entropy": 0.8171651303768158,
|
| 2354 |
+
"epoch": 1.8445839874411303,
|
| 2355 |
+
"grad_norm": 1.0631944976572028,
|
| 2356 |
+
"learning_rate": 3.734040554588514e-07,
|
| 2357 |
+
"loss": 0.7965,
|
| 2358 |
+
"mean_token_accuracy": 0.79867804646492,
|
| 2359 |
+
"num_tokens": 39407568.0,
|
| 2360 |
+
"step": 1175
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"entropy": 0.8014465630054474,
|
| 2364 |
+
"epoch": 1.8524332810047097,
|
| 2365 |
+
"grad_norm": 1.0547329667601317,
|
| 2366 |
+
"learning_rate": 3.372032376986034e-07,
|
| 2367 |
+
"loss": 0.7731,
|
| 2368 |
+
"mean_token_accuracy": 0.8016147315502167,
|
| 2369 |
+
"num_tokens": 39573096.0,
|
| 2370 |
+
"step": 1180
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"entropy": 0.8390758752822876,
|
| 2374 |
+
"epoch": 1.8602825745682887,
|
| 2375 |
+
"grad_norm": 1.1016119489612444,
|
| 2376 |
+
"learning_rate": 3.028177958332512e-07,
|
| 2377 |
+
"loss": 0.7973,
|
| 2378 |
+
"mean_token_accuracy": 0.7978279173374176,
|
| 2379 |
+
"num_tokens": 39752468.0,
|
| 2380 |
+
"step": 1185
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"entropy": 0.7954071164131165,
|
| 2384 |
+
"epoch": 1.8681318681318682,
|
| 2385 |
+
"grad_norm": 1.0681384062877102,
|
| 2386 |
+
"learning_rate": 2.7025418994922835e-07,
|
| 2387 |
+
"loss": 0.7699,
|
| 2388 |
+
"mean_token_accuracy": 0.8014423370361328,
|
| 2389 |
+
"num_tokens": 39925000.0,
|
| 2390 |
+
"step": 1190
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"entropy": 0.7690263390541077,
|
| 2394 |
+
"epoch": 1.8759811616954474,
|
| 2395 |
+
"grad_norm": 1.0897071679243313,
|
| 2396 |
+
"learning_rate": 2.3951853785969535e-07,
|
| 2397 |
+
"loss": 0.7247,
|
| 2398 |
+
"mean_token_accuracy": 0.8128218948841095,
|
| 2399 |
+
"num_tokens": 40101755.0,
|
| 2400 |
+
"step": 1195
|
| 2401 |
+
},
|
| 2402 |
+
{
|
| 2403 |
+
"entropy": 0.7904431164264679,
|
| 2404 |
+
"epoch": 1.8838304552590266,
|
| 2405 |
+
"grad_norm": 1.024203647295983,
|
| 2406 |
+
"learning_rate": 2.106166139551602e-07,
|
| 2407 |
+
"loss": 0.7661,
|
| 2408 |
+
"mean_token_accuracy": 0.8057327687740325,
|
| 2409 |
+
"num_tokens": 40267453.0,
|
| 2410 |
+
"step": 1200
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"entropy": 0.777401077747345,
|
| 2414 |
+
"epoch": 1.891679748822606,
|
| 2415 |
+
"grad_norm": 1.0905304117203938,
|
| 2416 |
+
"learning_rate": 1.8355384811863274e-07,
|
| 2417 |
+
"loss": 0.7457,
|
| 2418 |
+
"mean_token_accuracy": 0.8098066449165344,
|
| 2419 |
+
"num_tokens": 40442554.0,
|
| 2420 |
+
"step": 1205
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"entropy": 0.7975175678730011,
|
| 2424 |
+
"epoch": 1.8995290423861853,
|
| 2425 |
+
"grad_norm": 1.0694029085898318,
|
| 2426 |
+
"learning_rate": 1.5833532470549862e-07,
|
| 2427 |
+
"loss": 0.7671,
|
| 2428 |
+
"mean_token_accuracy": 0.8054138720035553,
|
| 2429 |
+
"num_tokens": 40610140.0,
|
| 2430 |
+
"step": 1210
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"entropy": 0.7842841446399689,
|
| 2434 |
+
"epoch": 1.9073783359497645,
|
| 2435 |
+
"grad_norm": 1.0406692951081211,
|
| 2436 |
+
"learning_rate": 1.349657815883032e-07,
|
| 2437 |
+
"loss": 0.7595,
|
| 2438 |
+
"mean_token_accuracy": 0.8052766740322113,
|
| 2439 |
+
"num_tokens": 40790260.0,
|
| 2440 |
+
"step": 1215
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"entropy": 0.8115567684173584,
|
| 2444 |
+
"epoch": 1.915227629513344,
|
| 2445 |
+
"grad_norm": 1.043901419244555,
|
| 2446 |
+
"learning_rate": 1.134496092666415e-07,
|
| 2447 |
+
"loss": 0.7838,
|
| 2448 |
+
"mean_token_accuracy": 0.7996500730514526,
|
| 2449 |
+
"num_tokens": 40956127.0,
|
| 2450 |
+
"step": 1220
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"entropy": 0.8145834147930145,
|
| 2454 |
+
"epoch": 1.9230769230769231,
|
| 2455 |
+
"grad_norm": 1.1167095476772675,
|
| 2456 |
+
"learning_rate": 9.379085004229571e-08,
|
| 2457 |
+
"loss": 0.7947,
|
| 2458 |
+
"mean_token_accuracy": 0.8002051293849946,
|
| 2459 |
+
"num_tokens": 41129584.0,
|
| 2460 |
+
"step": 1225
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"entropy": 0.7975345253944397,
|
| 2464 |
+
"epoch": 1.9309262166405023,
|
| 2465 |
+
"grad_norm": 1.0430263140771778,
|
| 2466 |
+
"learning_rate": 7.599319725980047e-08,
|
| 2467 |
+
"loss": 0.7683,
|
| 2468 |
+
"mean_token_accuracy": 0.8059882283210754,
|
| 2469 |
+
"num_tokens": 41313026.0,
|
| 2470 |
+
"step": 1230
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"entropy": 0.7508834600448608,
|
| 2474 |
+
"epoch": 1.9387755102040818,
|
| 2475 |
+
"grad_norm": 1.007086000352449,
|
| 2476 |
+
"learning_rate": 6.005999461256684e-08,
|
| 2477 |
+
"loss": 0.7225,
|
| 2478 |
+
"mean_token_accuracy": 0.8129178524017334,
|
| 2479 |
+
"num_tokens": 41474640.0,
|
| 2480 |
+
"step": 1235
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"entropy": 0.7730951249599457,
|
| 2484 |
+
"epoch": 1.9466248037676608,
|
| 2485 |
+
"grad_norm": 0.8290716851619099,
|
| 2486 |
+
"learning_rate": 4.599423551468807e-08,
|
| 2487 |
+
"loss": 0.7361,
|
| 2488 |
+
"mean_token_accuracy": 0.8124108791351319,
|
| 2489 |
+
"num_tokens": 41649075.0,
|
| 2490 |
+
"step": 1240
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"entropy": 0.8062951028347015,
|
| 2494 |
+
"epoch": 1.9544740973312402,
|
| 2495 |
+
"grad_norm": 0.991088255327716,
|
| 2496 |
+
"learning_rate": 3.379856253855951e-08,
|
| 2497 |
+
"loss": 0.7888,
|
| 2498 |
+
"mean_token_accuracy": 0.7999569416046143,
|
| 2499 |
+
"num_tokens": 41826656.0,
|
| 2500 |
+
"step": 1245
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"entropy": 0.8706649959087371,
|
| 2504 |
+
"epoch": 1.9623233908948194,
|
| 2505 |
+
"grad_norm": 1.0157252292092094,
|
| 2506 |
+
"learning_rate": 2.347526691841906e-08,
|
| 2507 |
+
"loss": 0.8462,
|
| 2508 |
+
"mean_token_accuracy": 0.784136426448822,
|
| 2509 |
+
"num_tokens": 41997967.0,
|
| 2510 |
+
"step": 1250
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"entropy": 0.8358751595020294,
|
| 2514 |
+
"epoch": 1.9701726844583987,
|
| 2515 |
+
"grad_norm": 0.9496866968781222,
|
| 2516 |
+
"learning_rate": 1.5026288119874833e-08,
|
| 2517 |
+
"loss": 0.798,
|
| 2518 |
+
"mean_token_accuracy": 0.7961611032485962,
|
| 2519 |
+
"num_tokens": 42162565.0,
|
| 2520 |
+
"step": 1255
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"entropy": 0.7746085405349732,
|
| 2524 |
+
"epoch": 1.978021978021978,
|
| 2525 |
+
"grad_norm": 1.0390321214337483,
|
| 2526 |
+
"learning_rate": 8.453213475543287e-09,
|
| 2527 |
+
"loss": 0.7536,
|
| 2528 |
+
"mean_token_accuracy": 0.8045296788215637,
|
| 2529 |
+
"num_tokens": 42343567.0,
|
| 2530 |
+
"step": 1260
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"entropy": 0.8106873512268067,
|
| 2534 |
+
"epoch": 1.9858712715855573,
|
| 2535 |
+
"grad_norm": 0.8551488607901518,
|
| 2536 |
+
"learning_rate": 3.757277886824451e-09,
|
| 2537 |
+
"loss": 0.7896,
|
| 2538 |
+
"mean_token_accuracy": 0.8025787651538849,
|
| 2539 |
+
"num_tokens": 42498880.0,
|
| 2540 |
+
"step": 1265
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"entropy": 0.7977005839347839,
|
| 2544 |
+
"epoch": 1.9937205651491365,
|
| 2545 |
+
"grad_norm": 1.0550479791927134,
|
| 2546 |
+
"learning_rate": 9.393635919041632e-10,
|
| 2547 |
+
"loss": 0.7593,
|
| 2548 |
+
"mean_token_accuracy": 0.8044815063476562,
|
| 2549 |
+
"num_tokens": 42661916.0,
|
| 2550 |
+
"step": 1270
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"entropy": 0.8685899376869202,
|
| 2554 |
+
"epoch": 2.0,
|
| 2555 |
+
"mean_token_accuracy": 0.7839328944683075,
|
| 2556 |
+
"num_tokens": 42794273.0,
|
| 2557 |
+
"step": 1274,
|
| 2558 |
+
"total_flos": 79316336967680.0,
|
| 2559 |
+
"train_loss": 0.8601785458425525,
|
| 2560 |
+
"train_runtime": 2840.349,
|
| 2561 |
+
"train_samples_per_second": 28.698,
|
| 2562 |
+
"train_steps_per_second": 0.449
|
| 2563 |
+
}
|
| 2564 |
+
],
|
| 2565 |
+
"logging_steps": 5,
|
| 2566 |
+
"max_steps": 1274,
|
| 2567 |
+
"num_input_tokens_seen": 0,
|
| 2568 |
+
"num_train_epochs": 2,
|
| 2569 |
+
"save_steps": 100,
|
| 2570 |
+
"stateful_callbacks": {
|
| 2571 |
+
"TrainerControl": {
|
| 2572 |
+
"args": {
|
| 2573 |
+
"should_epoch_stop": false,
|
| 2574 |
+
"should_evaluate": false,
|
| 2575 |
+
"should_log": false,
|
| 2576 |
+
"should_save": true,
|
| 2577 |
+
"should_training_stop": true
|
| 2578 |
+
},
|
| 2579 |
+
"attributes": {}
|
| 2580 |
+
}
|
| 2581 |
+
},
|
| 2582 |
+
"total_flos": 79316336967680.0,
|
| 2583 |
+
"train_batch_size": 8,
|
| 2584 |
+
"trial_name": null,
|
| 2585 |
+
"trial_params": null
|
| 2586 |
+
}
|