End of training
Browse files- README.md +1 -1
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +9 -0
- trainer_state.json +2504 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -17,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
# train_cb_42_1760623601
|
| 19 |
|
| 20 |
-
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on
|
| 21 |
It achieves the following results on the evaluation set:
|
| 22 |
- Loss: 0.1031
|
| 23 |
- Num Input Tokens Seen: 725992
|
|
|
|
| 17 |
|
| 18 |
# train_cb_42_1760623601
|
| 19 |
|
| 20 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the cb dataset.
|
| 21 |
It achieves the following results on the evaluation set:
|
| 22 |
- Loss: 0.1031
|
| 23 |
- Num Input Tokens Seen: 725992
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_loss": 0.10313867032527924,
|
| 4 |
+
"eval_runtime": 0.6842,
|
| 5 |
+
"eval_samples_per_second": 36.54,
|
| 6 |
+
"eval_steps_per_second": 10.231,
|
| 7 |
+
"num_input_tokens_seen": 725992,
|
| 8 |
+
"total_flos": 3.269194804985856e+16,
|
| 9 |
+
"train_loss": 0.22796724926222833,
|
| 10 |
+
"train_runtime": 288.2339,
|
| 11 |
+
"train_samples_per_second": 15.612,
|
| 12 |
+
"train_steps_per_second": 3.955
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_loss": 0.10313867032527924,
|
| 4 |
+
"eval_runtime": 0.6842,
|
| 5 |
+
"eval_samples_per_second": 36.54,
|
| 6 |
+
"eval_steps_per_second": 10.231,
|
| 7 |
+
"num_input_tokens_seen": 725992
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"num_input_tokens_seen": 725992,
|
| 4 |
+
"total_flos": 3.269194804985856e+16,
|
| 5 |
+
"train_loss": 0.22796724926222833,
|
| 6 |
+
"train_runtime": 288.2339,
|
| 7 |
+
"train_samples_per_second": 15.612,
|
| 8 |
+
"train_steps_per_second": 3.955
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2504 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1140,
|
| 3 |
+
"best_metric": 0.10313867032527924,
|
| 4 |
+
"best_model_checkpoint": "saves_multiple/ia3/llama-3-8b-instruct/train_cb_42_1760623601/checkpoint-1140",
|
| 5 |
+
"epoch": 20.0,
|
| 6 |
+
"eval_steps": 57,
|
| 7 |
+
"global_step": 1140,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08771929824561403,
|
| 14 |
+
"grad_norm": 3.354792833328247,
|
| 15 |
+
"learning_rate": 1.7543859649122807e-06,
|
| 16 |
+
"loss": 1.0538,
|
| 17 |
+
"num_input_tokens_seen": 3232,
|
| 18 |
+
"step": 5,
|
| 19 |
+
"train_runtime": 4.0544,
|
| 20 |
+
"train_tokens_per_second": 797.153
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.17543859649122806,
|
| 24 |
+
"grad_norm": 3.0176336765289307,
|
| 25 |
+
"learning_rate": 3.9473684210526315e-06,
|
| 26 |
+
"loss": 1.0309,
|
| 27 |
+
"num_input_tokens_seen": 6432,
|
| 28 |
+
"step": 10,
|
| 29 |
+
"train_runtime": 5.4768,
|
| 30 |
+
"train_tokens_per_second": 1174.404
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.2631578947368421,
|
| 34 |
+
"grad_norm": 2.661266565322876,
|
| 35 |
+
"learning_rate": 6.140350877192982e-06,
|
| 36 |
+
"loss": 1.0359,
|
| 37 |
+
"num_input_tokens_seen": 9376,
|
| 38 |
+
"step": 15,
|
| 39 |
+
"train_runtime": 6.6968,
|
| 40 |
+
"train_tokens_per_second": 1400.074
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.3508771929824561,
|
| 44 |
+
"grad_norm": 3.1714117527008057,
|
| 45 |
+
"learning_rate": 8.333333333333334e-06,
|
| 46 |
+
"loss": 1.0625,
|
| 47 |
+
"num_input_tokens_seen": 12448,
|
| 48 |
+
"step": 20,
|
| 49 |
+
"train_runtime": 7.7696,
|
| 50 |
+
"train_tokens_per_second": 1602.146
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.43859649122807015,
|
| 54 |
+
"grad_norm": 3.6407687664031982,
|
| 55 |
+
"learning_rate": 1.0526315789473684e-05,
|
| 56 |
+
"loss": 1.1393,
|
| 57 |
+
"num_input_tokens_seen": 15456,
|
| 58 |
+
"step": 25,
|
| 59 |
+
"train_runtime": 8.7928,
|
| 60 |
+
"train_tokens_per_second": 1757.808
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.5263157894736842,
|
| 64 |
+
"grad_norm": 3.01566481590271,
|
| 65 |
+
"learning_rate": 1.2719298245614037e-05,
|
| 66 |
+
"loss": 1.0549,
|
| 67 |
+
"num_input_tokens_seen": 18592,
|
| 68 |
+
"step": 30,
|
| 69 |
+
"train_runtime": 9.8688,
|
| 70 |
+
"train_tokens_per_second": 1883.908
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.6140350877192983,
|
| 74 |
+
"grad_norm": 3.68526291847229,
|
| 75 |
+
"learning_rate": 1.4912280701754386e-05,
|
| 76 |
+
"loss": 1.1235,
|
| 77 |
+
"num_input_tokens_seen": 23136,
|
| 78 |
+
"step": 35,
|
| 79 |
+
"train_runtime": 11.2885,
|
| 80 |
+
"train_tokens_per_second": 2049.515
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7017543859649122,
|
| 84 |
+
"grad_norm": 3.1010959148406982,
|
| 85 |
+
"learning_rate": 1.7105263157894737e-05,
|
| 86 |
+
"loss": 1.1976,
|
| 87 |
+
"num_input_tokens_seen": 26496,
|
| 88 |
+
"step": 40,
|
| 89 |
+
"train_runtime": 12.4133,
|
| 90 |
+
"train_tokens_per_second": 2134.492
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.7894736842105263,
|
| 94 |
+
"grad_norm": 2.408888816833496,
|
| 95 |
+
"learning_rate": 1.929824561403509e-05,
|
| 96 |
+
"loss": 1.1655,
|
| 97 |
+
"num_input_tokens_seen": 29728,
|
| 98 |
+
"step": 45,
|
| 99 |
+
"train_runtime": 13.5113,
|
| 100 |
+
"train_tokens_per_second": 2200.238
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.8771929824561403,
|
| 104 |
+
"grad_norm": 2.720629930496216,
|
| 105 |
+
"learning_rate": 2.149122807017544e-05,
|
| 106 |
+
"loss": 1.184,
|
| 107 |
+
"num_input_tokens_seen": 33024,
|
| 108 |
+
"step": 50,
|
| 109 |
+
"train_runtime": 14.5961,
|
| 110 |
+
"train_tokens_per_second": 2262.517
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.9649122807017544,
|
| 114 |
+
"grad_norm": 3.56002140045166,
|
| 115 |
+
"learning_rate": 2.368421052631579e-05,
|
| 116 |
+
"loss": 1.0884,
|
| 117 |
+
"num_input_tokens_seen": 35904,
|
| 118 |
+
"step": 55,
|
| 119 |
+
"train_runtime": 15.6262,
|
| 120 |
+
"train_tokens_per_second": 2297.68
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 1.0,
|
| 124 |
+
"eval_loss": 1.0677411556243896,
|
| 125 |
+
"eval_runtime": 0.5825,
|
| 126 |
+
"eval_samples_per_second": 42.919,
|
| 127 |
+
"eval_steps_per_second": 12.017,
|
| 128 |
+
"num_input_tokens_seen": 36480,
|
| 129 |
+
"step": 57
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0526315789473684,
|
| 133 |
+
"grad_norm": 2.2148914337158203,
|
| 134 |
+
"learning_rate": 2.5877192982456143e-05,
|
| 135 |
+
"loss": 0.9972,
|
| 136 |
+
"num_input_tokens_seen": 38016,
|
| 137 |
+
"step": 60,
|
| 138 |
+
"train_runtime": 19.5204,
|
| 139 |
+
"train_tokens_per_second": 1947.499
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 1.1403508771929824,
|
| 143 |
+
"grad_norm": 2.6269009113311768,
|
| 144 |
+
"learning_rate": 2.8070175438596492e-05,
|
| 145 |
+
"loss": 1.1312,
|
| 146 |
+
"num_input_tokens_seen": 40800,
|
| 147 |
+
"step": 65,
|
| 148 |
+
"train_runtime": 20.511,
|
| 149 |
+
"train_tokens_per_second": 1989.179
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.2280701754385965,
|
| 153 |
+
"grad_norm": 3.3006234169006348,
|
| 154 |
+
"learning_rate": 3.0263157894736844e-05,
|
| 155 |
+
"loss": 1.1464,
|
| 156 |
+
"num_input_tokens_seen": 43648,
|
| 157 |
+
"step": 70,
|
| 158 |
+
"train_runtime": 21.543,
|
| 159 |
+
"train_tokens_per_second": 2026.084
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.3157894736842106,
|
| 163 |
+
"grad_norm": 3.3827505111694336,
|
| 164 |
+
"learning_rate": 3.24561403508772e-05,
|
| 165 |
+
"loss": 1.1615,
|
| 166 |
+
"num_input_tokens_seen": 46112,
|
| 167 |
+
"step": 75,
|
| 168 |
+
"train_runtime": 22.4546,
|
| 169 |
+
"train_tokens_per_second": 2053.561
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.4035087719298245,
|
| 173 |
+
"grad_norm": 2.930447578430176,
|
| 174 |
+
"learning_rate": 3.4649122807017546e-05,
|
| 175 |
+
"loss": 1.1548,
|
| 176 |
+
"num_input_tokens_seen": 49632,
|
| 177 |
+
"step": 80,
|
| 178 |
+
"train_runtime": 23.6151,
|
| 179 |
+
"train_tokens_per_second": 2101.707
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.4912280701754386,
|
| 183 |
+
"grad_norm": 2.7595036029815674,
|
| 184 |
+
"learning_rate": 3.6842105263157895e-05,
|
| 185 |
+
"loss": 1.0904,
|
| 186 |
+
"num_input_tokens_seen": 53024,
|
| 187 |
+
"step": 85,
|
| 188 |
+
"train_runtime": 24.7506,
|
| 189 |
+
"train_tokens_per_second": 2142.33
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.5789473684210527,
|
| 193 |
+
"grad_norm": 4.10291051864624,
|
| 194 |
+
"learning_rate": 3.9035087719298244e-05,
|
| 195 |
+
"loss": 1.0917,
|
| 196 |
+
"num_input_tokens_seen": 56672,
|
| 197 |
+
"step": 90,
|
| 198 |
+
"train_runtime": 25.9484,
|
| 199 |
+
"train_tokens_per_second": 2184.027
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.6666666666666665,
|
| 203 |
+
"grad_norm": 2.9336965084075928,
|
| 204 |
+
"learning_rate": 4.12280701754386e-05,
|
| 205 |
+
"loss": 1.1794,
|
| 206 |
+
"num_input_tokens_seen": 59296,
|
| 207 |
+
"step": 95,
|
| 208 |
+
"train_runtime": 26.8826,
|
| 209 |
+
"train_tokens_per_second": 2205.741
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.7543859649122808,
|
| 213 |
+
"grad_norm": 3.761521339416504,
|
| 214 |
+
"learning_rate": 4.342105263157895e-05,
|
| 215 |
+
"loss": 1.098,
|
| 216 |
+
"num_input_tokens_seen": 62528,
|
| 217 |
+
"step": 100,
|
| 218 |
+
"train_runtime": 28.0143,
|
| 219 |
+
"train_tokens_per_second": 2232.007
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.8421052631578947,
|
| 223 |
+
"grad_norm": 2.3180630207061768,
|
| 224 |
+
"learning_rate": 4.56140350877193e-05,
|
| 225 |
+
"loss": 0.6296,
|
| 226 |
+
"num_input_tokens_seen": 65856,
|
| 227 |
+
"step": 105,
|
| 228 |
+
"train_runtime": 29.1276,
|
| 229 |
+
"train_tokens_per_second": 2260.95
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 1.9298245614035088,
|
| 233 |
+
"grad_norm": 3.160735607147217,
|
| 234 |
+
"learning_rate": 4.780701754385965e-05,
|
| 235 |
+
"loss": 0.7728,
|
| 236 |
+
"num_input_tokens_seen": 69664,
|
| 237 |
+
"step": 110,
|
| 238 |
+
"train_runtime": 30.3849,
|
| 239 |
+
"train_tokens_per_second": 2292.717
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 2.0,
|
| 243 |
+
"eval_loss": 0.47614240646362305,
|
| 244 |
+
"eval_runtime": 0.5864,
|
| 245 |
+
"eval_samples_per_second": 42.63,
|
| 246 |
+
"eval_steps_per_second": 11.936,
|
| 247 |
+
"num_input_tokens_seen": 72112,
|
| 248 |
+
"step": 114
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.017543859649123,
|
| 252 |
+
"grad_norm": 2.040980339050293,
|
| 253 |
+
"learning_rate": 5e-05,
|
| 254 |
+
"loss": 0.7205,
|
| 255 |
+
"num_input_tokens_seen": 72848,
|
| 256 |
+
"step": 115,
|
| 257 |
+
"train_runtime": 33.2194,
|
| 258 |
+
"train_tokens_per_second": 2192.937
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.1052631578947367,
|
| 262 |
+
"grad_norm": 2.362532615661621,
|
| 263 |
+
"learning_rate": 4.999707014206475e-05,
|
| 264 |
+
"loss": 0.4737,
|
| 265 |
+
"num_input_tokens_seen": 75888,
|
| 266 |
+
"step": 120,
|
| 267 |
+
"train_runtime": 34.3036,
|
| 268 |
+
"train_tokens_per_second": 2212.246
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 2.192982456140351,
|
| 272 |
+
"grad_norm": 0.6496001482009888,
|
| 273 |
+
"learning_rate": 4.9988281254984414e-05,
|
| 274 |
+
"loss": 0.2419,
|
| 275 |
+
"num_input_tokens_seen": 78896,
|
| 276 |
+
"step": 125,
|
| 277 |
+
"train_runtime": 35.3364,
|
| 278 |
+
"train_tokens_per_second": 2232.71
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.280701754385965,
|
| 282 |
+
"grad_norm": 0.8470844626426697,
|
| 283 |
+
"learning_rate": 4.997363539877422e-05,
|
| 284 |
+
"loss": 0.1957,
|
| 285 |
+
"num_input_tokens_seen": 82288,
|
| 286 |
+
"step": 130,
|
| 287 |
+
"train_runtime": 36.469,
|
| 288 |
+
"train_tokens_per_second": 2256.384
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.3684210526315788,
|
| 292 |
+
"grad_norm": 2.7049901485443115,
|
| 293 |
+
"learning_rate": 4.9953136006256415e-05,
|
| 294 |
+
"loss": 0.3056,
|
| 295 |
+
"num_input_tokens_seen": 85936,
|
| 296 |
+
"step": 135,
|
| 297 |
+
"train_runtime": 37.6794,
|
| 298 |
+
"train_tokens_per_second": 2280.717
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.456140350877193,
|
| 302 |
+
"grad_norm": 0.4993942081928253,
|
| 303 |
+
"learning_rate": 4.9926787882255636e-05,
|
| 304 |
+
"loss": 0.46,
|
| 305 |
+
"num_input_tokens_seen": 88784,
|
| 306 |
+
"step": 140,
|
| 307 |
+
"train_runtime": 38.6737,
|
| 308 |
+
"train_tokens_per_second": 2295.723
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 2.543859649122807,
|
| 312 |
+
"grad_norm": 1.035400390625,
|
| 313 |
+
"learning_rate": 4.9894597202472696e-05,
|
| 314 |
+
"loss": 0.2249,
|
| 315 |
+
"num_input_tokens_seen": 91696,
|
| 316 |
+
"step": 145,
|
| 317 |
+
"train_runtime": 39.6938,
|
| 318 |
+
"train_tokens_per_second": 2310.085
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.6315789473684212,
|
| 322 |
+
"grad_norm": 1.3134247064590454,
|
| 323 |
+
"learning_rate": 4.985657151203706e-05,
|
| 324 |
+
"loss": 0.1069,
|
| 325 |
+
"num_input_tokens_seen": 95024,
|
| 326 |
+
"step": 150,
|
| 327 |
+
"train_runtime": 40.8049,
|
| 328 |
+
"train_tokens_per_second": 2328.738
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.719298245614035,
|
| 332 |
+
"grad_norm": 0.3608320653438568,
|
| 333 |
+
"learning_rate": 4.9812719723738435e-05,
|
| 334 |
+
"loss": 0.1625,
|
| 335 |
+
"num_input_tokens_seen": 98512,
|
| 336 |
+
"step": 155,
|
| 337 |
+
"train_runtime": 41.9799,
|
| 338 |
+
"train_tokens_per_second": 2346.65
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.807017543859649,
|
| 342 |
+
"grad_norm": 0.513091504573822,
|
| 343 |
+
"learning_rate": 4.976305211593758e-05,
|
| 344 |
+
"loss": 0.1937,
|
| 345 |
+
"num_input_tokens_seen": 101744,
|
| 346 |
+
"step": 160,
|
| 347 |
+
"train_runtime": 43.0862,
|
| 348 |
+
"train_tokens_per_second": 2361.407
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 2.8947368421052633,
|
| 352 |
+
"grad_norm": 0.9133481979370117,
|
| 353 |
+
"learning_rate": 4.970758033015731e-05,
|
| 354 |
+
"loss": 0.1252,
|
| 355 |
+
"num_input_tokens_seen": 105264,
|
| 356 |
+
"step": 165,
|
| 357 |
+
"train_runtime": 44.2516,
|
| 358 |
+
"train_tokens_per_second": 2378.76
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 2.982456140350877,
|
| 362 |
+
"grad_norm": 0.6826978921890259,
|
| 363 |
+
"learning_rate": 4.9646317368353743e-05,
|
| 364 |
+
"loss": 0.1341,
|
| 365 |
+
"num_input_tokens_seen": 108592,
|
| 366 |
+
"step": 170,
|
| 367 |
+
"train_runtime": 45.3979,
|
| 368 |
+
"train_tokens_per_second": 2392.003
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 3.0,
|
| 372 |
+
"eval_loss": 0.1234603002667427,
|
| 373 |
+
"eval_runtime": 0.5871,
|
| 374 |
+
"eval_samples_per_second": 42.579,
|
| 375 |
+
"eval_steps_per_second": 11.922,
|
| 376 |
+
"num_input_tokens_seen": 108712,
|
| 377 |
+
"step": 171
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 3.0701754385964914,
|
| 381 |
+
"grad_norm": 1.365899920463562,
|
| 382 |
+
"learning_rate": 4.957927758986888e-05,
|
| 383 |
+
"loss": 0.6538,
|
| 384 |
+
"num_input_tokens_seen": 111976,
|
| 385 |
+
"step": 175,
|
| 386 |
+
"train_runtime": 48.1579,
|
| 387 |
+
"train_tokens_per_second": 2325.186
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.1578947368421053,
|
| 391 |
+
"grad_norm": 0.25882503390312195,
|
| 392 |
+
"learning_rate": 4.9506476708064865e-05,
|
| 393 |
+
"loss": 0.1908,
|
| 394 |
+
"num_input_tokens_seen": 114568,
|
| 395 |
+
"step": 180,
|
| 396 |
+
"train_runtime": 49.0895,
|
| 397 |
+
"train_tokens_per_second": 2333.859
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 3.245614035087719,
|
| 401 |
+
"grad_norm": 0.8045045733451843,
|
| 402 |
+
"learning_rate": 4.9427931786641e-05,
|
| 403 |
+
"loss": 0.1842,
|
| 404 |
+
"num_input_tokens_seen": 117384,
|
| 405 |
+
"step": 185,
|
| 406 |
+
"train_runtime": 50.0762,
|
| 407 |
+
"train_tokens_per_second": 2344.106
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.3333333333333335,
|
| 411 |
+
"grad_norm": 1.152528166770935,
|
| 412 |
+
"learning_rate": 4.93436612356342e-05,
|
| 413 |
+
"loss": 0.1629,
|
| 414 |
+
"num_input_tokens_seen": 120968,
|
| 415 |
+
"step": 190,
|
| 416 |
+
"train_runtime": 51.2816,
|
| 417 |
+
"train_tokens_per_second": 2358.898
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 3.4210526315789473,
|
| 421 |
+
"grad_norm": 2.5998055934906006,
|
| 422 |
+
"learning_rate": 4.925368480710385e-05,
|
| 423 |
+
"loss": 0.1379,
|
| 424 |
+
"num_input_tokens_seen": 124616,
|
| 425 |
+
"step": 195,
|
| 426 |
+
"train_runtime": 52.4735,
|
| 427 |
+
"train_tokens_per_second": 2374.837
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 3.5087719298245617,
|
| 431 |
+
"grad_norm": 0.23191165924072266,
|
| 432 |
+
"learning_rate": 4.915802359050222e-05,
|
| 433 |
+
"loss": 0.2213,
|
| 434 |
+
"num_input_tokens_seen": 128104,
|
| 435 |
+
"step": 200,
|
| 436 |
+
"train_runtime": 53.658,
|
| 437 |
+
"train_tokens_per_second": 2387.417
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.5964912280701755,
|
| 441 |
+
"grad_norm": 0.7694133520126343,
|
| 442 |
+
"learning_rate": 4.905670000773126e-05,
|
| 443 |
+
"loss": 0.1228,
|
| 444 |
+
"num_input_tokens_seen": 131528,
|
| 445 |
+
"step": 205,
|
| 446 |
+
"train_runtime": 54.7944,
|
| 447 |
+
"train_tokens_per_second": 2400.393
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 3.6842105263157894,
|
| 451 |
+
"grad_norm": 3.1285037994384766,
|
| 452 |
+
"learning_rate": 4.894973780788722e-05,
|
| 453 |
+
"loss": 0.2468,
|
| 454 |
+
"num_input_tokens_seen": 134600,
|
| 455 |
+
"step": 210,
|
| 456 |
+
"train_runtime": 55.8504,
|
| 457 |
+
"train_tokens_per_second": 2410.011
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 3.7719298245614032,
|
| 461 |
+
"grad_norm": 0.4966896176338196,
|
| 462 |
+
"learning_rate": 4.88371620616941e-05,
|
| 463 |
+
"loss": 0.1399,
|
| 464 |
+
"num_input_tokens_seen": 137896,
|
| 465 |
+
"step": 215,
|
| 466 |
+
"train_runtime": 56.9513,
|
| 467 |
+
"train_tokens_per_second": 2421.296
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 3.8596491228070176,
|
| 471 |
+
"grad_norm": 0.8622232675552368,
|
| 472 |
+
"learning_rate": 4.871899915562736e-05,
|
| 473 |
+
"loss": 0.1856,
|
| 474 |
+
"num_input_tokens_seen": 141064,
|
| 475 |
+
"step": 220,
|
| 476 |
+
"train_runtime": 58.0238,
|
| 477 |
+
"train_tokens_per_second": 2431.141
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.9473684210526314,
|
| 481 |
+
"grad_norm": 0.5208907723426819,
|
| 482 |
+
"learning_rate": 4.8595276785729236e-05,
|
| 483 |
+
"loss": 0.0575,
|
| 484 |
+
"num_input_tokens_seen": 144104,
|
| 485 |
+
"step": 225,
|
| 486 |
+
"train_runtime": 59.0531,
|
| 487 |
+
"train_tokens_per_second": 2440.242
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 4.0,
|
| 491 |
+
"eval_loss": 0.12084977328777313,
|
| 492 |
+
"eval_runtime": 0.588,
|
| 493 |
+
"eval_samples_per_second": 42.516,
|
| 494 |
+
"eval_steps_per_second": 11.905,
|
| 495 |
+
"num_input_tokens_seen": 145296,
|
| 496 |
+
"step": 228
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.035087719298246,
|
| 500 |
+
"grad_norm": 2.5595507621765137,
|
| 501 |
+
"learning_rate": 4.846602395111711e-05,
|
| 502 |
+
"loss": 0.2,
|
| 503 |
+
"num_input_tokens_seen": 146576,
|
| 504 |
+
"step": 230,
|
| 505 |
+
"train_runtime": 61.6276,
|
| 506 |
+
"train_tokens_per_second": 2378.416
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.12280701754386,
|
| 510 |
+
"grad_norm": 2.71787166595459,
|
| 511 |
+
"learning_rate": 4.833127094718643e-05,
|
| 512 |
+
"loss": 0.2242,
|
| 513 |
+
"num_input_tokens_seen": 149456,
|
| 514 |
+
"step": 235,
|
| 515 |
+
"train_runtime": 62.6238,
|
| 516 |
+
"train_tokens_per_second": 2386.567
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.2105263157894735,
|
| 520 |
+
"grad_norm": 0.4318540692329407,
|
| 521 |
+
"learning_rate": 4.819104935850983e-05,
|
| 522 |
+
"loss": 0.1248,
|
| 523 |
+
"num_input_tokens_seen": 152496,
|
| 524 |
+
"step": 240,
|
| 525 |
+
"train_runtime": 63.7523,
|
| 526 |
+
"train_tokens_per_second": 2392.006
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.298245614035087,
|
| 530 |
+
"grad_norm": 0.40779757499694824,
|
| 531 |
+
"learning_rate": 4.804539205143405e-05,
|
| 532 |
+
"loss": 0.1094,
|
| 533 |
+
"num_input_tokens_seen": 155504,
|
| 534 |
+
"step": 245,
|
| 535 |
+
"train_runtime": 64.7945,
|
| 536 |
+
"train_tokens_per_second": 2399.955
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 4.385964912280702,
|
| 540 |
+
"grad_norm": 0.5033511519432068,
|
| 541 |
+
"learning_rate": 4.789433316637644e-05,
|
| 542 |
+
"loss": 0.2048,
|
| 543 |
+
"num_input_tokens_seen": 158640,
|
| 544 |
+
"step": 250,
|
| 545 |
+
"train_runtime": 65.8462,
|
| 546 |
+
"train_tokens_per_second": 2409.25
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 4.473684210526316,
|
| 550 |
+
"grad_norm": 0.175580233335495,
|
| 551 |
+
"learning_rate": 4.7737908109822854e-05,
|
| 552 |
+
"loss": 0.1292,
|
| 553 |
+
"num_input_tokens_seen": 161296,
|
| 554 |
+
"step": 255,
|
| 555 |
+
"train_runtime": 66.7852,
|
| 556 |
+
"train_tokens_per_second": 2415.145
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.56140350877193,
|
| 560 |
+
"grad_norm": 0.5260064601898193,
|
| 561 |
+
"learning_rate": 4.757615354602874e-05,
|
| 562 |
+
"loss": 0.3014,
|
| 563 |
+
"num_input_tokens_seen": 164656,
|
| 564 |
+
"step": 260,
|
| 565 |
+
"train_runtime": 67.92,
|
| 566 |
+
"train_tokens_per_second": 2424.263
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 4.649122807017544,
|
| 570 |
+
"grad_norm": 0.22309179604053497,
|
| 571 |
+
"learning_rate": 4.7409107388425504e-05,
|
| 572 |
+
"loss": 0.1858,
|
| 573 |
+
"num_input_tokens_seen": 168240,
|
| 574 |
+
"step": 265,
|
| 575 |
+
"train_runtime": 69.1058,
|
| 576 |
+
"train_tokens_per_second": 2434.528
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 4.7368421052631575,
|
| 580 |
+
"grad_norm": 1.1554216146469116,
|
| 581 |
+
"learning_rate": 4.723680879073396e-05,
|
| 582 |
+
"loss": 0.0617,
|
| 583 |
+
"num_input_tokens_seen": 171568,
|
| 584 |
+
"step": 270,
|
| 585 |
+
"train_runtime": 70.2239,
|
| 586 |
+
"train_tokens_per_second": 2443.156
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.824561403508772,
|
| 590 |
+
"grad_norm": 0.8314106464385986,
|
| 591 |
+
"learning_rate": 4.70592981377872e-05,
|
| 592 |
+
"loss": 0.0912,
|
| 593 |
+
"num_input_tokens_seen": 174320,
|
| 594 |
+
"step": 275,
|
| 595 |
+
"train_runtime": 71.1856,
|
| 596 |
+
"train_tokens_per_second": 2448.81
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 4.912280701754386,
|
| 600 |
+
"grad_norm": 0.33456218242645264,
|
| 601 |
+
"learning_rate": 4.6876617036064844e-05,
|
| 602 |
+
"loss": 0.0936,
|
| 603 |
+
"num_input_tokens_seen": 178448,
|
| 604 |
+
"step": 280,
|
| 605 |
+
"train_runtime": 72.5198,
|
| 606 |
+
"train_tokens_per_second": 2460.681
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 5.0,
|
| 610 |
+
"grad_norm": 0.03806276246905327,
|
| 611 |
+
"learning_rate": 4.668880830394093e-05,
|
| 612 |
+
"loss": 0.0977,
|
| 613 |
+
"num_input_tokens_seen": 181408,
|
| 614 |
+
"step": 285,
|
| 615 |
+
"train_runtime": 73.6465,
|
| 616 |
+
"train_tokens_per_second": 2463.226
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 5.0,
|
| 620 |
+
"eval_loss": 0.11310870945453644,
|
| 621 |
+
"eval_runtime": 0.5808,
|
| 622 |
+
"eval_samples_per_second": 43.044,
|
| 623 |
+
"eval_steps_per_second": 12.052,
|
| 624 |
+
"num_input_tokens_seen": 181408,
|
| 625 |
+
"step": 285
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.087719298245614,
|
| 629 |
+
"grad_norm": 0.6238030195236206,
|
| 630 |
+
"learning_rate": 4.649591596164778e-05,
|
| 631 |
+
"loss": 0.0658,
|
| 632 |
+
"num_input_tokens_seen": 184352,
|
| 633 |
+
"step": 290,
|
| 634 |
+
"train_runtime": 76.2491,
|
| 635 |
+
"train_tokens_per_second": 2417.761
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 5.175438596491228,
|
| 639 |
+
"grad_norm": 0.5384188890457153,
|
| 640 |
+
"learning_rate": 4.629798522095818e-05,
|
| 641 |
+
"loss": 0.0977,
|
| 642 |
+
"num_input_tokens_seen": 187680,
|
| 643 |
+
"step": 295,
|
| 644 |
+
"train_runtime": 77.4202,
|
| 645 |
+
"train_tokens_per_second": 2424.174
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 5.2631578947368425,
|
| 649 |
+
"grad_norm": 0.6808844804763794,
|
| 650 |
+
"learning_rate": 4.6095062474588225e-05,
|
| 651 |
+
"loss": 0.1278,
|
| 652 |
+
"num_input_tokens_seen": 190880,
|
| 653 |
+
"step": 300,
|
| 654 |
+
"train_runtime": 78.5301,
|
| 655 |
+
"train_tokens_per_second": 2430.66
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 5.350877192982456,
|
| 659 |
+
"grad_norm": 0.35950031876564026,
|
| 660 |
+
"learning_rate": 4.588719528532342e-05,
|
| 661 |
+
"loss": 0.1046,
|
| 662 |
+
"num_input_tokens_seen": 194592,
|
| 663 |
+
"step": 305,
|
| 664 |
+
"train_runtime": 79.752,
|
| 665 |
+
"train_tokens_per_second": 2439.965
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 5.43859649122807,
|
| 669 |
+
"grad_norm": 0.3269582688808441,
|
| 670 |
+
"learning_rate": 4.5674432374870455e-05,
|
| 671 |
+
"loss": 0.075,
|
| 672 |
+
"num_input_tokens_seen": 198720,
|
| 673 |
+
"step": 310,
|
| 674 |
+
"train_runtime": 81.0899,
|
| 675 |
+
"train_tokens_per_second": 2450.614
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 5.526315789473684,
|
| 679 |
+
"grad_norm": 0.5484315752983093,
|
| 680 |
+
"learning_rate": 4.545682361243748e-05,
|
| 681 |
+
"loss": 0.379,
|
| 682 |
+
"num_input_tokens_seen": 201856,
|
| 683 |
+
"step": 315,
|
| 684 |
+
"train_runtime": 82.1892,
|
| 685 |
+
"train_tokens_per_second": 2455.991
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 5.614035087719298,
|
| 689 |
+
"grad_norm": 0.9835972189903259,
|
| 690 |
+
"learning_rate": 4.5234420003045236e-05,
|
| 691 |
+
"loss": 0.1695,
|
| 692 |
+
"num_input_tokens_seen": 204800,
|
| 693 |
+
"step": 320,
|
| 694 |
+
"train_runtime": 83.2065,
|
| 695 |
+
"train_tokens_per_second": 2461.345
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 5.701754385964913,
|
| 699 |
+
"grad_norm": 1.7251195907592773,
|
| 700 |
+
"learning_rate": 4.5007273675572104e-05,
|
| 701 |
+
"loss": 0.2308,
|
| 702 |
+
"num_input_tokens_seen": 208064,
|
| 703 |
+
"step": 325,
|
| 704 |
+
"train_runtime": 84.3235,
|
| 705 |
+
"train_tokens_per_second": 2467.45
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 5.7894736842105265,
|
| 709 |
+
"grad_norm": 1.3619377613067627,
|
| 710 |
+
"learning_rate": 4.4775437870535685e-05,
|
| 711 |
+
"loss": 0.1602,
|
| 712 |
+
"num_input_tokens_seen": 211168,
|
| 713 |
+
"step": 330,
|
| 714 |
+
"train_runtime": 85.3782,
|
| 715 |
+
"train_tokens_per_second": 2473.325
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 5.87719298245614,
|
| 719 |
+
"grad_norm": 0.35113874077796936,
|
| 720 |
+
"learning_rate": 4.4538966927613836e-05,
|
| 721 |
+
"loss": 0.0766,
|
| 722 |
+
"num_input_tokens_seen": 214144,
|
| 723 |
+
"step": 335,
|
| 724 |
+
"train_runtime": 86.4216,
|
| 725 |
+
"train_tokens_per_second": 2477.9
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 5.964912280701754,
|
| 729 |
+
"grad_norm": 0.24905243515968323,
|
| 730 |
+
"learning_rate": 4.4297916272908024e-05,
|
| 731 |
+
"loss": 0.1258,
|
| 732 |
+
"num_input_tokens_seen": 217216,
|
| 733 |
+
"step": 340,
|
| 734 |
+
"train_runtime": 87.4816,
|
| 735 |
+
"train_tokens_per_second": 2482.992
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 6.0,
|
| 739 |
+
"eval_loss": 0.1171339601278305,
|
| 740 |
+
"eval_runtime": 0.5892,
|
| 741 |
+
"eval_samples_per_second": 42.433,
|
| 742 |
+
"eval_steps_per_second": 11.881,
|
| 743 |
+
"num_input_tokens_seen": 217760,
|
| 744 |
+
"step": 342
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 6.052631578947368,
|
| 748 |
+
"grad_norm": 0.6464580297470093,
|
| 749 |
+
"learning_rate": 4.405234240595214e-05,
|
| 750 |
+
"loss": 0.0566,
|
| 751 |
+
"num_input_tokens_seen": 219808,
|
| 752 |
+
"step": 345,
|
| 753 |
+
"train_runtime": 90.143,
|
| 754 |
+
"train_tokens_per_second": 2438.438
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 6.140350877192983,
|
| 758 |
+
"grad_norm": 0.3178783357143402,
|
| 759 |
+
"learning_rate": 4.3802302886469606e-05,
|
| 760 |
+
"loss": 0.2094,
|
| 761 |
+
"num_input_tokens_seen": 223232,
|
| 762 |
+
"step": 350,
|
| 763 |
+
"train_runtime": 91.3086,
|
| 764 |
+
"train_tokens_per_second": 2444.807
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 6.228070175438597,
|
| 768 |
+
"grad_norm": 0.7545506954193115,
|
| 769 |
+
"learning_rate": 4.3547856320882044e-05,
|
| 770 |
+
"loss": 0.206,
|
| 771 |
+
"num_input_tokens_seen": 227008,
|
| 772 |
+
"step": 355,
|
| 773 |
+
"train_runtime": 92.5365,
|
| 774 |
+
"train_tokens_per_second": 2453.172
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 6.315789473684211,
|
| 778 |
+
"grad_norm": 0.19895721971988678,
|
| 779 |
+
"learning_rate": 4.328906234857259e-05,
|
| 780 |
+
"loss": 0.1101,
|
| 781 |
+
"num_input_tokens_seen": 230016,
|
| 782 |
+
"step": 360,
|
| 783 |
+
"train_runtime": 93.5746,
|
| 784 |
+
"train_tokens_per_second": 2458.103
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 6.4035087719298245,
|
| 788 |
+
"grad_norm": 1.6584346294403076,
|
| 789 |
+
"learning_rate": 4.302598162790712e-05,
|
| 790 |
+
"loss": 0.1999,
|
| 791 |
+
"num_input_tokens_seen": 233632,
|
| 792 |
+
"step": 365,
|
| 793 |
+
"train_runtime": 94.7787,
|
| 794 |
+
"train_tokens_per_second": 2465.027
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 6.491228070175438,
|
| 798 |
+
"grad_norm": 1.0977205038070679,
|
| 799 |
+
"learning_rate": 4.27586758220166e-05,
|
| 800 |
+
"loss": 0.1736,
|
| 801 |
+
"num_input_tokens_seen": 236960,
|
| 802 |
+
"step": 370,
|
| 803 |
+
"train_runtime": 95.8823,
|
| 804 |
+
"train_tokens_per_second": 2471.362
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 6.578947368421053,
|
| 808 |
+
"grad_norm": 1.279994249343872,
|
| 809 |
+
"learning_rate": 4.2487207584343955e-05,
|
| 810 |
+
"loss": 0.116,
|
| 811 |
+
"num_input_tokens_seen": 240160,
|
| 812 |
+
"step": 375,
|
| 813 |
+
"train_runtime": 96.9751,
|
| 814 |
+
"train_tokens_per_second": 2476.511
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 6.666666666666667,
|
| 818 |
+
"grad_norm": 1.1907933950424194,
|
| 819 |
+
"learning_rate": 4.2211640543958796e-05,
|
| 820 |
+
"loss": 0.1461,
|
| 821 |
+
"num_input_tokens_seen": 243264,
|
| 822 |
+
"step": 380,
|
| 823 |
+
"train_runtime": 98.0739,
|
| 824 |
+
"train_tokens_per_second": 2480.417
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 6.754385964912281,
|
| 828 |
+
"grad_norm": 0.3227393925189972,
|
| 829 |
+
"learning_rate": 4.193203929064353e-05,
|
| 830 |
+
"loss": 0.1582,
|
| 831 |
+
"num_input_tokens_seen": 246560,
|
| 832 |
+
"step": 385,
|
| 833 |
+
"train_runtime": 99.1776,
|
| 834 |
+
"train_tokens_per_second": 2486.046
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 6.842105263157895,
|
| 838 |
+
"grad_norm": 0.8687011003494263,
|
| 839 |
+
"learning_rate": 4.164846935975421e-05,
|
| 840 |
+
"loss": 0.1395,
|
| 841 |
+
"num_input_tokens_seen": 249408,
|
| 842 |
+
"step": 390,
|
| 843 |
+
"train_runtime": 100.1683,
|
| 844 |
+
"train_tokens_per_second": 2489.89
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 6.9298245614035086,
|
| 848 |
+
"grad_norm": 0.2596314251422882,
|
| 849 |
+
"learning_rate": 4.136099721685983e-05,
|
| 850 |
+
"loss": 0.0496,
|
| 851 |
+
"num_input_tokens_seen": 252192,
|
| 852 |
+
"step": 395,
|
| 853 |
+
"train_runtime": 101.1549,
|
| 854 |
+
"train_tokens_per_second": 2493.126
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 7.0,
|
| 858 |
+
"eval_loss": 0.11278051882982254,
|
| 859 |
+
"eval_runtime": 0.585,
|
| 860 |
+
"eval_samples_per_second": 42.735,
|
| 861 |
+
"eval_steps_per_second": 11.966,
|
| 862 |
+
"num_input_tokens_seen": 254568,
|
| 863 |
+
"step": 399
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 7.017543859649122,
|
| 867 |
+
"grad_norm": 0.154341921210289,
|
| 868 |
+
"learning_rate": 4.1069690242163484e-05,
|
| 869 |
+
"loss": 0.0433,
|
| 870 |
+
"num_input_tokens_seen": 255432,
|
| 871 |
+
"step": 400,
|
| 872 |
+
"train_runtime": 103.8416,
|
| 873 |
+
"train_tokens_per_second": 2459.823
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 7.105263157894737,
|
| 877 |
+
"grad_norm": 0.5618856549263,
|
| 878 |
+
"learning_rate": 4.0774616714709316e-05,
|
| 879 |
+
"loss": 0.1658,
|
| 880 |
+
"num_input_tokens_seen": 258408,
|
| 881 |
+
"step": 405,
|
| 882 |
+
"train_runtime": 104.9452,
|
| 883 |
+
"train_tokens_per_second": 2462.314
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 7.192982456140351,
|
| 887 |
+
"grad_norm": 0.22434987127780914,
|
| 888 |
+
"learning_rate": 4.047584579637857e-05,
|
| 889 |
+
"loss": 0.0465,
|
| 890 |
+
"num_input_tokens_seen": 262024,
|
| 891 |
+
"step": 410,
|
| 892 |
+
"train_runtime": 106.1142,
|
| 893 |
+
"train_tokens_per_second": 2469.264
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 7.280701754385965,
|
| 897 |
+
"grad_norm": 0.10421310365200043,
|
| 898 |
+
"learning_rate": 4.0173447515678916e-05,
|
| 899 |
+
"loss": 0.0636,
|
| 900 |
+
"num_input_tokens_seen": 265032,
|
| 901 |
+
"step": 415,
|
| 902 |
+
"train_runtime": 107.1595,
|
| 903 |
+
"train_tokens_per_second": 2473.248
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 7.368421052631579,
|
| 907 |
+
"grad_norm": 3.0621981620788574,
|
| 908 |
+
"learning_rate": 3.986749275133057e-05,
|
| 909 |
+
"loss": 0.3187,
|
| 910 |
+
"num_input_tokens_seen": 267656,
|
| 911 |
+
"step": 420,
|
| 912 |
+
"train_runtime": 108.1184,
|
| 913 |
+
"train_tokens_per_second": 2475.583
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 7.456140350877193,
|
| 917 |
+
"grad_norm": 0.2600061297416687,
|
| 918 |
+
"learning_rate": 3.955805321565304e-05,
|
| 919 |
+
"loss": 0.0868,
|
| 920 |
+
"num_input_tokens_seen": 271592,
|
| 921 |
+
"step": 425,
|
| 922 |
+
"train_runtime": 109.3564,
|
| 923 |
+
"train_tokens_per_second": 2483.548
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 7.543859649122807,
|
| 927 |
+
"grad_norm": 0.14819173514842987,
|
| 928 |
+
"learning_rate": 3.9245201437756654e-05,
|
| 929 |
+
"loss": 0.0824,
|
| 930 |
+
"num_input_tokens_seen": 275016,
|
| 931 |
+
"step": 430,
|
| 932 |
+
"train_runtime": 110.5238,
|
| 933 |
+
"train_tokens_per_second": 2488.296
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 7.631578947368421,
|
| 937 |
+
"grad_norm": 0.38911575078964233,
|
| 938 |
+
"learning_rate": 3.892901074654255e-05,
|
| 939 |
+
"loss": 0.1303,
|
| 940 |
+
"num_input_tokens_seen": 278248,
|
| 941 |
+
"step": 435,
|
| 942 |
+
"train_runtime": 111.6434,
|
| 943 |
+
"train_tokens_per_second": 2492.293
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 7.719298245614035,
|
| 947 |
+
"grad_norm": 0.18203085660934448,
|
| 948 |
+
"learning_rate": 3.860955525351516e-05,
|
| 949 |
+
"loss": 0.0562,
|
| 950 |
+
"num_input_tokens_seen": 281320,
|
| 951 |
+
"step": 440,
|
| 952 |
+
"train_runtime": 112.6743,
|
| 953 |
+
"train_tokens_per_second": 2496.753
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 7.807017543859649,
|
| 957 |
+
"grad_norm": 0.27270665764808655,
|
| 958 |
+
"learning_rate": 3.82869098354114e-05,
|
| 959 |
+
"loss": 0.0551,
|
| 960 |
+
"num_input_tokens_seen": 284360,
|
| 961 |
+
"step": 445,
|
| 962 |
+
"train_runtime": 113.7167,
|
| 963 |
+
"train_tokens_per_second": 2500.601
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 7.894736842105263,
|
| 967 |
+
"grad_norm": 1.1799060106277466,
|
| 968 |
+
"learning_rate": 3.796115011665034e-05,
|
| 969 |
+
"loss": 0.1744,
|
| 970 |
+
"num_input_tokens_seen": 287944,
|
| 971 |
+
"step": 450,
|
| 972 |
+
"train_runtime": 114.905,
|
| 973 |
+
"train_tokens_per_second": 2505.931
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 7.982456140350877,
|
| 977 |
+
"grad_norm": 1.0825549364089966,
|
| 978 |
+
"learning_rate": 3.763235245160775e-05,
|
| 979 |
+
"loss": 0.3172,
|
| 980 |
+
"num_input_tokens_seen": 290696,
|
| 981 |
+
"step": 455,
|
| 982 |
+
"train_runtime": 115.87,
|
| 983 |
+
"train_tokens_per_second": 2508.811
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 8.0,
|
| 987 |
+
"eval_loss": 0.11445556581020355,
|
| 988 |
+
"eval_runtime": 0.5874,
|
| 989 |
+
"eval_samples_per_second": 42.564,
|
| 990 |
+
"eval_steps_per_second": 11.918,
|
| 991 |
+
"num_input_tokens_seen": 291000,
|
| 992 |
+
"step": 456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 8.070175438596491,
|
| 996 |
+
"grad_norm": 0.09401629120111465,
|
| 997 |
+
"learning_rate": 3.7300593906719464e-05,
|
| 998 |
+
"loss": 0.1706,
|
| 999 |
+
"num_input_tokens_seen": 293432,
|
| 1000 |
+
"step": 460,
|
| 1001 |
+
"train_runtime": 119.3961,
|
| 1002 |
+
"train_tokens_per_second": 2457.635
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 8.157894736842104,
|
| 1006 |
+
"grad_norm": 0.689762532711029,
|
| 1007 |
+
"learning_rate": 3.69659522424179e-05,
|
| 1008 |
+
"loss": 0.0841,
|
| 1009 |
+
"num_input_tokens_seen": 296888,
|
| 1010 |
+
"step": 465,
|
| 1011 |
+
"train_runtime": 120.5629,
|
| 1012 |
+
"train_tokens_per_second": 2462.516
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 8.24561403508772,
|
| 1016 |
+
"grad_norm": 0.9293464422225952,
|
| 1017 |
+
"learning_rate": 3.662850589490592e-05,
|
| 1018 |
+
"loss": 0.1268,
|
| 1019 |
+
"num_input_tokens_seen": 300312,
|
| 1020 |
+
"step": 470,
|
| 1021 |
+
"train_runtime": 121.7003,
|
| 1022 |
+
"train_tokens_per_second": 2467.636
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 8.333333333333334,
|
| 1026 |
+
"grad_norm": 0.3482949435710907,
|
| 1027 |
+
"learning_rate": 3.628833395777224e-05,
|
| 1028 |
+
"loss": 0.1068,
|
| 1029 |
+
"num_input_tokens_seen": 303544,
|
| 1030 |
+
"step": 475,
|
| 1031 |
+
"train_runtime": 122.7941,
|
| 1032 |
+
"train_tokens_per_second": 2471.975
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 8.421052631578947,
|
| 1036 |
+
"grad_norm": 1.2682899236679077,
|
| 1037 |
+
"learning_rate": 3.59455161634528e-05,
|
| 1038 |
+
"loss": 0.1337,
|
| 1039 |
+
"num_input_tokens_seen": 306872,
|
| 1040 |
+
"step": 480,
|
| 1041 |
+
"train_runtime": 123.9208,
|
| 1042 |
+
"train_tokens_per_second": 2476.356
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 8.508771929824562,
|
| 1046 |
+
"grad_norm": 0.7482170462608337,
|
| 1047 |
+
"learning_rate": 3.560013286454242e-05,
|
| 1048 |
+
"loss": 0.1132,
|
| 1049 |
+
"num_input_tokens_seen": 309592,
|
| 1050 |
+
"step": 485,
|
| 1051 |
+
"train_runtime": 124.9127,
|
| 1052 |
+
"train_tokens_per_second": 2478.468
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 8.596491228070175,
|
| 1056 |
+
"grad_norm": 0.5950433015823364,
|
| 1057 |
+
"learning_rate": 3.5252265014961006e-05,
|
| 1058 |
+
"loss": 0.0949,
|
| 1059 |
+
"num_input_tokens_seen": 312728,
|
| 1060 |
+
"step": 490,
|
| 1061 |
+
"train_runtime": 125.9779,
|
| 1062 |
+
"train_tokens_per_second": 2482.403
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 8.68421052631579,
|
| 1066 |
+
"grad_norm": 1.1345596313476562,
|
| 1067 |
+
"learning_rate": 3.490199415097892e-05,
|
| 1068 |
+
"loss": 0.2452,
|
| 1069 |
+
"num_input_tokens_seen": 316088,
|
| 1070 |
+
"step": 495,
|
| 1071 |
+
"train_runtime": 127.1117,
|
| 1072 |
+
"train_tokens_per_second": 2486.694
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 8.771929824561404,
|
| 1076 |
+
"grad_norm": 0.1347583383321762,
|
| 1077 |
+
"learning_rate": 3.45494023721058e-05,
|
| 1078 |
+
"loss": 0.1226,
|
| 1079 |
+
"num_input_tokens_seen": 319352,
|
| 1080 |
+
"step": 500,
|
| 1081 |
+
"train_runtime": 128.2476,
|
| 1082 |
+
"train_tokens_per_second": 2490.121
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 8.859649122807017,
|
| 1086 |
+
"grad_norm": 0.5796114206314087,
|
| 1087 |
+
"learning_rate": 3.4194572321847336e-05,
|
| 1088 |
+
"loss": 0.0672,
|
| 1089 |
+
"num_input_tokens_seen": 323256,
|
| 1090 |
+
"step": 505,
|
| 1091 |
+
"train_runtime": 129.5565,
|
| 1092 |
+
"train_tokens_per_second": 2495.098
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 8.947368421052632,
|
| 1096 |
+
"grad_norm": 0.4608415961265564,
|
| 1097 |
+
"learning_rate": 3.383758716833459e-05,
|
| 1098 |
+
"loss": 0.0538,
|
| 1099 |
+
"num_input_tokens_seen": 326200,
|
| 1100 |
+
"step": 510,
|
| 1101 |
+
"train_runtime": 130.551,
|
| 1102 |
+
"train_tokens_per_second": 2498.641
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 9.0,
|
| 1106 |
+
"eval_loss": 0.11233614385128021,
|
| 1107 |
+
"eval_runtime": 0.5858,
|
| 1108 |
+
"eval_samples_per_second": 42.679,
|
| 1109 |
+
"eval_steps_per_second": 11.95,
|
| 1110 |
+
"num_input_tokens_seen": 327792,
|
| 1111 |
+
"step": 513
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 9.035087719298245,
|
| 1115 |
+
"grad_norm": 1.1988062858581543,
|
| 1116 |
+
"learning_rate": 3.347853058483037e-05,
|
| 1117 |
+
"loss": 0.2043,
|
| 1118 |
+
"num_input_tokens_seen": 329008,
|
| 1119 |
+
"step": 515,
|
| 1120 |
+
"train_runtime": 133.5482,
|
| 1121 |
+
"train_tokens_per_second": 2463.59
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 9.12280701754386,
|
| 1125 |
+
"grad_norm": 0.12025721371173859,
|
| 1126 |
+
"learning_rate": 3.311748673011709e-05,
|
| 1127 |
+
"loss": 0.0332,
|
| 1128 |
+
"num_input_tokens_seen": 332304,
|
| 1129 |
+
"step": 520,
|
| 1130 |
+
"train_runtime": 134.7048,
|
| 1131 |
+
"train_tokens_per_second": 2466.905
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 9.210526315789474,
|
| 1135 |
+
"grad_norm": 1.076543927192688,
|
| 1136 |
+
"learning_rate": 3.275454022877097e-05,
|
| 1137 |
+
"loss": 0.1037,
|
| 1138 |
+
"num_input_tokens_seen": 335792,
|
| 1139 |
+
"step": 525,
|
| 1140 |
+
"train_runtime": 135.8434,
|
| 1141 |
+
"train_tokens_per_second": 2471.906
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 9.298245614035087,
|
| 1145 |
+
"grad_norm": 0.06491763144731522,
|
| 1146 |
+
"learning_rate": 3.238977615132697e-05,
|
| 1147 |
+
"loss": 0.1265,
|
| 1148 |
+
"num_input_tokens_seen": 338672,
|
| 1149 |
+
"step": 530,
|
| 1150 |
+
"train_runtime": 136.8441,
|
| 1151 |
+
"train_tokens_per_second": 2474.875
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 9.385964912280702,
|
| 1155 |
+
"grad_norm": 0.24254095554351807,
|
| 1156 |
+
"learning_rate": 3.202327999433924e-05,
|
| 1157 |
+
"loss": 0.0537,
|
| 1158 |
+
"num_input_tokens_seen": 341968,
|
| 1159 |
+
"step": 535,
|
| 1160 |
+
"train_runtime": 137.946,
|
| 1161 |
+
"train_tokens_per_second": 2478.999
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 9.473684210526315,
|
| 1165 |
+
"grad_norm": 0.9142289161682129,
|
| 1166 |
+
"learning_rate": 3.165513766034167e-05,
|
| 1167 |
+
"loss": 0.18,
|
| 1168 |
+
"num_input_tokens_seen": 345232,
|
| 1169 |
+
"step": 540,
|
| 1170 |
+
"train_runtime": 139.0867,
|
| 1171 |
+
"train_tokens_per_second": 2482.135
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 9.56140350877193,
|
| 1175 |
+
"grad_norm": 0.5542981028556824,
|
| 1176 |
+
"learning_rate": 3.128543543771336e-05,
|
| 1177 |
+
"loss": 0.0903,
|
| 1178 |
+
"num_input_tokens_seen": 348656,
|
| 1179 |
+
"step": 545,
|
| 1180 |
+
"train_runtime": 140.2208,
|
| 1181 |
+
"train_tokens_per_second": 2486.478
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 9.649122807017545,
|
| 1185 |
+
"grad_norm": 0.9289880394935608,
|
| 1186 |
+
"learning_rate": 3.091425998045356e-05,
|
| 1187 |
+
"loss": 0.363,
|
| 1188 |
+
"num_input_tokens_seen": 351248,
|
| 1189 |
+
"step": 550,
|
| 1190 |
+
"train_runtime": 141.1596,
|
| 1191 |
+
"train_tokens_per_second": 2488.304
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 9.736842105263158,
|
| 1195 |
+
"grad_norm": 0.8378196954727173,
|
| 1196 |
+
"learning_rate": 3.0541698287870965e-05,
|
| 1197 |
+
"loss": 0.0762,
|
| 1198 |
+
"num_input_tokens_seen": 355024,
|
| 1199 |
+
"step": 555,
|
| 1200 |
+
"train_runtime": 142.4167,
|
| 1201 |
+
"train_tokens_per_second": 2492.854
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 9.824561403508772,
|
| 1205 |
+
"grad_norm": 0.09416269510984421,
|
| 1206 |
+
"learning_rate": 3.01678376841921e-05,
|
| 1207 |
+
"loss": 0.0364,
|
| 1208 |
+
"num_input_tokens_seen": 357776,
|
| 1209 |
+
"step": 560,
|
| 1210 |
+
"train_runtime": 143.3826,
|
| 1211 |
+
"train_tokens_per_second": 2495.253
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 9.912280701754385,
|
| 1215 |
+
"grad_norm": 0.16007107496261597,
|
| 1216 |
+
"learning_rate": 2.9792765798093465e-05,
|
| 1217 |
+
"loss": 0.1419,
|
| 1218 |
+
"num_input_tokens_seen": 360784,
|
| 1219 |
+
"step": 565,
|
| 1220 |
+
"train_runtime": 144.4207,
|
| 1221 |
+
"train_tokens_per_second": 2498.145
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 10.0,
|
| 1225 |
+
"grad_norm": 1.0040969848632812,
|
| 1226 |
+
"learning_rate": 2.94165705421624e-05,
|
| 1227 |
+
"loss": 0.1347,
|
| 1228 |
+
"num_input_tokens_seen": 363864,
|
| 1229 |
+
"step": 570,
|
| 1230 |
+
"train_runtime": 145.5797,
|
| 1231 |
+
"train_tokens_per_second": 2499.414
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 10.0,
|
| 1235 |
+
"eval_loss": 0.10720973461866379,
|
| 1236 |
+
"eval_runtime": 0.5841,
|
| 1237 |
+
"eval_samples_per_second": 42.802,
|
| 1238 |
+
"eval_steps_per_second": 11.984,
|
| 1239 |
+
"num_input_tokens_seen": 363864,
|
| 1240 |
+
"step": 570
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 10.087719298245615,
|
| 1244 |
+
"grad_norm": 0.06115713343024254,
|
| 1245 |
+
"learning_rate": 2.9039340092291373e-05,
|
| 1246 |
+
"loss": 0.2521,
|
| 1247 |
+
"num_input_tokens_seen": 368120,
|
| 1248 |
+
"step": 575,
|
| 1249 |
+
"train_runtime": 148.456,
|
| 1250 |
+
"train_tokens_per_second": 2479.657
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 10.175438596491228,
|
| 1254 |
+
"grad_norm": 0.9393717050552368,
|
| 1255 |
+
"learning_rate": 2.8661162867010543e-05,
|
| 1256 |
+
"loss": 0.1127,
|
| 1257 |
+
"num_input_tokens_seen": 371416,
|
| 1258 |
+
"step": 580,
|
| 1259 |
+
"train_runtime": 149.5706,
|
| 1260 |
+
"train_tokens_per_second": 2483.216
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 10.263157894736842,
|
| 1264 |
+
"grad_norm": 0.436775803565979,
|
| 1265 |
+
"learning_rate": 2.8282127506763456e-05,
|
| 1266 |
+
"loss": 0.0666,
|
| 1267 |
+
"num_input_tokens_seen": 374168,
|
| 1268 |
+
"step": 585,
|
| 1269 |
+
"train_runtime": 150.5321,
|
| 1270 |
+
"train_tokens_per_second": 2485.637
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 10.350877192982455,
|
| 1274 |
+
"grad_norm": 0.4993436634540558,
|
| 1275 |
+
"learning_rate": 2.7902322853130757e-05,
|
| 1276 |
+
"loss": 0.0503,
|
| 1277 |
+
"num_input_tokens_seen": 377496,
|
| 1278 |
+
"step": 590,
|
| 1279 |
+
"train_runtime": 151.6608,
|
| 1280 |
+
"train_tokens_per_second": 2489.08
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 10.43859649122807,
|
| 1284 |
+
"grad_norm": 2.0963611602783203,
|
| 1285 |
+
"learning_rate": 2.752183792800671e-05,
|
| 1286 |
+
"loss": 0.256,
|
| 1287 |
+
"num_input_tokens_seen": 381592,
|
| 1288 |
+
"step": 595,
|
| 1289 |
+
"train_runtime": 152.9709,
|
| 1290 |
+
"train_tokens_per_second": 2494.539
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 10.526315789473685,
|
| 1294 |
+
"grad_norm": 0.2084749937057495,
|
| 1295 |
+
"learning_rate": 2.7140761912733474e-05,
|
| 1296 |
+
"loss": 0.0612,
|
| 1297 |
+
"num_input_tokens_seen": 384760,
|
| 1298 |
+
"step": 600,
|
| 1299 |
+
"train_runtime": 154.0824,
|
| 1300 |
+
"train_tokens_per_second": 2497.105
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 10.614035087719298,
|
| 1304 |
+
"grad_norm": 0.20247477293014526,
|
| 1305 |
+
"learning_rate": 2.6759184127198046e-05,
|
| 1306 |
+
"loss": 0.1003,
|
| 1307 |
+
"num_input_tokens_seen": 387512,
|
| 1308 |
+
"step": 605,
|
| 1309 |
+
"train_runtime": 155.0647,
|
| 1310 |
+
"train_tokens_per_second": 2499.035
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 10.701754385964913,
|
| 1314 |
+
"grad_norm": 1.0483530759811401,
|
| 1315 |
+
"learning_rate": 2.6377194008896637e-05,
|
| 1316 |
+
"loss": 0.082,
|
| 1317 |
+
"num_input_tokens_seen": 390680,
|
| 1318 |
+
"step": 610,
|
| 1319 |
+
"train_runtime": 156.1362,
|
| 1320 |
+
"train_tokens_per_second": 2502.175
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 10.789473684210526,
|
| 1324 |
+
"grad_norm": 0.9714593291282654,
|
| 1325 |
+
"learning_rate": 2.5994881091971605e-05,
|
| 1326 |
+
"loss": 0.1216,
|
| 1327 |
+
"num_input_tokens_seen": 393944,
|
| 1328 |
+
"step": 615,
|
| 1329 |
+
"train_runtime": 157.2556,
|
| 1330 |
+
"train_tokens_per_second": 2505.119
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 10.87719298245614,
|
| 1334 |
+
"grad_norm": 1.19658362865448,
|
| 1335 |
+
"learning_rate": 2.5612334986225623e-05,
|
| 1336 |
+
"loss": 0.12,
|
| 1337 |
+
"num_input_tokens_seen": 396632,
|
| 1338 |
+
"step": 620,
|
| 1339 |
+
"train_runtime": 158.2119,
|
| 1340 |
+
"train_tokens_per_second": 2506.967
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 10.964912280701755,
|
| 1344 |
+
"grad_norm": 0.9653410911560059,
|
| 1345 |
+
"learning_rate": 2.5229645356118163e-05,
|
| 1346 |
+
"loss": 0.1298,
|
| 1347 |
+
"num_input_tokens_seen": 399288,
|
| 1348 |
+
"step": 625,
|
| 1349 |
+
"train_runtime": 159.1947,
|
| 1350 |
+
"train_tokens_per_second": 2508.173
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 11.0,
|
| 1354 |
+
"eval_loss": 0.10691339522600174,
|
| 1355 |
+
"eval_runtime": 0.587,
|
| 1356 |
+
"eval_samples_per_second": 42.586,
|
| 1357 |
+
"eval_steps_per_second": 11.924,
|
| 1358 |
+
"num_input_tokens_seen": 400104,
|
| 1359 |
+
"step": 627
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 11.052631578947368,
|
| 1363 |
+
"grad_norm": 0.33754128217697144,
|
| 1364 |
+
"learning_rate": 2.4846901899749185e-05,
|
| 1365 |
+
"loss": 0.0611,
|
| 1366 |
+
"num_input_tokens_seen": 401928,
|
| 1367 |
+
"step": 630,
|
| 1368 |
+
"train_runtime": 162.1813,
|
| 1369 |
+
"train_tokens_per_second": 2478.263
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 11.140350877192983,
|
| 1373 |
+
"grad_norm": 0.5505395531654358,
|
| 1374 |
+
"learning_rate": 2.4464194327834926e-05,
|
| 1375 |
+
"loss": 0.1293,
|
| 1376 |
+
"num_input_tokens_seen": 404360,
|
| 1377 |
+
"step": 635,
|
| 1378 |
+
"train_runtime": 163.1461,
|
| 1379 |
+
"train_tokens_per_second": 2478.515
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 11.228070175438596,
|
| 1383 |
+
"grad_norm": 0.8922757506370544,
|
| 1384 |
+
"learning_rate": 2.4081612342680694e-05,
|
| 1385 |
+
"loss": 0.0894,
|
| 1386 |
+
"num_input_tokens_seen": 407496,
|
| 1387 |
+
"step": 640,
|
| 1388 |
+
"train_runtime": 164.2456,
|
| 1389 |
+
"train_tokens_per_second": 2481.016
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 11.31578947368421,
|
| 1393 |
+
"grad_norm": 0.22152556478977203,
|
| 1394 |
+
"learning_rate": 2.369924561715569e-05,
|
| 1395 |
+
"loss": 0.0581,
|
| 1396 |
+
"num_input_tokens_seen": 410760,
|
| 1397 |
+
"step": 645,
|
| 1398 |
+
"train_runtime": 165.3215,
|
| 1399 |
+
"train_tokens_per_second": 2484.613
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 11.403508771929825,
|
| 1403 |
+
"grad_norm": 1.2716141939163208,
|
| 1404 |
+
"learning_rate": 2.3317183773674718e-05,
|
| 1405 |
+
"loss": 0.2198,
|
| 1406 |
+
"num_input_tokens_seen": 413832,
|
| 1407 |
+
"step": 650,
|
| 1408 |
+
"train_runtime": 166.3533,
|
| 1409 |
+
"train_tokens_per_second": 2487.67
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 11.491228070175438,
|
| 1413 |
+
"grad_norm": 0.23000480234622955,
|
| 1414 |
+
"learning_rate": 2.2935516363191693e-05,
|
| 1415 |
+
"loss": 0.1275,
|
| 1416 |
+
"num_input_tokens_seen": 417320,
|
| 1417 |
+
"step": 655,
|
| 1418 |
+
"train_runtime": 167.5146,
|
| 1419 |
+
"train_tokens_per_second": 2491.246
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 11.578947368421053,
|
| 1423 |
+
"grad_norm": 0.5997104644775391,
|
| 1424 |
+
"learning_rate": 2.2554332844209904e-05,
|
| 1425 |
+
"loss": 0.0887,
|
| 1426 |
+
"num_input_tokens_seen": 420232,
|
| 1427 |
+
"step": 660,
|
| 1428 |
+
"train_runtime": 168.5486,
|
| 1429 |
+
"train_tokens_per_second": 2493.239
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 11.666666666666666,
|
| 1433 |
+
"grad_norm": 0.38558098673820496,
|
| 1434 |
+
"learning_rate": 2.2173722561813987e-05,
|
| 1435 |
+
"loss": 0.0379,
|
| 1436 |
+
"num_input_tokens_seen": 423528,
|
| 1437 |
+
"step": 665,
|
| 1438 |
+
"train_runtime": 169.6363,
|
| 1439 |
+
"train_tokens_per_second": 2496.683
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 11.75438596491228,
|
| 1443 |
+
"grad_norm": 0.6237869262695312,
|
| 1444 |
+
"learning_rate": 2.179377472672842e-05,
|
| 1445 |
+
"loss": 0.1143,
|
| 1446 |
+
"num_input_tokens_seen": 426920,
|
| 1447 |
+
"step": 670,
|
| 1448 |
+
"train_runtime": 170.7838,
|
| 1449 |
+
"train_tokens_per_second": 2499.768
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 11.842105263157894,
|
| 1453 |
+
"grad_norm": 0.6095879673957825,
|
| 1454 |
+
"learning_rate": 2.1414578394407597e-05,
|
| 1455 |
+
"loss": 0.0651,
|
| 1456 |
+
"num_input_tokens_seen": 431016,
|
| 1457 |
+
"step": 675,
|
| 1458 |
+
"train_runtime": 172.1147,
|
| 1459 |
+
"train_tokens_per_second": 2504.236
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 11.929824561403509,
|
| 1463 |
+
"grad_norm": 2.0376551151275635,
|
| 1464 |
+
"learning_rate": 2.1036222444162147e-05,
|
| 1465 |
+
"loss": 0.2924,
|
| 1466 |
+
"num_input_tokens_seen": 434600,
|
| 1467 |
+
"step": 680,
|
| 1468 |
+
"train_runtime": 173.3197,
|
| 1469 |
+
"train_tokens_per_second": 2507.506
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 12.0,
|
| 1473 |
+
"eval_loss": 0.10777657479047775,
|
| 1474 |
+
"eval_runtime": 0.5862,
|
| 1475 |
+
"eval_samples_per_second": 42.649,
|
| 1476 |
+
"eval_steps_per_second": 11.942,
|
| 1477 |
+
"num_input_tokens_seen": 436440,
|
| 1478 |
+
"step": 684
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 12.017543859649123,
|
| 1482 |
+
"grad_norm": 0.6392739415168762,
|
| 1483 |
+
"learning_rate": 2.0658795558326743e-05,
|
| 1484 |
+
"loss": 0.0623,
|
| 1485 |
+
"num_input_tokens_seen": 437016,
|
| 1486 |
+
"step": 685,
|
| 1487 |
+
"train_runtime": 175.7733,
|
| 1488 |
+
"train_tokens_per_second": 2486.249
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 12.105263157894736,
|
| 1492 |
+
"grad_norm": 0.3666384518146515,
|
| 1493 |
+
"learning_rate": 2.0282386201473894e-05,
|
| 1494 |
+
"loss": 0.041,
|
| 1495 |
+
"num_input_tokens_seen": 440280,
|
| 1496 |
+
"step": 690,
|
| 1497 |
+
"train_runtime": 176.8869,
|
| 1498 |
+
"train_tokens_per_second": 2489.048
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 12.192982456140351,
|
| 1502 |
+
"grad_norm": 0.843946635723114,
|
| 1503 |
+
"learning_rate": 1.99070825996789e-05,
|
| 1504 |
+
"loss": 0.0815,
|
| 1505 |
+
"num_input_tokens_seen": 443800,
|
| 1506 |
+
"step": 695,
|
| 1507 |
+
"train_runtime": 178.0941,
|
| 1508 |
+
"train_tokens_per_second": 2491.941
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 12.280701754385966,
|
| 1512 |
+
"grad_norm": 0.1081356629729271,
|
| 1513 |
+
"learning_rate": 1.9532972719840607e-05,
|
| 1514 |
+
"loss": 0.0367,
|
| 1515 |
+
"num_input_tokens_seen": 446168,
|
| 1516 |
+
"step": 700,
|
| 1517 |
+
"train_runtime": 178.9971,
|
| 1518 |
+
"train_tokens_per_second": 2492.599
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 12.368421052631579,
|
| 1522 |
+
"grad_norm": 1.2140501737594604,
|
| 1523 |
+
"learning_rate": 1.9160144249063035e-05,
|
| 1524 |
+
"loss": 0.4138,
|
| 1525 |
+
"num_input_tokens_seen": 449560,
|
| 1526 |
+
"step": 705,
|
| 1527 |
+
"train_runtime": 180.1421,
|
| 1528 |
+
"train_tokens_per_second": 2495.586
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 12.456140350877194,
|
| 1532 |
+
"grad_norm": 0.10039330273866653,
|
| 1533 |
+
"learning_rate": 1.8788684574102467e-05,
|
| 1534 |
+
"loss": 0.0396,
|
| 1535 |
+
"num_input_tokens_seen": 452824,
|
| 1536 |
+
"step": 710,
|
| 1537 |
+
"train_runtime": 181.246,
|
| 1538 |
+
"train_tokens_per_second": 2498.394
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 12.543859649122806,
|
| 1542 |
+
"grad_norm": 1.3357945680618286,
|
| 1543 |
+
"learning_rate": 1.8418680760885027e-05,
|
| 1544 |
+
"loss": 0.166,
|
| 1545 |
+
"num_input_tokens_seen": 456120,
|
| 1546 |
+
"step": 715,
|
| 1547 |
+
"train_runtime": 182.3751,
|
| 1548 |
+
"train_tokens_per_second": 2500.999
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 12.631578947368421,
|
| 1552 |
+
"grad_norm": 0.7184349894523621,
|
| 1553 |
+
"learning_rate": 1.805021953409934e-05,
|
| 1554 |
+
"loss": 0.1237,
|
| 1555 |
+
"num_input_tokens_seen": 459736,
|
| 1556 |
+
"step": 720,
|
| 1557 |
+
"train_runtime": 183.5708,
|
| 1558 |
+
"train_tokens_per_second": 2504.407
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 12.719298245614034,
|
| 1562 |
+
"grad_norm": 0.8796572089195251,
|
| 1563 |
+
"learning_rate": 1.7683387256869353e-05,
|
| 1564 |
+
"loss": 0.1099,
|
| 1565 |
+
"num_input_tokens_seen": 462584,
|
| 1566 |
+
"step": 725,
|
| 1567 |
+
"train_runtime": 184.56,
|
| 1568 |
+
"train_tokens_per_second": 2506.415
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 12.807017543859649,
|
| 1572 |
+
"grad_norm": 0.0641786977648735,
|
| 1573 |
+
"learning_rate": 1.7318269910511736e-05,
|
| 1574 |
+
"loss": 0.0561,
|
| 1575 |
+
"num_input_tokens_seen": 465688,
|
| 1576 |
+
"step": 730,
|
| 1577 |
+
"train_runtime": 185.6328,
|
| 1578 |
+
"train_tokens_per_second": 2508.652
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 12.894736842105264,
|
| 1582 |
+
"grad_norm": 0.7770268321037292,
|
| 1583 |
+
"learning_rate": 1.6954953074382863e-05,
|
| 1584 |
+
"loss": 0.0666,
|
| 1585 |
+
"num_input_tokens_seen": 468856,
|
| 1586 |
+
"step": 735,
|
| 1587 |
+
"train_runtime": 186.7038,
|
| 1588 |
+
"train_tokens_per_second": 2511.229
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 12.982456140350877,
|
| 1592 |
+
"grad_norm": 0.7374283075332642,
|
| 1593 |
+
"learning_rate": 1.659352190581993e-05,
|
| 1594 |
+
"loss": 0.1389,
|
| 1595 |
+
"num_input_tokens_seen": 471768,
|
| 1596 |
+
"step": 740,
|
| 1597 |
+
"train_runtime": 187.7037,
|
| 1598 |
+
"train_tokens_per_second": 2513.365
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 13.0,
|
| 1602 |
+
"eval_loss": 0.10621943324804306,
|
| 1603 |
+
"eval_runtime": 0.5863,
|
| 1604 |
+
"eval_samples_per_second": 42.641,
|
| 1605 |
+
"eval_steps_per_second": 11.94,
|
| 1606 |
+
"num_input_tokens_seen": 471944,
|
| 1607 |
+
"step": 741
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 13.070175438596491,
|
| 1611 |
+
"grad_norm": 0.05659092217683792,
|
| 1612 |
+
"learning_rate": 1.6234061120181142e-05,
|
| 1613 |
+
"loss": 0.2374,
|
| 1614 |
+
"num_input_tokens_seen": 474344,
|
| 1615 |
+
"step": 745,
|
| 1616 |
+
"train_runtime": 190.9684,
|
| 1617 |
+
"train_tokens_per_second": 2483.888
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 13.157894736842104,
|
| 1621 |
+
"grad_norm": 1.235370397567749,
|
| 1622 |
+
"learning_rate": 1.5876654970989308e-05,
|
| 1623 |
+
"loss": 0.1211,
|
| 1624 |
+
"num_input_tokens_seen": 478120,
|
| 1625 |
+
"step": 750,
|
| 1626 |
+
"train_runtime": 192.2311,
|
| 1627 |
+
"train_tokens_per_second": 2487.215
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 13.24561403508772,
|
| 1631 |
+
"grad_norm": 0.5860034227371216,
|
| 1632 |
+
"learning_rate": 1.552138723018382e-05,
|
| 1633 |
+
"loss": 0.1409,
|
| 1634 |
+
"num_input_tokens_seen": 480840,
|
| 1635 |
+
"step": 755,
|
| 1636 |
+
"train_runtime": 193.1775,
|
| 1637 |
+
"train_tokens_per_second": 2489.109
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 13.333333333333334,
|
| 1641 |
+
"grad_norm": 0.6669791340827942,
|
| 1642 |
+
"learning_rate": 1.5168341168485423e-05,
|
| 1643 |
+
"loss": 0.1713,
|
| 1644 |
+
"num_input_tokens_seen": 484552,
|
| 1645 |
+
"step": 760,
|
| 1646 |
+
"train_runtime": 194.4098,
|
| 1647 |
+
"train_tokens_per_second": 2492.425
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 13.421052631578947,
|
| 1651 |
+
"grad_norm": 0.3427886366844177,
|
| 1652 |
+
"learning_rate": 1.4817599535878565e-05,
|
| 1653 |
+
"loss": 0.1061,
|
| 1654 |
+
"num_input_tokens_seen": 487304,
|
| 1655 |
+
"step": 765,
|
| 1656 |
+
"train_runtime": 195.3955,
|
| 1657 |
+
"train_tokens_per_second": 2493.936
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 13.508771929824562,
|
| 1661 |
+
"grad_norm": 0.4120057225227356,
|
| 1662 |
+
"learning_rate": 1.4469244542215682e-05,
|
| 1663 |
+
"loss": 0.0367,
|
| 1664 |
+
"num_input_tokens_seen": 490536,
|
| 1665 |
+
"step": 770,
|
| 1666 |
+
"train_runtime": 196.4949,
|
| 1667 |
+
"train_tokens_per_second": 2496.431
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 13.596491228070175,
|
| 1671 |
+
"grad_norm": 0.5584914684295654,
|
| 1672 |
+
"learning_rate": 1.4123357837948175e-05,
|
| 1673 |
+
"loss": 0.0849,
|
| 1674 |
+
"num_input_tokens_seen": 493128,
|
| 1675 |
+
"step": 775,
|
| 1676 |
+
"train_runtime": 197.4396,
|
| 1677 |
+
"train_tokens_per_second": 2497.614
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 13.68421052631579,
|
| 1681 |
+
"grad_norm": 0.942362368106842,
|
| 1682 |
+
"learning_rate": 1.3780020494988446e-05,
|
| 1683 |
+
"loss": 0.2508,
|
| 1684 |
+
"num_input_tokens_seen": 496136,
|
| 1685 |
+
"step": 780,
|
| 1686 |
+
"train_runtime": 198.4693,
|
| 1687 |
+
"train_tokens_per_second": 2499.813
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 13.771929824561404,
|
| 1691 |
+
"grad_norm": 0.29723063111305237,
|
| 1692 |
+
"learning_rate": 1.3439312987707615e-05,
|
| 1693 |
+
"loss": 0.0526,
|
| 1694 |
+
"num_input_tokens_seen": 499432,
|
| 1695 |
+
"step": 785,
|
| 1696 |
+
"train_runtime": 199.6161,
|
| 1697 |
+
"train_tokens_per_second": 2501.962
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 13.859649122807017,
|
| 1701 |
+
"grad_norm": 1.1332062482833862,
|
| 1702 |
+
"learning_rate": 1.3101315174073162e-05,
|
| 1703 |
+
"loss": 0.1816,
|
| 1704 |
+
"num_input_tokens_seen": 502984,
|
| 1705 |
+
"step": 790,
|
| 1706 |
+
"train_runtime": 200.7801,
|
| 1707 |
+
"train_tokens_per_second": 2505.149
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 13.947368421052632,
|
| 1711 |
+
"grad_norm": 0.8299708962440491,
|
| 1712 |
+
"learning_rate": 1.2766106276931223e-05,
|
| 1713 |
+
"loss": 0.0521,
|
| 1714 |
+
"num_input_tokens_seen": 506376,
|
| 1715 |
+
"step": 795,
|
| 1716 |
+
"train_runtime": 201.9139,
|
| 1717 |
+
"train_tokens_per_second": 2507.881
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 14.0,
|
| 1721 |
+
"eval_loss": 0.10609572380781174,
|
| 1722 |
+
"eval_runtime": 0.5874,
|
| 1723 |
+
"eval_samples_per_second": 42.562,
|
| 1724 |
+
"eval_steps_per_second": 11.917,
|
| 1725 |
+
"num_input_tokens_seen": 508424,
|
| 1726 |
+
"step": 798
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 14.035087719298245,
|
| 1730 |
+
"grad_norm": 1.2610480785369873,
|
| 1731 |
+
"learning_rate": 1.243376486543755e-05,
|
| 1732 |
+
"loss": 0.1066,
|
| 1733 |
+
"num_input_tokens_seen": 509672,
|
| 1734 |
+
"step": 800,
|
| 1735 |
+
"train_runtime": 204.608,
|
| 1736 |
+
"train_tokens_per_second": 2490.968
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 14.12280701754386,
|
| 1740 |
+
"grad_norm": 0.10829579085111618,
|
| 1741 |
+
"learning_rate": 1.2104368836641908e-05,
|
| 1742 |
+
"loss": 0.1141,
|
| 1743 |
+
"num_input_tokens_seen": 513320,
|
| 1744 |
+
"step": 805,
|
| 1745 |
+
"train_runtime": 205.8216,
|
| 1746 |
+
"train_tokens_per_second": 2494.004
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 14.210526315789474,
|
| 1750 |
+
"grad_norm": 0.283528596162796,
|
| 1751 |
+
"learning_rate": 1.1777995397229771e-05,
|
| 1752 |
+
"loss": 0.2027,
|
| 1753 |
+
"num_input_tokens_seen": 516296,
|
| 1754 |
+
"step": 810,
|
| 1755 |
+
"train_runtime": 206.8214,
|
| 1756 |
+
"train_tokens_per_second": 2496.337
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 14.298245614035087,
|
| 1760 |
+
"grad_norm": 1.0561656951904297,
|
| 1761 |
+
"learning_rate": 1.1454721045426073e-05,
|
| 1762 |
+
"loss": 0.0441,
|
| 1763 |
+
"num_input_tokens_seen": 520040,
|
| 1764 |
+
"step": 815,
|
| 1765 |
+
"train_runtime": 208.0835,
|
| 1766 |
+
"train_tokens_per_second": 2499.189
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 14.385964912280702,
|
| 1770 |
+
"grad_norm": 0.39123740792274475,
|
| 1771 |
+
"learning_rate": 1.113462155306478e-05,
|
| 1772 |
+
"loss": 0.0417,
|
| 1773 |
+
"num_input_tokens_seen": 523272,
|
| 1774 |
+
"step": 820,
|
| 1775 |
+
"train_runtime": 209.1813,
|
| 1776 |
+
"train_tokens_per_second": 2501.524
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 14.473684210526315,
|
| 1780 |
+
"grad_norm": 0.12004759907722473,
|
| 1781 |
+
"learning_rate": 1.0817771947828934e-05,
|
| 1782 |
+
"loss": 0.0771,
|
| 1783 |
+
"num_input_tokens_seen": 527112,
|
| 1784 |
+
"step": 825,
|
| 1785 |
+
"train_runtime": 210.4461,
|
| 1786 |
+
"train_tokens_per_second": 2504.736
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 14.56140350877193,
|
| 1790 |
+
"grad_norm": 0.46222877502441406,
|
| 1791 |
+
"learning_rate": 1.0504246495664932e-05,
|
| 1792 |
+
"loss": 0.0922,
|
| 1793 |
+
"num_input_tokens_seen": 530568,
|
| 1794 |
+
"step": 830,
|
| 1795 |
+
"train_runtime": 211.5932,
|
| 1796 |
+
"train_tokens_per_second": 2507.491
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 14.649122807017545,
|
| 1800 |
+
"grad_norm": 1.991820216178894,
|
| 1801 |
+
"learning_rate": 1.0194118683375503e-05,
|
| 1802 |
+
"loss": 0.1317,
|
| 1803 |
+
"num_input_tokens_seen": 533192,
|
| 1804 |
+
"step": 835,
|
| 1805 |
+
"train_runtime": 212.5471,
|
| 1806 |
+
"train_tokens_per_second": 2508.582
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 14.736842105263158,
|
| 1810 |
+
"grad_norm": 0.1747927963733673,
|
| 1811 |
+
"learning_rate": 9.887461201395176e-06,
|
| 1812 |
+
"loss": 0.0789,
|
| 1813 |
+
"num_input_tokens_seen": 536232,
|
| 1814 |
+
"step": 840,
|
| 1815 |
+
"train_runtime": 213.6015,
|
| 1816 |
+
"train_tokens_per_second": 2510.432
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 14.824561403508772,
|
| 1820 |
+
"grad_norm": 0.7753216028213501,
|
| 1821 |
+
"learning_rate": 9.584345926752524e-06,
|
| 1822 |
+
"loss": 0.0801,
|
| 1823 |
+
"num_input_tokens_seen": 539624,
|
| 1824 |
+
"step": 845,
|
| 1825 |
+
"train_runtime": 214.7284,
|
| 1826 |
+
"train_tokens_per_second": 2513.053
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 14.912280701754385,
|
| 1830 |
+
"grad_norm": 1.4348740577697754,
|
| 1831 |
+
"learning_rate": 9.284843906222948e-06,
|
| 1832 |
+
"loss": 0.2411,
|
| 1833 |
+
"num_input_tokens_seen": 542664,
|
| 1834 |
+
"step": 850,
|
| 1835 |
+
"train_runtime": 215.7733,
|
| 1836 |
+
"train_tokens_per_second": 2514.973
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 15.0,
|
| 1840 |
+
"grad_norm": 1.8949453830718994,
|
| 1841 |
+
"learning_rate": 8.98902533967618e-06,
|
| 1842 |
+
"loss": 0.1696,
|
| 1843 |
+
"num_input_tokens_seen": 545352,
|
| 1844 |
+
"step": 855,
|
| 1845 |
+
"train_runtime": 216.8312,
|
| 1846 |
+
"train_tokens_per_second": 2515.1
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 15.0,
|
| 1850 |
+
"eval_loss": 0.10493011027574539,
|
| 1851 |
+
"eval_runtime": 0.5851,
|
| 1852 |
+
"eval_samples_per_second": 42.728,
|
| 1853 |
+
"eval_steps_per_second": 11.964,
|
| 1854 |
+
"num_input_tokens_seen": 545352,
|
| 1855 |
+
"step": 855
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 15.087719298245615,
|
| 1859 |
+
"grad_norm": 0.3690445125102997,
|
| 1860 |
+
"learning_rate": 8.696959563622174e-06,
|
| 1861 |
+
"loss": 0.064,
|
| 1862 |
+
"num_input_tokens_seen": 548488,
|
| 1863 |
+
"step": 860,
|
| 1864 |
+
"train_runtime": 219.4293,
|
| 1865 |
+
"train_tokens_per_second": 2499.611
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 15.175438596491228,
|
| 1869 |
+
"grad_norm": 3.105729103088379,
|
| 1870 |
+
"learning_rate": 8.40871503495947e-06,
|
| 1871 |
+
"loss": 0.1508,
|
| 1872 |
+
"num_input_tokens_seen": 552008,
|
| 1873 |
+
"step": 865,
|
| 1874 |
+
"train_runtime": 220.6298,
|
| 1875 |
+
"train_tokens_per_second": 2501.964
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 15.263157894736842,
|
| 1879 |
+
"grad_norm": 0.8427509665489197,
|
| 1880 |
+
"learning_rate": 8.124359314929622e-06,
|
| 1881 |
+
"loss": 0.1135,
|
| 1882 |
+
"num_input_tokens_seen": 555944,
|
| 1883 |
+
"step": 870,
|
| 1884 |
+
"train_runtime": 221.9416,
|
| 1885 |
+
"train_tokens_per_second": 2504.911
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 15.350877192982455,
|
| 1889 |
+
"grad_norm": 0.44307586550712585,
|
| 1890 |
+
"learning_rate": 7.843959053281663e-06,
|
| 1891 |
+
"loss": 0.1025,
|
| 1892 |
+
"num_input_tokens_seen": 558888,
|
| 1893 |
+
"step": 875,
|
| 1894 |
+
"train_runtime": 222.9853,
|
| 1895 |
+
"train_tokens_per_second": 2506.39
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 15.43859649122807,
|
| 1899 |
+
"grad_norm": 0.9975228905677795,
|
| 1900 |
+
"learning_rate": 7.5675799726501155e-06,
|
| 1901 |
+
"loss": 0.1617,
|
| 1902 |
+
"num_input_tokens_seen": 562536,
|
| 1903 |
+
"step": 880,
|
| 1904 |
+
"train_runtime": 224.1818,
|
| 1905 |
+
"train_tokens_per_second": 2509.285
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 15.526315789473685,
|
| 1909 |
+
"grad_norm": 0.6071796417236328,
|
| 1910 |
+
"learning_rate": 7.295286853150391e-06,
|
| 1911 |
+
"loss": 0.0518,
|
| 1912 |
+
"num_input_tokens_seen": 565160,
|
| 1913 |
+
"step": 885,
|
| 1914 |
+
"train_runtime": 225.1357,
|
| 1915 |
+
"train_tokens_per_second": 2510.308
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 15.614035087719298,
|
| 1919 |
+
"grad_norm": 0.6162160038948059,
|
| 1920 |
+
"learning_rate": 7.027143517195023e-06,
|
| 1921 |
+
"loss": 0.0476,
|
| 1922 |
+
"num_input_tokens_seen": 568232,
|
| 1923 |
+
"step": 890,
|
| 1924 |
+
"train_runtime": 226.1893,
|
| 1925 |
+
"train_tokens_per_second": 2512.197
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 15.701754385964913,
|
| 1929 |
+
"grad_norm": 0.6943656206130981,
|
| 1930 |
+
"learning_rate": 6.763212814534484e-06,
|
| 1931 |
+
"loss": 0.2381,
|
| 1932 |
+
"num_input_tokens_seen": 571048,
|
| 1933 |
+
"step": 895,
|
| 1934 |
+
"train_runtime": 227.1756,
|
| 1935 |
+
"train_tokens_per_second": 2513.686
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 15.789473684210526,
|
| 1939 |
+
"grad_norm": 0.7030104994773865,
|
| 1940 |
+
"learning_rate": 6.503556607525838e-06,
|
| 1941 |
+
"loss": 0.0584,
|
| 1942 |
+
"num_input_tokens_seen": 574664,
|
| 1943 |
+
"step": 900,
|
| 1944 |
+
"train_runtime": 228.3878,
|
| 1945 |
+
"train_tokens_per_second": 2516.177
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 15.87719298245614,
|
| 1949 |
+
"grad_norm": 0.8077787756919861,
|
| 1950 |
+
"learning_rate": 6.248235756632984e-06,
|
| 1951 |
+
"loss": 0.1518,
|
| 1952 |
+
"num_input_tokens_seen": 577608,
|
| 1953 |
+
"step": 905,
|
| 1954 |
+
"train_runtime": 229.4146,
|
| 1955 |
+
"train_tokens_per_second": 2517.747
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 15.964912280701755,
|
| 1959 |
+
"grad_norm": 0.8616611361503601,
|
| 1960 |
+
"learning_rate": 5.997310106161589e-06,
|
| 1961 |
+
"loss": 0.1024,
|
| 1962 |
+
"num_input_tokens_seen": 580680,
|
| 1963 |
+
"step": 910,
|
| 1964 |
+
"train_runtime": 230.4206,
|
| 1965 |
+
"train_tokens_per_second": 2520.088
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 16.0,
|
| 1969 |
+
"eval_loss": 0.10666210204362869,
|
| 1970 |
+
"eval_runtime": 0.5897,
|
| 1971 |
+
"eval_samples_per_second": 42.392,
|
| 1972 |
+
"eval_steps_per_second": 11.87,
|
| 1973 |
+
"num_input_tokens_seen": 581368,
|
| 1974 |
+
"step": 912
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 16.05263157894737,
|
| 1978 |
+
"grad_norm": 0.42227837443351746,
|
| 1979 |
+
"learning_rate": 5.7508384702323226e-06,
|
| 1980 |
+
"loss": 0.0354,
|
| 1981 |
+
"num_input_tokens_seen": 583544,
|
| 1982 |
+
"step": 915,
|
| 1983 |
+
"train_runtime": 232.9828,
|
| 1984 |
+
"train_tokens_per_second": 2504.666
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 16.140350877192983,
|
| 1988 |
+
"grad_norm": 0.22992517054080963,
|
| 1989 |
+
"learning_rate": 5.508878618995439e-06,
|
| 1990 |
+
"loss": 0.1062,
|
| 1991 |
+
"num_input_tokens_seen": 586424,
|
| 1992 |
+
"step": 920,
|
| 1993 |
+
"train_runtime": 234.019,
|
| 1994 |
+
"train_tokens_per_second": 2505.882
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 16.228070175438596,
|
| 1998 |
+
"grad_norm": 0.6025696396827698,
|
| 1999 |
+
"learning_rate": 5.271487265090163e-06,
|
| 2000 |
+
"loss": 0.1369,
|
| 2001 |
+
"num_input_tokens_seen": 589976,
|
| 2002 |
+
"step": 925,
|
| 2003 |
+
"train_runtime": 235.2016,
|
| 2004 |
+
"train_tokens_per_second": 2508.384
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 16.31578947368421,
|
| 2008 |
+
"grad_norm": 0.1732168048620224,
|
| 2009 |
+
"learning_rate": 5.038720050351842e-06,
|
| 2010 |
+
"loss": 0.0959,
|
| 2011 |
+
"num_input_tokens_seen": 593048,
|
| 2012 |
+
"step": 930,
|
| 2013 |
+
"train_runtime": 236.2527,
|
| 2014 |
+
"train_tokens_per_second": 2510.227
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 16.403508771929825,
|
| 2018 |
+
"grad_norm": 0.4299314320087433,
|
| 2019 |
+
"learning_rate": 4.810631532770182e-06,
|
| 2020 |
+
"loss": 0.0776,
|
| 2021 |
+
"num_input_tokens_seen": 596472,
|
| 2022 |
+
"step": 935,
|
| 2023 |
+
"train_runtime": 237.3915,
|
| 2024 |
+
"train_tokens_per_second": 2512.609
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 16.49122807017544,
|
| 2028 |
+
"grad_norm": 0.14484725892543793,
|
| 2029 |
+
"learning_rate": 4.587275173701428e-06,
|
| 2030 |
+
"loss": 0.2379,
|
| 2031 |
+
"num_input_tokens_seen": 599544,
|
| 2032 |
+
"step": 940,
|
| 2033 |
+
"train_runtime": 238.4376,
|
| 2034 |
+
"train_tokens_per_second": 2514.469
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 16.57894736842105,
|
| 2038 |
+
"grad_norm": 1.2683119773864746,
|
| 2039 |
+
"learning_rate": 4.368703325337667e-06,
|
| 2040 |
+
"loss": 0.0572,
|
| 2041 |
+
"num_input_tokens_seen": 603032,
|
| 2042 |
+
"step": 945,
|
| 2043 |
+
"train_runtime": 239.5986,
|
| 2044 |
+
"train_tokens_per_second": 2516.842
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 16.666666666666668,
|
| 2048 |
+
"grad_norm": 0.46702468395233154,
|
| 2049 |
+
"learning_rate": 4.154967218436037e-06,
|
| 2050 |
+
"loss": 0.0573,
|
| 2051 |
+
"num_input_tokens_seen": 605560,
|
| 2052 |
+
"step": 950,
|
| 2053 |
+
"train_runtime": 240.5398,
|
| 2054 |
+
"train_tokens_per_second": 2517.504
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 16.75438596491228,
|
| 2058 |
+
"grad_norm": 0.1284426748752594,
|
| 2059 |
+
"learning_rate": 3.94611695031086e-06,
|
| 2060 |
+
"loss": 0.1545,
|
| 2061 |
+
"num_input_tokens_seen": 609176,
|
| 2062 |
+
"step": 955,
|
| 2063 |
+
"train_runtime": 241.7744,
|
| 2064 |
+
"train_tokens_per_second": 2519.605
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 16.842105263157894,
|
| 2068 |
+
"grad_norm": 0.9069837927818298,
|
| 2069 |
+
"learning_rate": 3.74220147309135e-06,
|
| 2070 |
+
"loss": 0.1051,
|
| 2071 |
+
"num_input_tokens_seen": 612024,
|
| 2072 |
+
"step": 960,
|
| 2073 |
+
"train_runtime": 242.7614,
|
| 2074 |
+
"train_tokens_per_second": 2521.092
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 16.92982456140351,
|
| 2078 |
+
"grad_norm": 0.25688862800598145,
|
| 2079 |
+
"learning_rate": 3.543268582247844e-06,
|
| 2080 |
+
"loss": 0.0598,
|
| 2081 |
+
"num_input_tokens_seen": 615064,
|
| 2082 |
+
"step": 965,
|
| 2083 |
+
"train_runtime": 243.8119,
|
| 2084 |
+
"train_tokens_per_second": 2522.698
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 17.0,
|
| 2088 |
+
"eval_loss": 0.10478480905294418,
|
| 2089 |
+
"eval_runtime": 0.5911,
|
| 2090 |
+
"eval_samples_per_second": 42.294,
|
| 2091 |
+
"eval_steps_per_second": 11.842,
|
| 2092 |
+
"num_input_tokens_seen": 616776,
|
| 2093 |
+
"step": 969
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 17.017543859649123,
|
| 2097 |
+
"grad_norm": 0.20608949661254883,
|
| 2098 |
+
"learning_rate": 3.3493649053890326e-06,
|
| 2099 |
+
"loss": 0.132,
|
| 2100 |
+
"num_input_tokens_seen": 617640,
|
| 2101 |
+
"step": 970,
|
| 2102 |
+
"train_runtime": 246.4873,
|
| 2103 |
+
"train_tokens_per_second": 2505.768
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 17.105263157894736,
|
| 2107 |
+
"grad_norm": 1.5790821313858032,
|
| 2108 |
+
"learning_rate": 3.1605358913330385e-06,
|
| 2109 |
+
"loss": 0.1705,
|
| 2110 |
+
"num_input_tokens_seen": 620424,
|
| 2111 |
+
"step": 975,
|
| 2112 |
+
"train_runtime": 247.4926,
|
| 2113 |
+
"train_tokens_per_second": 2506.839
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 17.19298245614035,
|
| 2117 |
+
"grad_norm": 0.38555988669395447,
|
| 2118 |
+
"learning_rate": 2.9768257994546662e-06,
|
| 2119 |
+
"loss": 0.0891,
|
| 2120 |
+
"num_input_tokens_seen": 623176,
|
| 2121 |
+
"step": 980,
|
| 2122 |
+
"train_runtime": 248.4814,
|
| 2123 |
+
"train_tokens_per_second": 2507.938
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 17.280701754385966,
|
| 2127 |
+
"grad_norm": 0.05342285707592964,
|
| 2128 |
+
"learning_rate": 2.7982776893115627e-06,
|
| 2129 |
+
"loss": 0.029,
|
| 2130 |
+
"num_input_tokens_seen": 626600,
|
| 2131 |
+
"step": 985,
|
| 2132 |
+
"train_runtime": 249.6185,
|
| 2133 |
+
"train_tokens_per_second": 2510.231
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 17.36842105263158,
|
| 2137 |
+
"grad_norm": 0.2652755081653595,
|
| 2138 |
+
"learning_rate": 2.624933410551508e-06,
|
| 2139 |
+
"loss": 0.036,
|
| 2140 |
+
"num_input_tokens_seen": 629992,
|
| 2141 |
+
"step": 990,
|
| 2142 |
+
"train_runtime": 250.7293,
|
| 2143 |
+
"train_tokens_per_second": 2512.638
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 17.45614035087719,
|
| 2147 |
+
"grad_norm": 1.1502621173858643,
|
| 2148 |
+
"learning_rate": 2.456833593103361e-06,
|
| 2149 |
+
"loss": 0.0807,
|
| 2150 |
+
"num_input_tokens_seen": 633128,
|
| 2151 |
+
"step": 995,
|
| 2152 |
+
"train_runtime": 251.8025,
|
| 2153 |
+
"train_tokens_per_second": 2514.383
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 17.54385964912281,
|
| 2157 |
+
"grad_norm": 1.4065945148468018,
|
| 2158 |
+
"learning_rate": 2.2940176376538445e-06,
|
| 2159 |
+
"loss": 0.098,
|
| 2160 |
+
"num_input_tokens_seen": 635880,
|
| 2161 |
+
"step": 1000,
|
| 2162 |
+
"train_runtime": 252.7529,
|
| 2163 |
+
"train_tokens_per_second": 2515.817
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 17.63157894736842,
|
| 2167 |
+
"grad_norm": 0.12156794220209122,
|
| 2168 |
+
"learning_rate": 2.136523706412477e-06,
|
| 2169 |
+
"loss": 0.075,
|
| 2170 |
+
"num_input_tokens_seen": 639208,
|
| 2171 |
+
"step": 1005,
|
| 2172 |
+
"train_runtime": 253.8752,
|
| 2173 |
+
"train_tokens_per_second": 2517.804
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 17.719298245614034,
|
| 2177 |
+
"grad_norm": 0.29780030250549316,
|
| 2178 |
+
"learning_rate": 1.984388714166799e-06,
|
| 2179 |
+
"loss": 0.2354,
|
| 2180 |
+
"num_input_tokens_seen": 642152,
|
| 2181 |
+
"step": 1010,
|
| 2182 |
+
"train_runtime": 254.896,
|
| 2183 |
+
"train_tokens_per_second": 2519.27
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 17.80701754385965,
|
| 2187 |
+
"grad_norm": 0.4195615351200104,
|
| 2188 |
+
"learning_rate": 1.837648319629956e-06,
|
| 2189 |
+
"loss": 0.1131,
|
| 2190 |
+
"num_input_tokens_seen": 645800,
|
| 2191 |
+
"step": 1015,
|
| 2192 |
+
"train_runtime": 256.1081,
|
| 2193 |
+
"train_tokens_per_second": 2521.591
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 17.894736842105264,
|
| 2197 |
+
"grad_norm": 1.0894429683685303,
|
| 2198 |
+
"learning_rate": 1.6963369170826943e-06,
|
| 2199 |
+
"loss": 0.2088,
|
| 2200 |
+
"num_input_tokens_seen": 649736,
|
| 2201 |
+
"step": 1020,
|
| 2202 |
+
"train_runtime": 257.3817,
|
| 2203 |
+
"train_tokens_per_second": 2524.406
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 17.982456140350877,
|
| 2207 |
+
"grad_norm": 0.8357132077217102,
|
| 2208 |
+
"learning_rate": 1.5604876283117326e-06,
|
| 2209 |
+
"loss": 0.0994,
|
| 2210 |
+
"num_input_tokens_seen": 653064,
|
| 2211 |
+
"step": 1025,
|
| 2212 |
+
"train_runtime": 258.4933,
|
| 2213 |
+
"train_tokens_per_second": 2526.425
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 18.0,
|
| 2217 |
+
"eval_loss": 0.10452605038881302,
|
| 2218 |
+
"eval_runtime": 0.5882,
|
| 2219 |
+
"eval_samples_per_second": 42.499,
|
| 2220 |
+
"eval_steps_per_second": 11.9,
|
| 2221 |
+
"num_input_tokens_seen": 653152,
|
| 2222 |
+
"step": 1026
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 18.07017543859649,
|
| 2226 |
+
"grad_norm": 0.881624698638916,
|
| 2227 |
+
"learning_rate": 1.4301322948464147e-06,
|
| 2228 |
+
"loss": 0.2031,
|
| 2229 |
+
"num_input_tokens_seen": 655424,
|
| 2230 |
+
"step": 1030,
|
| 2231 |
+
"train_runtime": 261.0397,
|
| 2232 |
+
"train_tokens_per_second": 2510.822
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 18.157894736842106,
|
| 2236 |
+
"grad_norm": 0.88265061378479,
|
| 2237 |
+
"learning_rate": 1.3053014704953987e-06,
|
| 2238 |
+
"loss": 0.1471,
|
| 2239 |
+
"num_input_tokens_seen": 658848,
|
| 2240 |
+
"step": 1035,
|
| 2241 |
+
"train_runtime": 262.1635,
|
| 2242 |
+
"train_tokens_per_second": 2513.119
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 18.24561403508772,
|
| 2246 |
+
"grad_norm": 0.20473460853099823,
|
| 2247 |
+
"learning_rate": 1.1860244141851773e-06,
|
| 2248 |
+
"loss": 0.0341,
|
| 2249 |
+
"num_input_tokens_seen": 661888,
|
| 2250 |
+
"step": 1040,
|
| 2251 |
+
"train_runtime": 263.2303,
|
| 2252 |
+
"train_tokens_per_second": 2514.483
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 18.333333333333332,
|
| 2256 |
+
"grad_norm": 0.7834029197692871,
|
| 2257 |
+
"learning_rate": 1.0723290831021471e-06,
|
| 2258 |
+
"loss": 0.0833,
|
| 2259 |
+
"num_input_tokens_seen": 665664,
|
| 2260 |
+
"step": 1045,
|
| 2261 |
+
"train_runtime": 264.4893,
|
| 2262 |
+
"train_tokens_per_second": 2516.79
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 18.42105263157895,
|
| 2266 |
+
"grad_norm": 0.6192132234573364,
|
| 2267 |
+
"learning_rate": 9.642421261397472e-07,
|
| 2268 |
+
"loss": 0.0973,
|
| 2269 |
+
"num_input_tokens_seen": 668896,
|
| 2270 |
+
"step": 1050,
|
| 2271 |
+
"train_runtime": 265.5927,
|
| 2272 |
+
"train_tokens_per_second": 2518.503
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 18.50877192982456,
|
| 2276 |
+
"grad_norm": 0.48073986172676086,
|
| 2277 |
+
"learning_rate": 8.617888776522642e-07,
|
| 2278 |
+
"loss": 0.064,
|
| 2279 |
+
"num_input_tokens_seen": 672352,
|
| 2280 |
+
"step": 1055,
|
| 2281 |
+
"train_runtime": 266.7674,
|
| 2282 |
+
"train_tokens_per_second": 2520.368
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 18.596491228070175,
|
| 2286 |
+
"grad_norm": 0.39590880274772644,
|
| 2287 |
+
"learning_rate": 7.649933515167407e-07,
|
| 2288 |
+
"loss": 0.1197,
|
| 2289 |
+
"num_input_tokens_seen": 675392,
|
| 2290 |
+
"step": 1060,
|
| 2291 |
+
"train_runtime": 267.8115,
|
| 2292 |
+
"train_tokens_per_second": 2521.893
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 18.68421052631579,
|
| 2296 |
+
"grad_norm": 0.5385900735855103,
|
| 2297 |
+
"learning_rate": 6.738782355044049e-07,
|
| 2298 |
+
"loss": 0.0847,
|
| 2299 |
+
"num_input_tokens_seen": 678080,
|
| 2300 |
+
"step": 1065,
|
| 2301 |
+
"train_runtime": 268.7582,
|
| 2302 |
+
"train_tokens_per_second": 2523.012
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 18.771929824561404,
|
| 2306 |
+
"grad_norm": 0.2719675302505493,
|
| 2307 |
+
"learning_rate": 5.88464885962911e-07,
|
| 2308 |
+
"loss": 0.0278,
|
| 2309 |
+
"num_input_tokens_seen": 681440,
|
| 2310 |
+
"step": 1070,
|
| 2311 |
+
"train_runtime": 269.8918,
|
| 2312 |
+
"train_tokens_per_second": 2524.864
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 18.859649122807017,
|
| 2316 |
+
"grad_norm": 0.6708596348762512,
|
| 2317 |
+
"learning_rate": 5.087733228106517e-07,
|
| 2318 |
+
"loss": 0.0769,
|
| 2319 |
+
"num_input_tokens_seen": 684608,
|
| 2320 |
+
"step": 1075,
|
| 2321 |
+
"train_runtime": 270.9537,
|
| 2322 |
+
"train_tokens_per_second": 2526.661
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 18.94736842105263,
|
| 2326 |
+
"grad_norm": 0.47063499689102173,
|
| 2327 |
+
"learning_rate": 4.3482222484432513e-07,
|
| 2328 |
+
"loss": 0.0973,
|
| 2329 |
+
"num_input_tokens_seen": 688480,
|
| 2330 |
+
"step": 1080,
|
| 2331 |
+
"train_runtime": 272.2182,
|
| 2332 |
+
"train_tokens_per_second": 2529.148
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 19.0,
|
| 2336 |
+
"eval_loss": 0.1085146814584732,
|
| 2337 |
+
"eval_runtime": 0.5884,
|
| 2338 |
+
"eval_samples_per_second": 42.488,
|
| 2339 |
+
"eval_steps_per_second": 11.897,
|
| 2340 |
+
"num_input_tokens_seen": 689856,
|
| 2341 |
+
"step": 1083
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 19.035087719298247,
|
| 2345 |
+
"grad_norm": 0.8871979117393494,
|
| 2346 |
+
"learning_rate": 3.666289253608235e-07,
|
| 2347 |
+
"loss": 0.5679,
|
| 2348 |
+
"num_input_tokens_seen": 691200,
|
| 2349 |
+
"step": 1085,
|
| 2350 |
+
"train_runtime": 274.7464,
|
| 2351 |
+
"train_tokens_per_second": 2515.774
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 19.12280701754386,
|
| 2355 |
+
"grad_norm": 2.014510154724121,
|
| 2356 |
+
"learning_rate": 3.0420940809451624e-07,
|
| 2357 |
+
"loss": 0.2126,
|
| 2358 |
+
"num_input_tokens_seen": 694240,
|
| 2359 |
+
"step": 1090,
|
| 2360 |
+
"train_runtime": 275.8315,
|
| 2361 |
+
"train_tokens_per_second": 2516.899
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 19.210526315789473,
|
| 2365 |
+
"grad_norm": 0.20251081883907318,
|
| 2366 |
+
"learning_rate": 2.47578303470844e-07,
|
| 2367 |
+
"loss": 0.0572,
|
| 2368 |
+
"num_input_tokens_seen": 697472,
|
| 2369 |
+
"step": 1095,
|
| 2370 |
+
"train_runtime": 276.9291,
|
| 2371 |
+
"train_tokens_per_second": 2518.594
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 19.29824561403509,
|
| 2375 |
+
"grad_norm": 0.9537681341171265,
|
| 2376 |
+
"learning_rate": 1.96748885177106e-07,
|
| 2377 |
+
"loss": 0.2066,
|
| 2378 |
+
"num_input_tokens_seen": 699872,
|
| 2379 |
+
"step": 1100,
|
| 2380 |
+
"train_runtime": 277.8148,
|
| 2381 |
+
"train_tokens_per_second": 2519.204
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 19.385964912280702,
|
| 2385 |
+
"grad_norm": 0.5312708616256714,
|
| 2386 |
+
"learning_rate": 1.517330670512629e-07,
|
| 2387 |
+
"loss": 0.2041,
|
| 2388 |
+
"num_input_tokens_seen": 704736,
|
| 2389 |
+
"step": 1105,
|
| 2390 |
+
"train_runtime": 279.3407,
|
| 2391 |
+
"train_tokens_per_second": 2522.855
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 19.473684210526315,
|
| 2395 |
+
"grad_norm": 0.2024543583393097,
|
| 2396 |
+
"learning_rate": 1.125414002894759e-07,
|
| 2397 |
+
"loss": 0.0422,
|
| 2398 |
+
"num_input_tokens_seen": 707968,
|
| 2399 |
+
"step": 1110,
|
| 2400 |
+
"train_runtime": 280.4278,
|
| 2401 |
+
"train_tokens_per_second": 2524.6
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 19.56140350877193,
|
| 2405 |
+
"grad_norm": 0.7978918552398682,
|
| 2406 |
+
"learning_rate": 7.918307097301014e-08,
|
| 2407 |
+
"loss": 0.1165,
|
| 2408 |
+
"num_input_tokens_seen": 711136,
|
| 2409 |
+
"step": 1115,
|
| 2410 |
+
"train_runtime": 281.5081,
|
| 2411 |
+
"train_tokens_per_second": 2526.166
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 19.649122807017545,
|
| 2415 |
+
"grad_norm": 0.37032362818717957,
|
| 2416 |
+
"learning_rate": 5.166589791513465e-08,
|
| 2417 |
+
"loss": 0.0263,
|
| 2418 |
+
"num_input_tokens_seen": 714432,
|
| 2419 |
+
"step": 1120,
|
| 2420 |
+
"train_runtime": 282.6136,
|
| 2421 |
+
"train_tokens_per_second": 2527.947
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 19.736842105263158,
|
| 2425 |
+
"grad_norm": 0.10990393161773682,
|
| 2426 |
+
"learning_rate": 2.999633082847453e-08,
|
| 2427 |
+
"loss": 0.0151,
|
| 2428 |
+
"num_input_tokens_seen": 717440,
|
| 2429 |
+
"step": 1125,
|
| 2430 |
+
"train_runtime": 283.6603,
|
| 2431 |
+
"train_tokens_per_second": 2529.223
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 19.82456140350877,
|
| 2435 |
+
"grad_norm": 0.7865080833435059,
|
| 2436 |
+
"learning_rate": 1.4179448813278484e-08,
|
| 2437 |
+
"loss": 0.1254,
|
| 2438 |
+
"num_input_tokens_seen": 720480,
|
| 2439 |
+
"step": 1130,
|
| 2440 |
+
"train_runtime": 284.7252,
|
| 2441 |
+
"train_tokens_per_second": 2530.44
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 19.912280701754387,
|
| 2445 |
+
"grad_norm": 0.35326337814331055,
|
| 2446 |
+
"learning_rate": 4.218959166932268e-09,
|
| 2447 |
+
"loss": 0.1746,
|
| 2448 |
+
"num_input_tokens_seen": 723552,
|
| 2449 |
+
"step": 1135,
|
| 2450 |
+
"train_runtime": 285.8181,
|
| 2451 |
+
"train_tokens_per_second": 2531.512
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 20.0,
|
| 2455 |
+
"grad_norm": 0.42669206857681274,
|
| 2456 |
+
"learning_rate": 1.1719651499819683e-10,
|
| 2457 |
+
"loss": 0.0289,
|
| 2458 |
+
"num_input_tokens_seen": 725992,
|
| 2459 |
+
"step": 1140,
|
| 2460 |
+
"train_runtime": 286.7994,
|
| 2461 |
+
"train_tokens_per_second": 2531.358
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 20.0,
|
| 2465 |
+
"eval_loss": 0.10313867032527924,
|
| 2466 |
+
"eval_runtime": 0.5919,
|
| 2467 |
+
"eval_samples_per_second": 42.24,
|
| 2468 |
+
"eval_steps_per_second": 11.827,
|
| 2469 |
+
"num_input_tokens_seen": 725992,
|
| 2470 |
+
"step": 1140
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 20.0,
|
| 2474 |
+
"num_input_tokens_seen": 725992,
|
| 2475 |
+
"step": 1140,
|
| 2476 |
+
"total_flos": 3.269194804985856e+16,
|
| 2477 |
+
"train_loss": 0.22796724926222833,
|
| 2478 |
+
"train_runtime": 288.2339,
|
| 2479 |
+
"train_samples_per_second": 15.612,
|
| 2480 |
+
"train_steps_per_second": 3.955
|
| 2481 |
+
}
|
| 2482 |
+
],
|
| 2483 |
+
"logging_steps": 5,
|
| 2484 |
+
"max_steps": 1140,
|
| 2485 |
+
"num_input_tokens_seen": 725992,
|
| 2486 |
+
"num_train_epochs": 20,
|
| 2487 |
+
"save_steps": 57,
|
| 2488 |
+
"stateful_callbacks": {
|
| 2489 |
+
"TrainerControl": {
|
| 2490 |
+
"args": {
|
| 2491 |
+
"should_epoch_stop": false,
|
| 2492 |
+
"should_evaluate": false,
|
| 2493 |
+
"should_log": false,
|
| 2494 |
+
"should_save": true,
|
| 2495 |
+
"should_training_stop": true
|
| 2496 |
+
},
|
| 2497 |
+
"attributes": {}
|
| 2498 |
+
}
|
| 2499 |
+
},
|
| 2500 |
+
"total_flos": 3.269194804985856e+16,
|
| 2501 |
+
"train_batch_size": 4,
|
| 2502 |
+
"trial_name": null,
|
| 2503 |
+
"trial_params": null
|
| 2504 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|