End of training
Browse files- README.md +2 -1
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +9 -0
- trainer_state.json +1568 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: train_wsc_42_1760466772
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# train_wsc_42_1760466772
|
| 17 |
|
| 18 |
-
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
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.3524
|
| 21 |
- Num Input Tokens Seen: 1468632
|
|
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- prefix-tuning
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: train_wsc_42_1760466772
|
|
|
|
| 16 |
|
| 17 |
# train_wsc_42_1760466772
|
| 18 |
|
| 19 |
+
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 wsc dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.3524
|
| 22 |
- Num Input Tokens Seen: 1468632
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_loss": 0.3523610234260559,
|
| 4 |
+
"eval_runtime": 1.1287,
|
| 5 |
+
"eval_samples_per_second": 98.345,
|
| 6 |
+
"eval_steps_per_second": 12.404,
|
| 7 |
+
"num_input_tokens_seen": 1468632,
|
| 8 |
+
"total_flos": 6.613183518533222e+16,
|
| 9 |
+
"train_loss": 0.47382661629290806,
|
| 10 |
+
"train_runtime": 311.6966,
|
| 11 |
+
"train_samples_per_second": 42.638,
|
| 12 |
+
"train_steps_per_second": 2.695
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_loss": 0.3523610234260559,
|
| 4 |
+
"eval_runtime": 1.1287,
|
| 5 |
+
"eval_samples_per_second": 98.345,
|
| 6 |
+
"eval_steps_per_second": 12.404,
|
| 7 |
+
"num_input_tokens_seen": 1468632
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"num_input_tokens_seen": 1468632,
|
| 4 |
+
"total_flos": 6.613183518533222e+16,
|
| 5 |
+
"train_loss": 0.47382661629290806,
|
| 6 |
+
"train_runtime": 311.6966,
|
| 7 |
+
"train_samples_per_second": 42.638,
|
| 8 |
+
"train_steps_per_second": 2.695
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1568 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 210,
|
| 3 |
+
"best_metric": 0.34583956003189087,
|
| 4 |
+
"best_model_checkpoint": "saves/prefix-tuning/llama-3-8b-instruct/train_wsc_42_1760466772/checkpoint-210",
|
| 5 |
+
"epoch": 30.0,
|
| 6 |
+
"eval_steps": 42,
|
| 7 |
+
"global_step": 840,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.17857142857142858,
|
| 14 |
+
"grad_norm": 175.3666534423828,
|
| 15 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 16 |
+
"loss": 11.3214,
|
| 17 |
+
"num_input_tokens_seen": 9216,
|
| 18 |
+
"step": 5
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.35714285714285715,
|
| 22 |
+
"grad_norm": 145.5986785888672,
|
| 23 |
+
"learning_rate": 7.142857142857143e-06,
|
| 24 |
+
"loss": 6.8093,
|
| 25 |
+
"num_input_tokens_seen": 18112,
|
| 26 |
+
"step": 10
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.5357142857142857,
|
| 30 |
+
"grad_norm": 58.49702453613281,
|
| 31 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 32 |
+
"loss": 2.1384,
|
| 33 |
+
"num_input_tokens_seen": 26624,
|
| 34 |
+
"step": 15
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.7142857142857143,
|
| 38 |
+
"grad_norm": 57.37192153930664,
|
| 39 |
+
"learning_rate": 1.5079365079365079e-05,
|
| 40 |
+
"loss": 0.7391,
|
| 41 |
+
"num_input_tokens_seen": 35456,
|
| 42 |
+
"step": 20
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.8928571428571429,
|
| 46 |
+
"grad_norm": 20.545907974243164,
|
| 47 |
+
"learning_rate": 1.9047619047619046e-05,
|
| 48 |
+
"loss": 0.5556,
|
| 49 |
+
"num_input_tokens_seen": 44096,
|
| 50 |
+
"step": 25
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0714285714285714,
|
| 54 |
+
"grad_norm": 25.21103286743164,
|
| 55 |
+
"learning_rate": 2.3015873015873015e-05,
|
| 56 |
+
"loss": 0.5487,
|
| 57 |
+
"num_input_tokens_seen": 52192,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.25,
|
| 62 |
+
"grad_norm": 8.62804126739502,
|
| 63 |
+
"learning_rate": 2.6984126984126984e-05,
|
| 64 |
+
"loss": 0.4509,
|
| 65 |
+
"num_input_tokens_seen": 60896,
|
| 66 |
+
"step": 35
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.4285714285714286,
|
| 70 |
+
"grad_norm": 23.560665130615234,
|
| 71 |
+
"learning_rate": 3.095238095238095e-05,
|
| 72 |
+
"loss": 0.4818,
|
| 73 |
+
"num_input_tokens_seen": 69920,
|
| 74 |
+
"step": 40
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.5,
|
| 78 |
+
"eval_loss": 0.3624400794506073,
|
| 79 |
+
"eval_runtime": 1.118,
|
| 80 |
+
"eval_samples_per_second": 99.287,
|
| 81 |
+
"eval_steps_per_second": 12.523,
|
| 82 |
+
"num_input_tokens_seen": 73824,
|
| 83 |
+
"step": 42
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.6071428571428572,
|
| 87 |
+
"grad_norm": 13.207117080688477,
|
| 88 |
+
"learning_rate": 3.492063492063492e-05,
|
| 89 |
+
"loss": 0.4103,
|
| 90 |
+
"num_input_tokens_seen": 78560,
|
| 91 |
+
"step": 45
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.7857142857142856,
|
| 95 |
+
"grad_norm": 11.927769660949707,
|
| 96 |
+
"learning_rate": 3.888888888888889e-05,
|
| 97 |
+
"loss": 0.4405,
|
| 98 |
+
"num_input_tokens_seen": 87136,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.9642857142857144,
|
| 103 |
+
"grad_norm": 7.089166641235352,
|
| 104 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 105 |
+
"loss": 0.408,
|
| 106 |
+
"num_input_tokens_seen": 96480,
|
| 107 |
+
"step": 55
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.142857142857143,
|
| 111 |
+
"grad_norm": 2.959946393966675,
|
| 112 |
+
"learning_rate": 4.682539682539683e-05,
|
| 113 |
+
"loss": 0.3548,
|
| 114 |
+
"num_input_tokens_seen": 104248,
|
| 115 |
+
"step": 60
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 2.3214285714285716,
|
| 119 |
+
"grad_norm": 1.298177719116211,
|
| 120 |
+
"learning_rate": 5.0793650793650794e-05,
|
| 121 |
+
"loss": 0.377,
|
| 122 |
+
"num_input_tokens_seen": 113272,
|
| 123 |
+
"step": 65
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.5,
|
| 127 |
+
"grad_norm": 8.108621597290039,
|
| 128 |
+
"learning_rate": 5.4761904761904766e-05,
|
| 129 |
+
"loss": 0.3922,
|
| 130 |
+
"num_input_tokens_seen": 122168,
|
| 131 |
+
"step": 70
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 2.678571428571429,
|
| 135 |
+
"grad_norm": 2.4724233150482178,
|
| 136 |
+
"learning_rate": 5.873015873015873e-05,
|
| 137 |
+
"loss": 0.4379,
|
| 138 |
+
"num_input_tokens_seen": 131576,
|
| 139 |
+
"step": 75
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.857142857142857,
|
| 143 |
+
"grad_norm": 4.108138084411621,
|
| 144 |
+
"learning_rate": 6.26984126984127e-05,
|
| 145 |
+
"loss": 0.3847,
|
| 146 |
+
"num_input_tokens_seen": 139832,
|
| 147 |
+
"step": 80
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.0,
|
| 151 |
+
"eval_loss": 0.3848850131034851,
|
| 152 |
+
"eval_runtime": 1.1397,
|
| 153 |
+
"eval_samples_per_second": 97.398,
|
| 154 |
+
"eval_steps_per_second": 12.284,
|
| 155 |
+
"num_input_tokens_seen": 146552,
|
| 156 |
+
"step": 84
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.0357142857142856,
|
| 160 |
+
"grad_norm": 0.5216990113258362,
|
| 161 |
+
"learning_rate": 6.666666666666667e-05,
|
| 162 |
+
"loss": 0.3722,
|
| 163 |
+
"num_input_tokens_seen": 147832,
|
| 164 |
+
"step": 85
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 3.2142857142857144,
|
| 168 |
+
"grad_norm": 0.9133898615837097,
|
| 169 |
+
"learning_rate": 7.063492063492065e-05,
|
| 170 |
+
"loss": 0.4223,
|
| 171 |
+
"num_input_tokens_seen": 157176,
|
| 172 |
+
"step": 90
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.392857142857143,
|
| 176 |
+
"grad_norm": 4.493597030639648,
|
| 177 |
+
"learning_rate": 7.460317460317461e-05,
|
| 178 |
+
"loss": 0.388,
|
| 179 |
+
"num_input_tokens_seen": 166392,
|
| 180 |
+
"step": 95
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 3.571428571428571,
|
| 184 |
+
"grad_norm": 0.2193046361207962,
|
| 185 |
+
"learning_rate": 7.857142857142858e-05,
|
| 186 |
+
"loss": 0.3655,
|
| 187 |
+
"num_input_tokens_seen": 174904,
|
| 188 |
+
"step": 100
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.75,
|
| 192 |
+
"grad_norm": 3.4654486179351807,
|
| 193 |
+
"learning_rate": 8.253968253968255e-05,
|
| 194 |
+
"loss": 0.4285,
|
| 195 |
+
"num_input_tokens_seen": 184312,
|
| 196 |
+
"step": 105
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 3.928571428571429,
|
| 200 |
+
"grad_norm": 0.2327776700258255,
|
| 201 |
+
"learning_rate": 8.650793650793651e-05,
|
| 202 |
+
"loss": 0.3877,
|
| 203 |
+
"num_input_tokens_seen": 193464,
|
| 204 |
+
"step": 110
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 4.107142857142857,
|
| 208 |
+
"grad_norm": 0.6309562921524048,
|
| 209 |
+
"learning_rate": 9.047619047619048e-05,
|
| 210 |
+
"loss": 0.3744,
|
| 211 |
+
"num_input_tokens_seen": 201168,
|
| 212 |
+
"step": 115
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.285714285714286,
|
| 216 |
+
"grad_norm": 0.7549311518669128,
|
| 217 |
+
"learning_rate": 9.444444444444444e-05,
|
| 218 |
+
"loss": 0.4027,
|
| 219 |
+
"num_input_tokens_seen": 210384,
|
| 220 |
+
"step": 120
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 4.464285714285714,
|
| 224 |
+
"grad_norm": 0.4532535672187805,
|
| 225 |
+
"learning_rate": 9.841269841269841e-05,
|
| 226 |
+
"loss": 0.3464,
|
| 227 |
+
"num_input_tokens_seen": 219536,
|
| 228 |
+
"step": 125
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.5,
|
| 232 |
+
"eval_loss": 0.34812429547309875,
|
| 233 |
+
"eval_runtime": 1.0994,
|
| 234 |
+
"eval_samples_per_second": 100.966,
|
| 235 |
+
"eval_steps_per_second": 12.735,
|
| 236 |
+
"num_input_tokens_seen": 221264,
|
| 237 |
+
"step": 126
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 4.642857142857143,
|
| 241 |
+
"grad_norm": 0.14779126644134521,
|
| 242 |
+
"learning_rate": 9.999564408362054e-05,
|
| 243 |
+
"loss": 0.333,
|
| 244 |
+
"num_input_tokens_seen": 228432,
|
| 245 |
+
"step": 130
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 4.821428571428571,
|
| 249 |
+
"grad_norm": 0.35166576504707336,
|
| 250 |
+
"learning_rate": 9.996902734308346e-05,
|
| 251 |
+
"loss": 0.3557,
|
| 252 |
+
"num_input_tokens_seen": 238032,
|
| 253 |
+
"step": 135
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 5.0,
|
| 257 |
+
"grad_norm": 0.6182653307914734,
|
| 258 |
+
"learning_rate": 9.991822668185927e-05,
|
| 259 |
+
"loss": 0.3749,
|
| 260 |
+
"num_input_tokens_seen": 245376,
|
| 261 |
+
"step": 140
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.178571428571429,
|
| 265 |
+
"grad_norm": 0.5397220253944397,
|
| 266 |
+
"learning_rate": 9.984326668636131e-05,
|
| 267 |
+
"loss": 0.3563,
|
| 268 |
+
"num_input_tokens_seen": 253952,
|
| 269 |
+
"step": 145
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 5.357142857142857,
|
| 273 |
+
"grad_norm": 0.11434465646743774,
|
| 274 |
+
"learning_rate": 9.974418363559444e-05,
|
| 275 |
+
"loss": 0.3447,
|
| 276 |
+
"num_input_tokens_seen": 263296,
|
| 277 |
+
"step": 150
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 5.535714285714286,
|
| 281 |
+
"grad_norm": 0.9336232542991638,
|
| 282 |
+
"learning_rate": 9.96210254835968e-05,
|
| 283 |
+
"loss": 0.3606,
|
| 284 |
+
"num_input_tokens_seen": 271808,
|
| 285 |
+
"step": 155
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 5.714285714285714,
|
| 289 |
+
"grad_norm": 0.10893812775611877,
|
| 290 |
+
"learning_rate": 9.947385183623098e-05,
|
| 291 |
+
"loss": 0.3507,
|
| 292 |
+
"num_input_tokens_seen": 280704,
|
| 293 |
+
"step": 160
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 5.892857142857143,
|
| 297 |
+
"grad_norm": 0.15622828900814056,
|
| 298 |
+
"learning_rate": 9.930273392233624e-05,
|
| 299 |
+
"loss": 0.3578,
|
| 300 |
+
"num_input_tokens_seen": 289600,
|
| 301 |
+
"step": 165
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 6.0,
|
| 305 |
+
"eval_loss": 0.37425312399864197,
|
| 306 |
+
"eval_runtime": 1.2693,
|
| 307 |
+
"eval_samples_per_second": 87.447,
|
| 308 |
+
"eval_steps_per_second": 11.029,
|
| 309 |
+
"num_input_tokens_seen": 294256,
|
| 310 |
+
"step": 168
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.071428571428571,
|
| 314 |
+
"grad_norm": 0.267703652381897,
|
| 315 |
+
"learning_rate": 9.910775455925518e-05,
|
| 316 |
+
"loss": 0.3546,
|
| 317 |
+
"num_input_tokens_seen": 297520,
|
| 318 |
+
"step": 170
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 6.25,
|
| 322 |
+
"grad_norm": 0.12959004938602448,
|
| 323 |
+
"learning_rate": 9.888900811275204e-05,
|
| 324 |
+
"loss": 0.3507,
|
| 325 |
+
"num_input_tokens_seen": 306672,
|
| 326 |
+
"step": 175
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 6.428571428571429,
|
| 330 |
+
"grad_norm": 0.25042709708213806,
|
| 331 |
+
"learning_rate": 9.864660045134165e-05,
|
| 332 |
+
"loss": 0.3929,
|
| 333 |
+
"num_input_tokens_seen": 315824,
|
| 334 |
+
"step": 180
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 6.607142857142857,
|
| 338 |
+
"grad_norm": 0.1262078881263733,
|
| 339 |
+
"learning_rate": 9.838064889505141e-05,
|
| 340 |
+
"loss": 0.35,
|
| 341 |
+
"num_input_tokens_seen": 324464,
|
| 342 |
+
"step": 185
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 6.785714285714286,
|
| 346 |
+
"grad_norm": 0.2343859225511551,
|
| 347 |
+
"learning_rate": 9.809128215864097e-05,
|
| 348 |
+
"loss": 0.36,
|
| 349 |
+
"num_input_tokens_seen": 332976,
|
| 350 |
+
"step": 190
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.964285714285714,
|
| 354 |
+
"grad_norm": 0.2336055040359497,
|
| 355 |
+
"learning_rate": 9.777864028930705e-05,
|
| 356 |
+
"loss": 0.3521,
|
| 357 |
+
"num_input_tokens_seen": 341552,
|
| 358 |
+
"step": 195
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 7.142857142857143,
|
| 362 |
+
"grad_norm": 0.34513550996780396,
|
| 363 |
+
"learning_rate": 9.744287459890368e-05,
|
| 364 |
+
"loss": 0.352,
|
| 365 |
+
"num_input_tokens_seen": 349584,
|
| 366 |
+
"step": 200
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 7.321428571428571,
|
| 370 |
+
"grad_norm": 0.5254360437393188,
|
| 371 |
+
"learning_rate": 9.708414759071059e-05,
|
| 372 |
+
"loss": 0.3669,
|
| 373 |
+
"num_input_tokens_seen": 358672,
|
| 374 |
+
"step": 205
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 7.5,
|
| 378 |
+
"grad_norm": 0.04271647334098816,
|
| 379 |
+
"learning_rate": 9.670263288078502e-05,
|
| 380 |
+
"loss": 0.3492,
|
| 381 |
+
"num_input_tokens_seen": 368144,
|
| 382 |
+
"step": 210
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 7.5,
|
| 386 |
+
"eval_loss": 0.34583956003189087,
|
| 387 |
+
"eval_runtime": 1.1811,
|
| 388 |
+
"eval_samples_per_second": 93.983,
|
| 389 |
+
"eval_steps_per_second": 11.854,
|
| 390 |
+
"num_input_tokens_seen": 368144,
|
| 391 |
+
"step": 210
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 7.678571428571429,
|
| 395 |
+
"grad_norm": 0.2870471179485321,
|
| 396 |
+
"learning_rate": 9.629851511393555e-05,
|
| 397 |
+
"loss": 0.3575,
|
| 398 |
+
"num_input_tokens_seen": 376464,
|
| 399 |
+
"step": 215
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 7.857142857142857,
|
| 403 |
+
"grad_norm": 0.16233104467391968,
|
| 404 |
+
"learning_rate": 9.587198987435782e-05,
|
| 405 |
+
"loss": 0.347,
|
| 406 |
+
"num_input_tokens_seen": 385616,
|
| 407 |
+
"step": 220
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 8.035714285714286,
|
| 411 |
+
"grad_norm": 0.24467554688453674,
|
| 412 |
+
"learning_rate": 9.542326359097619e-05,
|
| 413 |
+
"loss": 0.3491,
|
| 414 |
+
"num_input_tokens_seen": 393384,
|
| 415 |
+
"step": 225
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 8.214285714285714,
|
| 419 |
+
"grad_norm": 0.2640959918498993,
|
| 420 |
+
"learning_rate": 9.495255343753657e-05,
|
| 421 |
+
"loss": 0.3562,
|
| 422 |
+
"num_input_tokens_seen": 402280,
|
| 423 |
+
"step": 230
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 8.392857142857142,
|
| 427 |
+
"grad_norm": 0.18736229836940765,
|
| 428 |
+
"learning_rate": 9.446008722749905e-05,
|
| 429 |
+
"loss": 0.3491,
|
| 430 |
+
"num_input_tokens_seen": 410280,
|
| 431 |
+
"step": 235
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 8.571428571428571,
|
| 435 |
+
"grad_norm": 0.039574023336172104,
|
| 436 |
+
"learning_rate": 9.394610330378124e-05,
|
| 437 |
+
"loss": 0.3404,
|
| 438 |
+
"num_input_tokens_seen": 418856,
|
| 439 |
+
"step": 240
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 8.75,
|
| 443 |
+
"grad_norm": 0.12487329542636871,
|
| 444 |
+
"learning_rate": 9.341085042340532e-05,
|
| 445 |
+
"loss": 0.3528,
|
| 446 |
+
"num_input_tokens_seen": 427816,
|
| 447 |
+
"step": 245
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 8.928571428571429,
|
| 451 |
+
"grad_norm": 0.08902257680892944,
|
| 452 |
+
"learning_rate": 9.285458763710524e-05,
|
| 453 |
+
"loss": 0.3542,
|
| 454 |
+
"num_input_tokens_seen": 437352,
|
| 455 |
+
"step": 250
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 9.0,
|
| 459 |
+
"eval_loss": 0.34945669770240784,
|
| 460 |
+
"eval_runtime": 1.1316,
|
| 461 |
+
"eval_samples_per_second": 98.092,
|
| 462 |
+
"eval_steps_per_second": 12.372,
|
| 463 |
+
"num_input_tokens_seen": 439768,
|
| 464 |
+
"step": 252
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 9.107142857142858,
|
| 468 |
+
"grad_norm": 0.12130332738161087,
|
| 469 |
+
"learning_rate": 9.227758416395169e-05,
|
| 470 |
+
"loss": 0.3435,
|
| 471 |
+
"num_input_tokens_seen": 444504,
|
| 472 |
+
"step": 255
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 9.285714285714286,
|
| 476 |
+
"grad_norm": 0.14609932899475098,
|
| 477 |
+
"learning_rate": 9.168011926105598e-05,
|
| 478 |
+
"loss": 0.3565,
|
| 479 |
+
"num_input_tokens_seen": 453144,
|
| 480 |
+
"step": 260
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 9.464285714285714,
|
| 484 |
+
"grad_norm": 0.025746047496795654,
|
| 485 |
+
"learning_rate": 9.106248208841569e-05,
|
| 486 |
+
"loss": 0.3499,
|
| 487 |
+
"num_input_tokens_seen": 462040,
|
| 488 |
+
"step": 265
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 9.642857142857142,
|
| 492 |
+
"grad_norm": 0.10566066205501556,
|
| 493 |
+
"learning_rate": 9.042497156896748e-05,
|
| 494 |
+
"loss": 0.349,
|
| 495 |
+
"num_input_tokens_seen": 471576,
|
| 496 |
+
"step": 270
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.821428571428571,
|
| 500 |
+
"grad_norm": 0.2756125032901764,
|
| 501 |
+
"learning_rate": 8.976789624391498e-05,
|
| 502 |
+
"loss": 0.3532,
|
| 503 |
+
"num_input_tokens_seen": 480280,
|
| 504 |
+
"step": 275
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 10.0,
|
| 508 |
+
"grad_norm": 0.05616496875882149,
|
| 509 |
+
"learning_rate": 8.90915741234015e-05,
|
| 510 |
+
"loss": 0.3506,
|
| 511 |
+
"num_input_tokens_seen": 489096,
|
| 512 |
+
"step": 280
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 10.178571428571429,
|
| 516 |
+
"grad_norm": 0.23147229850292206,
|
| 517 |
+
"learning_rate": 8.839633253260006e-05,
|
| 518 |
+
"loss": 0.3475,
|
| 519 |
+
"num_input_tokens_seen": 498952,
|
| 520 |
+
"step": 285
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 10.357142857142858,
|
| 524 |
+
"grad_norm": 0.17449620366096497,
|
| 525 |
+
"learning_rate": 8.768250795329518e-05,
|
| 526 |
+
"loss": 0.34,
|
| 527 |
+
"num_input_tokens_seen": 507976,
|
| 528 |
+
"step": 290
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 10.5,
|
| 532 |
+
"eval_loss": 0.35035043954849243,
|
| 533 |
+
"eval_runtime": 1.1011,
|
| 534 |
+
"eval_samples_per_second": 100.804,
|
| 535 |
+
"eval_steps_per_second": 12.714,
|
| 536 |
+
"num_input_tokens_seen": 514888,
|
| 537 |
+
"step": 294
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 10.535714285714286,
|
| 541 |
+
"grad_norm": 0.09387495368719101,
|
| 542 |
+
"learning_rate": 8.695044586103296e-05,
|
| 543 |
+
"loss": 0.355,
|
| 544 |
+
"num_input_tokens_seen": 516744,
|
| 545 |
+
"step": 295
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 10.714285714285714,
|
| 549 |
+
"grad_norm": 0.269589364528656,
|
| 550 |
+
"learning_rate": 8.620050055791851e-05,
|
| 551 |
+
"loss": 0.3472,
|
| 552 |
+
"num_input_tokens_seen": 525960,
|
| 553 |
+
"step": 300
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 10.892857142857142,
|
| 557 |
+
"grad_norm": 0.045772165060043335,
|
| 558 |
+
"learning_rate": 8.543303500114141e-05,
|
| 559 |
+
"loss": 0.3496,
|
| 560 |
+
"num_input_tokens_seen": 534344,
|
| 561 |
+
"step": 305
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 11.071428571428571,
|
| 565 |
+
"grad_norm": 0.23182272911071777,
|
| 566 |
+
"learning_rate": 8.464842062731235e-05,
|
| 567 |
+
"loss": 0.3481,
|
| 568 |
+
"num_input_tokens_seen": 541856,
|
| 569 |
+
"step": 310
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 11.25,
|
| 573 |
+
"grad_norm": 0.34456300735473633,
|
| 574 |
+
"learning_rate": 8.384703717269584e-05,
|
| 575 |
+
"loss": 0.3448,
|
| 576 |
+
"num_input_tokens_seen": 550176,
|
| 577 |
+
"step": 315
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 11.428571428571429,
|
| 581 |
+
"grad_norm": 0.04894077032804489,
|
| 582 |
+
"learning_rate": 8.302927248942627e-05,
|
| 583 |
+
"loss": 0.3405,
|
| 584 |
+
"num_input_tokens_seen": 558368,
|
| 585 |
+
"step": 320
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 11.607142857142858,
|
| 589 |
+
"grad_norm": 0.33567559719085693,
|
| 590 |
+
"learning_rate": 8.219552235779578e-05,
|
| 591 |
+
"loss": 0.3466,
|
| 592 |
+
"num_input_tokens_seen": 567392,
|
| 593 |
+
"step": 325
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 11.785714285714286,
|
| 597 |
+
"grad_norm": 0.2454521507024765,
|
| 598 |
+
"learning_rate": 8.134619029470534e-05,
|
| 599 |
+
"loss": 0.3594,
|
| 600 |
+
"num_input_tokens_seen": 576864,
|
| 601 |
+
"step": 330
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 11.964285714285714,
|
| 605 |
+
"grad_norm": 0.41574931144714355,
|
| 606 |
+
"learning_rate": 8.048168735837121e-05,
|
| 607 |
+
"loss": 0.3578,
|
| 608 |
+
"num_input_tokens_seen": 585376,
|
| 609 |
+
"step": 335
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 12.0,
|
| 613 |
+
"eval_loss": 0.3650008738040924,
|
| 614 |
+
"eval_runtime": 1.1362,
|
| 615 |
+
"eval_samples_per_second": 97.697,
|
| 616 |
+
"eval_steps_per_second": 12.322,
|
| 617 |
+
"num_input_tokens_seen": 586448,
|
| 618 |
+
"step": 336
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 12.142857142857142,
|
| 622 |
+
"grad_norm": 0.13403870165348053,
|
| 623 |
+
"learning_rate": 7.960243194938192e-05,
|
| 624 |
+
"loss": 0.3703,
|
| 625 |
+
"num_input_tokens_seen": 593296,
|
| 626 |
+
"step": 340
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 12.321428571428571,
|
| 630 |
+
"grad_norm": 0.03592513129115105,
|
| 631 |
+
"learning_rate": 7.87088496082013e-05,
|
| 632 |
+
"loss": 0.3383,
|
| 633 |
+
"num_input_tokens_seen": 602000,
|
| 634 |
+
"step": 345
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 12.5,
|
| 638 |
+
"grad_norm": 0.30938151478767395,
|
| 639 |
+
"learning_rate": 7.780137280921636e-05,
|
| 640 |
+
"loss": 0.3849,
|
| 641 |
+
"num_input_tokens_seen": 610960,
|
| 642 |
+
"step": 350
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 12.678571428571429,
|
| 646 |
+
"grad_norm": 0.047836095094680786,
|
| 647 |
+
"learning_rate": 7.688044075142887e-05,
|
| 648 |
+
"loss": 0.3439,
|
| 649 |
+
"num_input_tokens_seen": 619984,
|
| 650 |
+
"step": 355
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 12.857142857142858,
|
| 654 |
+
"grad_norm": 0.08763613551855087,
|
| 655 |
+
"learning_rate": 7.594649914589287e-05,
|
| 656 |
+
"loss": 0.3448,
|
| 657 |
+
"num_input_tokens_seen": 629776,
|
| 658 |
+
"step": 360
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 13.035714285714286,
|
| 662 |
+
"grad_norm": 0.0415462851524353,
|
| 663 |
+
"learning_rate": 7.500000000000001e-05,
|
| 664 |
+
"loss": 0.348,
|
| 665 |
+
"num_input_tokens_seen": 637888,
|
| 666 |
+
"step": 365
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 13.214285714285714,
|
| 670 |
+
"grad_norm": 0.12538637220859528,
|
| 671 |
+
"learning_rate": 7.404140139871797e-05,
|
| 672 |
+
"loss": 0.345,
|
| 673 |
+
"num_input_tokens_seen": 648128,
|
| 674 |
+
"step": 370
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 13.392857142857142,
|
| 678 |
+
"grad_norm": 0.17965561151504517,
|
| 679 |
+
"learning_rate": 7.307116728288727e-05,
|
| 680 |
+
"loss": 0.3462,
|
| 681 |
+
"num_input_tokens_seen": 656768,
|
| 682 |
+
"step": 375
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 13.5,
|
| 686 |
+
"eval_loss": 0.3538319170475006,
|
| 687 |
+
"eval_runtime": 1.1141,
|
| 688 |
+
"eval_samples_per_second": 99.628,
|
| 689 |
+
"eval_steps_per_second": 12.566,
|
| 690 |
+
"num_input_tokens_seen": 662016,
|
| 691 |
+
"step": 378
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 13.571428571428571,
|
| 695 |
+
"grad_norm": 0.22999395430088043,
|
| 696 |
+
"learning_rate": 7.208976722468392e-05,
|
| 697 |
+
"loss": 0.3606,
|
| 698 |
+
"num_input_tokens_seen": 665472,
|
| 699 |
+
"step": 380
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 13.75,
|
| 703 |
+
"grad_norm": 0.0893191397190094,
|
| 704 |
+
"learning_rate": 7.109767620035689e-05,
|
| 705 |
+
"loss": 0.3501,
|
| 706 |
+
"num_input_tokens_seen": 673664,
|
| 707 |
+
"step": 385
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 13.928571428571429,
|
| 711 |
+
"grad_norm": 0.09730294346809387,
|
| 712 |
+
"learning_rate": 7.00953743603498e-05,
|
| 713 |
+
"loss": 0.3461,
|
| 714 |
+
"num_input_tokens_seen": 682688,
|
| 715 |
+
"step": 390
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 14.107142857142858,
|
| 719 |
+
"grad_norm": 0.05350238084793091,
|
| 720 |
+
"learning_rate": 6.908334679691863e-05,
|
| 721 |
+
"loss": 0.3522,
|
| 722 |
+
"num_input_tokens_seen": 690936,
|
| 723 |
+
"step": 395
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 14.285714285714286,
|
| 727 |
+
"grad_norm": 0.04386021941900253,
|
| 728 |
+
"learning_rate": 6.806208330935766e-05,
|
| 729 |
+
"loss": 0.3553,
|
| 730 |
+
"num_input_tokens_seen": 700536,
|
| 731 |
+
"step": 400
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 14.464285714285714,
|
| 735 |
+
"grad_norm": 0.0879250317811966,
|
| 736 |
+
"learning_rate": 6.703207816694719e-05,
|
| 737 |
+
"loss": 0.3382,
|
| 738 |
+
"num_input_tokens_seen": 709560,
|
| 739 |
+
"step": 405
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 14.642857142857142,
|
| 743 |
+
"grad_norm": 0.31405237317085266,
|
| 744 |
+
"learning_rate": 6.599382986973808e-05,
|
| 745 |
+
"loss": 0.348,
|
| 746 |
+
"num_input_tokens_seen": 718584,
|
| 747 |
+
"step": 410
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 14.821428571428571,
|
| 751 |
+
"grad_norm": 0.3567558228969574,
|
| 752 |
+
"learning_rate": 6.494784090728852e-05,
|
| 753 |
+
"loss": 0.3544,
|
| 754 |
+
"num_input_tokens_seen": 727416,
|
| 755 |
+
"step": 415
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 15.0,
|
| 759 |
+
"grad_norm": 0.061139896512031555,
|
| 760 |
+
"learning_rate": 6.389461751547008e-05,
|
| 761 |
+
"loss": 0.3506,
|
| 762 |
+
"num_input_tokens_seen": 735680,
|
| 763 |
+
"step": 420
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 15.0,
|
| 767 |
+
"eval_loss": 0.35569673776626587,
|
| 768 |
+
"eval_runtime": 1.1923,
|
| 769 |
+
"eval_samples_per_second": 93.099,
|
| 770 |
+
"eval_steps_per_second": 11.742,
|
| 771 |
+
"num_input_tokens_seen": 735680,
|
| 772 |
+
"step": 420
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 15.178571428571429,
|
| 776 |
+
"grad_norm": 0.19610817730426788,
|
| 777 |
+
"learning_rate": 6.283466943146053e-05,
|
| 778 |
+
"loss": 0.3486,
|
| 779 |
+
"num_input_tokens_seen": 744832,
|
| 780 |
+
"step": 425
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 15.357142857142858,
|
| 784 |
+
"grad_norm": 0.05017664283514023,
|
| 785 |
+
"learning_rate": 6.176850964704213e-05,
|
| 786 |
+
"loss": 0.3573,
|
| 787 |
+
"num_input_tokens_seen": 753728,
|
| 788 |
+
"step": 430
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 15.535714285714286,
|
| 792 |
+
"grad_norm": 0.047355543822050095,
|
| 793 |
+
"learning_rate": 6.069665416032487e-05,
|
| 794 |
+
"loss": 0.3564,
|
| 795 |
+
"num_input_tokens_seen": 762752,
|
| 796 |
+
"step": 435
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 15.714285714285714,
|
| 800 |
+
"grad_norm": 0.16390874981880188,
|
| 801 |
+
"learning_rate": 5.961962172601458e-05,
|
| 802 |
+
"loss": 0.3461,
|
| 803 |
+
"num_input_tokens_seen": 771200,
|
| 804 |
+
"step": 440
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 15.892857142857142,
|
| 808 |
+
"grad_norm": 0.19481535255908966,
|
| 809 |
+
"learning_rate": 5.853793360434687e-05,
|
| 810 |
+
"loss": 0.3564,
|
| 811 |
+
"num_input_tokens_seen": 779776,
|
| 812 |
+
"step": 445
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 16.071428571428573,
|
| 816 |
+
"grad_norm": 0.04299188032746315,
|
| 817 |
+
"learning_rate": 5.745211330880872e-05,
|
| 818 |
+
"loss": 0.3214,
|
| 819 |
+
"num_input_tokens_seen": 788984,
|
| 820 |
+
"step": 450
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 16.25,
|
| 824 |
+
"grad_norm": 0.07021050900220871,
|
| 825 |
+
"learning_rate": 5.636268635276918e-05,
|
| 826 |
+
"loss": 0.355,
|
| 827 |
+
"num_input_tokens_seen": 798456,
|
| 828 |
+
"step": 455
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 16.428571428571427,
|
| 832 |
+
"grad_norm": 0.2501037120819092,
|
| 833 |
+
"learning_rate": 5.527017999514239e-05,
|
| 834 |
+
"loss": 0.3489,
|
| 835 |
+
"num_input_tokens_seen": 807352,
|
| 836 |
+
"step": 460
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 16.5,
|
| 840 |
+
"eval_loss": 0.35193607211112976,
|
| 841 |
+
"eval_runtime": 1.1402,
|
| 842 |
+
"eval_samples_per_second": 97.356,
|
| 843 |
+
"eval_steps_per_second": 12.279,
|
| 844 |
+
"num_input_tokens_seen": 810232,
|
| 845 |
+
"step": 462
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 16.607142857142858,
|
| 849 |
+
"grad_norm": 0.08860552310943604,
|
| 850 |
+
"learning_rate": 5.417512298520585e-05,
|
| 851 |
+
"loss": 0.3506,
|
| 852 |
+
"num_input_tokens_seen": 814712,
|
| 853 |
+
"step": 465
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 16.785714285714285,
|
| 857 |
+
"grad_norm": 0.08465074002742767,
|
| 858 |
+
"learning_rate": 5.307804530669716e-05,
|
| 859 |
+
"loss": 0.346,
|
| 860 |
+
"num_input_tokens_seen": 823608,
|
| 861 |
+
"step": 470
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 16.964285714285715,
|
| 865 |
+
"grad_norm": 0.14993856847286224,
|
| 866 |
+
"learning_rate": 5.197947792131348e-05,
|
| 867 |
+
"loss": 0.3467,
|
| 868 |
+
"num_input_tokens_seen": 832888,
|
| 869 |
+
"step": 475
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 17.142857142857142,
|
| 873 |
+
"grad_norm": 0.17248603701591492,
|
| 874 |
+
"learning_rate": 5.0879952511737696e-05,
|
| 875 |
+
"loss": 0.3452,
|
| 876 |
+
"num_input_tokens_seen": 840464,
|
| 877 |
+
"step": 480
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 17.321428571428573,
|
| 881 |
+
"grad_norm": 0.10642991960048676,
|
| 882 |
+
"learning_rate": 4.97800012243155e-05,
|
| 883 |
+
"loss": 0.3402,
|
| 884 |
+
"num_input_tokens_seen": 849488,
|
| 885 |
+
"step": 485
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 17.5,
|
| 889 |
+
"grad_norm": 0.07554838806390762,
|
| 890 |
+
"learning_rate": 4.86801564115082e-05,
|
| 891 |
+
"loss": 0.3428,
|
| 892 |
+
"num_input_tokens_seen": 858448,
|
| 893 |
+
"step": 490
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 17.678571428571427,
|
| 897 |
+
"grad_norm": 0.1515018343925476,
|
| 898 |
+
"learning_rate": 4.758095037424567e-05,
|
| 899 |
+
"loss": 0.352,
|
| 900 |
+
"num_input_tokens_seen": 867280,
|
| 901 |
+
"step": 495
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 17.857142857142858,
|
| 905 |
+
"grad_norm": 0.2787249684333801,
|
| 906 |
+
"learning_rate": 4.648291510430438e-05,
|
| 907 |
+
"loss": 0.3528,
|
| 908 |
+
"num_input_tokens_seen": 876880,
|
| 909 |
+
"step": 500
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 18.0,
|
| 913 |
+
"eval_loss": 0.3557737469673157,
|
| 914 |
+
"eval_runtime": 1.1986,
|
| 915 |
+
"eval_samples_per_second": 92.609,
|
| 916 |
+
"eval_steps_per_second": 11.68,
|
| 917 |
+
"num_input_tokens_seen": 882920,
|
| 918 |
+
"step": 504
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 18.035714285714285,
|
| 922 |
+
"grad_norm": 0.21712534129619598,
|
| 923 |
+
"learning_rate": 4.5386582026834906e-05,
|
| 924 |
+
"loss": 0.3418,
|
| 925 |
+
"num_input_tokens_seen": 885480,
|
| 926 |
+
"step": 505
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 18.214285714285715,
|
| 930 |
+
"grad_norm": 0.05514955148100853,
|
| 931 |
+
"learning_rate": 4.4292481743163755e-05,
|
| 932 |
+
"loss": 0.3441,
|
| 933 |
+
"num_input_tokens_seen": 893864,
|
| 934 |
+
"step": 510
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 18.392857142857142,
|
| 938 |
+
"grad_norm": 0.06789080053567886,
|
| 939 |
+
"learning_rate": 4.3201143773993865e-05,
|
| 940 |
+
"loss": 0.3524,
|
| 941 |
+
"num_input_tokens_seen": 901928,
|
| 942 |
+
"step": 515
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 18.571428571428573,
|
| 946 |
+
"grad_norm": 0.11630789190530777,
|
| 947 |
+
"learning_rate": 4.2113096303128125e-05,
|
| 948 |
+
"loss": 0.3495,
|
| 949 |
+
"num_input_tokens_seen": 910696,
|
| 950 |
+
"step": 520
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 18.75,
|
| 954 |
+
"grad_norm": 0.13940396904945374,
|
| 955 |
+
"learning_rate": 4.102886592183996e-05,
|
| 956 |
+
"loss": 0.3378,
|
| 957 |
+
"num_input_tokens_seen": 920104,
|
| 958 |
+
"step": 525
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 18.928571428571427,
|
| 962 |
+
"grad_norm": 0.051015470176935196,
|
| 963 |
+
"learning_rate": 3.9948977374014544e-05,
|
| 964 |
+
"loss": 0.3511,
|
| 965 |
+
"num_input_tokens_seen": 928936,
|
| 966 |
+
"step": 530
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 19.107142857142858,
|
| 970 |
+
"grad_norm": 0.09734626859426498,
|
| 971 |
+
"learning_rate": 3.887395330218429e-05,
|
| 972 |
+
"loss": 0.3465,
|
| 973 |
+
"num_input_tokens_seen": 937840,
|
| 974 |
+
"step": 535
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 19.285714285714285,
|
| 978 |
+
"grad_norm": 0.11513973772525787,
|
| 979 |
+
"learning_rate": 3.780431399458114e-05,
|
| 980 |
+
"loss": 0.3478,
|
| 981 |
+
"num_input_tokens_seen": 947248,
|
| 982 |
+
"step": 540
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 19.464285714285715,
|
| 986 |
+
"grad_norm": 0.11863286793231964,
|
| 987 |
+
"learning_rate": 3.6740577133328524e-05,
|
| 988 |
+
"loss": 0.3408,
|
| 989 |
+
"num_input_tokens_seen": 955568,
|
| 990 |
+
"step": 545
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 19.5,
|
| 994 |
+
"eval_loss": 0.3516767919063568,
|
| 995 |
+
"eval_runtime": 1.1222,
|
| 996 |
+
"eval_samples_per_second": 98.916,
|
| 997 |
+
"eval_steps_per_second": 12.476,
|
| 998 |
+
"num_input_tokens_seen": 957488,
|
| 999 |
+
"step": 546
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 19.642857142857142,
|
| 1003 |
+
"grad_norm": 0.04675190523266792,
|
| 1004 |
+
"learning_rate": 3.568325754389438e-05,
|
| 1005 |
+
"loss": 0.3459,
|
| 1006 |
+
"num_input_tokens_seen": 964400,
|
| 1007 |
+
"step": 550
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 19.821428571428573,
|
| 1011 |
+
"grad_norm": 0.13177751004695892,
|
| 1012 |
+
"learning_rate": 3.4632866945926855e-05,
|
| 1013 |
+
"loss": 0.3473,
|
| 1014 |
+
"num_input_tokens_seen": 972720,
|
| 1015 |
+
"step": 555
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 20.0,
|
| 1019 |
+
"grad_norm": 0.05624629184603691,
|
| 1020 |
+
"learning_rate": 3.3589913705593235e-05,
|
| 1021 |
+
"loss": 0.3508,
|
| 1022 |
+
"num_input_tokens_seen": 980760,
|
| 1023 |
+
"step": 560
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 20.178571428571427,
|
| 1027 |
+
"grad_norm": 0.04487062245607376,
|
| 1028 |
+
"learning_rate": 3.255490258954167e-05,
|
| 1029 |
+
"loss": 0.3414,
|
| 1030 |
+
"num_input_tokens_seen": 989464,
|
| 1031 |
+
"step": 565
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 20.357142857142858,
|
| 1035 |
+
"grad_norm": 0.3037759065628052,
|
| 1036 |
+
"learning_rate": 3.152833452060522e-05,
|
| 1037 |
+
"loss": 0.3518,
|
| 1038 |
+
"num_input_tokens_seen": 998680,
|
| 1039 |
+
"step": 570
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 20.535714285714285,
|
| 1043 |
+
"grad_norm": 0.34507808089256287,
|
| 1044 |
+
"learning_rate": 3.0510706335366035e-05,
|
| 1045 |
+
"loss": 0.3479,
|
| 1046 |
+
"num_input_tokens_seen": 1007768,
|
| 1047 |
+
"step": 575
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 20.714285714285715,
|
| 1051 |
+
"grad_norm": 0.32024624943733215,
|
| 1052 |
+
"learning_rate": 2.9502510543697325e-05,
|
| 1053 |
+
"loss": 0.3549,
|
| 1054 |
+
"num_input_tokens_seen": 1016536,
|
| 1055 |
+
"step": 580
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 20.892857142857142,
|
| 1059 |
+
"grad_norm": 0.0999528020620346,
|
| 1060 |
+
"learning_rate": 2.850423509039928e-05,
|
| 1061 |
+
"loss": 0.3469,
|
| 1062 |
+
"num_input_tokens_seen": 1025048,
|
| 1063 |
+
"step": 585
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 21.0,
|
| 1067 |
+
"eval_loss": 0.35418227314949036,
|
| 1068 |
+
"eval_runtime": 1.282,
|
| 1069 |
+
"eval_samples_per_second": 86.583,
|
| 1070 |
+
"eval_steps_per_second": 10.92,
|
| 1071 |
+
"num_input_tokens_seen": 1029792,
|
| 1072 |
+
"step": 588
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 21.071428571428573,
|
| 1076 |
+
"grad_norm": 0.10501652210950851,
|
| 1077 |
+
"learning_rate": 2.751636311904444e-05,
|
| 1078 |
+
"loss": 0.343,
|
| 1079 |
+
"num_input_tokens_seen": 1032864,
|
| 1080 |
+
"step": 590
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 21.25,
|
| 1084 |
+
"grad_norm": 0.04894215986132622,
|
| 1085 |
+
"learning_rate": 2.6539372738146695e-05,
|
| 1086 |
+
"loss": 0.3426,
|
| 1087 |
+
"num_input_tokens_seen": 1041248,
|
| 1088 |
+
"step": 595
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 21.428571428571427,
|
| 1092 |
+
"grad_norm": 0.11162778735160828,
|
| 1093 |
+
"learning_rate": 2.5573736789767232e-05,
|
| 1094 |
+
"loss": 0.3358,
|
| 1095 |
+
"num_input_tokens_seen": 1050720,
|
| 1096 |
+
"step": 600
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 21.607142857142858,
|
| 1100 |
+
"grad_norm": 0.11263741552829742,
|
| 1101 |
+
"learning_rate": 2.4619922620669218e-05,
|
| 1102 |
+
"loss": 0.3549,
|
| 1103 |
+
"num_input_tokens_seen": 1060064,
|
| 1104 |
+
"step": 605
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 21.785714285714285,
|
| 1108 |
+
"grad_norm": 0.2273244857788086,
|
| 1109 |
+
"learning_rate": 2.3678391856132204e-05,
|
| 1110 |
+
"loss": 0.3524,
|
| 1111 |
+
"num_input_tokens_seen": 1068256,
|
| 1112 |
+
"step": 610
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 21.964285714285715,
|
| 1116 |
+
"grad_norm": 0.23367320001125336,
|
| 1117 |
+
"learning_rate": 2.2749600176535534e-05,
|
| 1118 |
+
"loss": 0.3474,
|
| 1119 |
+
"num_input_tokens_seen": 1077024,
|
| 1120 |
+
"step": 615
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 22.142857142857142,
|
| 1124 |
+
"grad_norm": 0.17969422042369843,
|
| 1125 |
+
"learning_rate": 2.1833997096818898e-05,
|
| 1126 |
+
"loss": 0.3425,
|
| 1127 |
+
"num_input_tokens_seen": 1086328,
|
| 1128 |
+
"step": 620
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 22.321428571428573,
|
| 1132 |
+
"grad_norm": 0.20163026452064514,
|
| 1133 |
+
"learning_rate": 2.0932025748927013e-05,
|
| 1134 |
+
"loss": 0.3438,
|
| 1135 |
+
"num_input_tokens_seen": 1094520,
|
| 1136 |
+
"step": 625
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 22.5,
|
| 1140 |
+
"grad_norm": 0.23486794531345367,
|
| 1141 |
+
"learning_rate": 2.0044122667343297e-05,
|
| 1142 |
+
"loss": 0.3488,
|
| 1143 |
+
"num_input_tokens_seen": 1103160,
|
| 1144 |
+
"step": 630
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 22.5,
|
| 1148 |
+
"eval_loss": 0.35537073016166687,
|
| 1149 |
+
"eval_runtime": 1.1326,
|
| 1150 |
+
"eval_samples_per_second": 98.004,
|
| 1151 |
+
"eval_steps_per_second": 12.361,
|
| 1152 |
+
"num_input_tokens_seen": 1103160,
|
| 1153 |
+
"step": 630
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 22.678571428571427,
|
| 1157 |
+
"grad_norm": 0.04255451634526253,
|
| 1158 |
+
"learning_rate": 1.917071757781679e-05,
|
| 1159 |
+
"loss": 0.3465,
|
| 1160 |
+
"num_input_tokens_seen": 1112376,
|
| 1161 |
+
"step": 635
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 22.857142857142858,
|
| 1165 |
+
"grad_norm": 0.07782665640115738,
|
| 1166 |
+
"learning_rate": 1.831223318938419e-05,
|
| 1167 |
+
"loss": 0.3449,
|
| 1168 |
+
"num_input_tokens_seen": 1121464,
|
| 1169 |
+
"step": 640
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 23.035714285714285,
|
| 1173 |
+
"grad_norm": 0.11463826894760132,
|
| 1174 |
+
"learning_rate": 1.746908498978791e-05,
|
| 1175 |
+
"loss": 0.3526,
|
| 1176 |
+
"num_input_tokens_seen": 1130152,
|
| 1177 |
+
"step": 645
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 23.214285714285715,
|
| 1181 |
+
"grad_norm": 0.07137856632471085,
|
| 1182 |
+
"learning_rate": 1.6641681044389014e-05,
|
| 1183 |
+
"loss": 0.3495,
|
| 1184 |
+
"num_input_tokens_seen": 1138664,
|
| 1185 |
+
"step": 650
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 23.392857142857142,
|
| 1189 |
+
"grad_norm": 0.08257535845041275,
|
| 1190 |
+
"learning_rate": 1.5830421798672568e-05,
|
| 1191 |
+
"loss": 0.3447,
|
| 1192 |
+
"num_input_tokens_seen": 1146216,
|
| 1193 |
+
"step": 655
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 23.571428571428573,
|
| 1197 |
+
"grad_norm": 0.0766848772764206,
|
| 1198 |
+
"learning_rate": 1.5035699884440697e-05,
|
| 1199 |
+
"loss": 0.3509,
|
| 1200 |
+
"num_input_tokens_seen": 1155496,
|
| 1201 |
+
"step": 660
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 23.75,
|
| 1205 |
+
"grad_norm": 0.08535895496606827,
|
| 1206 |
+
"learning_rate": 1.4257899929787294e-05,
|
| 1207 |
+
"loss": 0.3515,
|
| 1208 |
+
"num_input_tokens_seen": 1164840,
|
| 1209 |
+
"step": 665
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 23.928571428571427,
|
| 1213 |
+
"grad_norm": 0.09435199946165085,
|
| 1214 |
+
"learning_rate": 1.3497398372946501e-05,
|
| 1215 |
+
"loss": 0.3402,
|
| 1216 |
+
"num_input_tokens_seen": 1173736,
|
| 1217 |
+
"step": 670
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 24.0,
|
| 1221 |
+
"eval_loss": 0.35293668508529663,
|
| 1222 |
+
"eval_runtime": 1.2061,
|
| 1223 |
+
"eval_samples_per_second": 92.034,
|
| 1224 |
+
"eval_steps_per_second": 11.608,
|
| 1225 |
+
"num_input_tokens_seen": 1176968,
|
| 1226 |
+
"step": 672
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 24.107142857142858,
|
| 1230 |
+
"grad_norm": 0.05331406742334366,
|
| 1231 |
+
"learning_rate": 1.2754563280104714e-05,
|
| 1232 |
+
"loss": 0.3404,
|
| 1233 |
+
"num_input_tokens_seen": 1182344,
|
| 1234 |
+
"step": 675
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 24.285714285714285,
|
| 1238 |
+
"grad_norm": 0.14441683888435364,
|
| 1239 |
+
"learning_rate": 1.202975416726464e-05,
|
| 1240 |
+
"loss": 0.357,
|
| 1241 |
+
"num_input_tokens_seen": 1191944,
|
| 1242 |
+
"step": 680
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 24.464285714285715,
|
| 1246 |
+
"grad_norm": 0.06084274500608444,
|
| 1247 |
+
"learning_rate": 1.1323321826247346e-05,
|
| 1248 |
+
"loss": 0.3355,
|
| 1249 |
+
"num_input_tokens_seen": 1200520,
|
| 1250 |
+
"step": 685
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 24.642857142857142,
|
| 1254 |
+
"grad_norm": 0.13699252903461456,
|
| 1255 |
+
"learning_rate": 1.0635608154916648e-05,
|
| 1256 |
+
"loss": 0.3433,
|
| 1257 |
+
"num_input_tokens_seen": 1209288,
|
| 1258 |
+
"step": 690
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 24.821428571428573,
|
| 1262 |
+
"grad_norm": 0.14677266776561737,
|
| 1263 |
+
"learning_rate": 9.966945991708005e-06,
|
| 1264 |
+
"loss": 0.3554,
|
| 1265 |
+
"num_input_tokens_seen": 1217544,
|
| 1266 |
+
"step": 695
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 25.0,
|
| 1270 |
+
"grad_norm": 0.23254753649234772,
|
| 1271 |
+
"learning_rate": 9.317658954541992e-06,
|
| 1272 |
+
"loss": 0.3439,
|
| 1273 |
+
"num_input_tokens_seen": 1224848,
|
| 1274 |
+
"step": 700
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 25.178571428571427,
|
| 1278 |
+
"grad_norm": 0.1202574372291565,
|
| 1279 |
+
"learning_rate": 8.688061284200266e-06,
|
| 1280 |
+
"loss": 0.3371,
|
| 1281 |
+
"num_input_tokens_seen": 1234064,
|
| 1282 |
+
"step": 705
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 25.357142857142858,
|
| 1286 |
+
"grad_norm": 0.055635951459407806,
|
| 1287 |
+
"learning_rate": 8.07845769223981e-06,
|
| 1288 |
+
"loss": 0.3422,
|
| 1289 |
+
"num_input_tokens_seen": 1242448,
|
| 1290 |
+
"step": 710
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 25.5,
|
| 1294 |
+
"eval_loss": 0.35582345724105835,
|
| 1295 |
+
"eval_runtime": 1.0722,
|
| 1296 |
+
"eval_samples_per_second": 103.524,
|
| 1297 |
+
"eval_steps_per_second": 13.057,
|
| 1298 |
+
"num_input_tokens_seen": 1250064,
|
| 1299 |
+
"step": 714
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 25.535714285714285,
|
| 1303 |
+
"grad_norm": 0.29359593987464905,
|
| 1304 |
+
"learning_rate": 7.489143213519301e-06,
|
| 1305 |
+
"loss": 0.352,
|
| 1306 |
+
"num_input_tokens_seen": 1251536,
|
| 1307 |
+
"step": 715
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 25.714285714285715,
|
| 1311 |
+
"grad_norm": 0.08035387843847275,
|
| 1312 |
+
"learning_rate": 6.920403063408526e-06,
|
| 1313 |
+
"loss": 0.3426,
|
| 1314 |
+
"num_input_tokens_seen": 1259792,
|
| 1315 |
+
"step": 720
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 25.892857142857142,
|
| 1319 |
+
"grad_norm": 0.21466954052448273,
|
| 1320 |
+
"learning_rate": 6.372512499750471e-06,
|
| 1321 |
+
"loss": 0.3496,
|
| 1322 |
+
"num_input_tokens_seen": 1268560,
|
| 1323 |
+
"step": 725
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 26.071428571428573,
|
| 1327 |
+
"grad_norm": 0.15136410295963287,
|
| 1328 |
+
"learning_rate": 5.845736689642472e-06,
|
| 1329 |
+
"loss": 0.3447,
|
| 1330 |
+
"num_input_tokens_seen": 1276648,
|
| 1331 |
+
"step": 730
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 26.25,
|
| 1335 |
+
"grad_norm": 0.13561971485614777,
|
| 1336 |
+
"learning_rate": 5.3403305811010885e-06,
|
| 1337 |
+
"loss": 0.3406,
|
| 1338 |
+
"num_input_tokens_seen": 1287144,
|
| 1339 |
+
"step": 735
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 26.428571428571427,
|
| 1343 |
+
"grad_norm": 0.07142584770917892,
|
| 1344 |
+
"learning_rate": 4.8565387796728865e-06,
|
| 1345 |
+
"loss": 0.3451,
|
| 1346 |
+
"num_input_tokens_seen": 1294504,
|
| 1347 |
+
"step": 740
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 26.607142857142858,
|
| 1351 |
+
"grad_norm": 0.2093130499124527,
|
| 1352 |
+
"learning_rate": 4.394595430050613e-06,
|
| 1353 |
+
"loss": 0.3485,
|
| 1354 |
+
"num_input_tokens_seen": 1303336,
|
| 1355 |
+
"step": 745
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 26.785714285714285,
|
| 1359 |
+
"grad_norm": 0.07084821909666061,
|
| 1360 |
+
"learning_rate": 3.954724102752316e-06,
|
| 1361 |
+
"loss": 0.3451,
|
| 1362 |
+
"num_input_tokens_seen": 1312296,
|
| 1363 |
+
"step": 750
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 26.964285714285715,
|
| 1367 |
+
"grad_norm": 0.1166943907737732,
|
| 1368 |
+
"learning_rate": 3.537137685918074e-06,
|
| 1369 |
+
"loss": 0.3474,
|
| 1370 |
+
"num_input_tokens_seen": 1319912,
|
| 1371 |
+
"step": 755
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 27.0,
|
| 1375 |
+
"eval_loss": 0.3538145124912262,
|
| 1376 |
+
"eval_runtime": 1.1107,
|
| 1377 |
+
"eval_samples_per_second": 99.936,
|
| 1378 |
+
"eval_steps_per_second": 12.605,
|
| 1379 |
+
"num_input_tokens_seen": 1321408,
|
| 1380 |
+
"step": 756
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 27.142857142857142,
|
| 1384 |
+
"grad_norm": 0.21413478255271912,
|
| 1385 |
+
"learning_rate": 3.1420382822767323e-06,
|
| 1386 |
+
"loss": 0.3382,
|
| 1387 |
+
"num_input_tokens_seen": 1329024,
|
| 1388 |
+
"step": 760
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 27.321428571428573,
|
| 1392 |
+
"grad_norm": 0.08071761578321457,
|
| 1393 |
+
"learning_rate": 2.7696171113326396e-06,
|
| 1394 |
+
"loss": 0.3531,
|
| 1395 |
+
"num_input_tokens_seen": 1337984,
|
| 1396 |
+
"step": 765
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 27.5,
|
| 1400 |
+
"grad_norm": 0.20051950216293335,
|
| 1401 |
+
"learning_rate": 2.420054416819556e-06,
|
| 1402 |
+
"loss": 0.3338,
|
| 1403 |
+
"num_input_tokens_seen": 1346112,
|
| 1404 |
+
"step": 770
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 27.678571428571427,
|
| 1408 |
+
"grad_norm": 0.05100846663117409,
|
| 1409 |
+
"learning_rate": 2.093519379466602e-06,
|
| 1410 |
+
"loss": 0.342,
|
| 1411 |
+
"num_input_tokens_seen": 1355584,
|
| 1412 |
+
"step": 775
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 27.857142857142858,
|
| 1416 |
+
"grad_norm": 0.11974883079528809,
|
| 1417 |
+
"learning_rate": 1.7901700351184659e-06,
|
| 1418 |
+
"loss": 0.3414,
|
| 1419 |
+
"num_input_tokens_seen": 1364864,
|
| 1420 |
+
"step": 780
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 28.035714285714285,
|
| 1424 |
+
"grad_norm": 0.0698491781949997,
|
| 1425 |
+
"learning_rate": 1.5101531982495308e-06,
|
| 1426 |
+
"loss": 0.3553,
|
| 1427 |
+
"num_input_tokens_seen": 1373400,
|
| 1428 |
+
"step": 785
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 28.214285714285715,
|
| 1432 |
+
"grad_norm": 0.18634414672851562,
|
| 1433 |
+
"learning_rate": 1.2536043909088191e-06,
|
| 1434 |
+
"loss": 0.349,
|
| 1435 |
+
"num_input_tokens_seen": 1382744,
|
| 1436 |
+
"step": 790
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 28.392857142857142,
|
| 1440 |
+
"grad_norm": 0.05743042752146721,
|
| 1441 |
+
"learning_rate": 1.0206477771303236e-06,
|
| 1442 |
+
"loss": 0.3409,
|
| 1443 |
+
"num_input_tokens_seen": 1390424,
|
| 1444 |
+
"step": 795
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 28.5,
|
| 1448 |
+
"eval_loss": 0.35238173604011536,
|
| 1449 |
+
"eval_runtime": 1.1779,
|
| 1450 |
+
"eval_samples_per_second": 94.233,
|
| 1451 |
+
"eval_steps_per_second": 11.885,
|
| 1452 |
+
"num_input_tokens_seen": 1394904,
|
| 1453 |
+
"step": 798
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 28.571428571428573,
|
| 1457 |
+
"grad_norm": 0.0811547264456749,
|
| 1458 |
+
"learning_rate": 8.113961028402894e-07,
|
| 1459 |
+
"loss": 0.3513,
|
| 1460 |
+
"num_input_tokens_seen": 1399256,
|
| 1461 |
+
"step": 800
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 28.75,
|
| 1465 |
+
"grad_norm": 0.06173446401953697,
|
| 1466 |
+
"learning_rate": 6.259506412906402e-07,
|
| 1467 |
+
"loss": 0.35,
|
| 1468 |
+
"num_input_tokens_seen": 1407896,
|
| 1469 |
+
"step": 805
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 28.928571428571427,
|
| 1473 |
+
"grad_norm": 0.05312683433294296,
|
| 1474 |
+
"learning_rate": 4.6440114404492363e-07,
|
| 1475 |
+
"loss": 0.338,
|
| 1476 |
+
"num_input_tokens_seen": 1416472,
|
| 1477 |
+
"step": 810
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 29.107142857142858,
|
| 1481 |
+
"grad_norm": 0.07188601791858673,
|
| 1482 |
+
"learning_rate": 3.268257975405697e-07,
|
| 1483 |
+
"loss": 0.3438,
|
| 1484 |
+
"num_input_tokens_seen": 1425120,
|
| 1485 |
+
"step": 815
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 29.285714285714285,
|
| 1489 |
+
"grad_norm": 0.07120722532272339,
|
| 1490 |
+
"learning_rate": 2.1329118524827662e-07,
|
| 1491 |
+
"loss": 0.3419,
|
| 1492 |
+
"num_input_tokens_seen": 1433248,
|
| 1493 |
+
"step": 820
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 29.464285714285715,
|
| 1497 |
+
"grad_norm": 0.12804925441741943,
|
| 1498 |
+
"learning_rate": 1.238522554470989e-07,
|
| 1499 |
+
"loss": 0.3417,
|
| 1500 |
+
"num_input_tokens_seen": 1442592,
|
| 1501 |
+
"step": 825
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 29.642857142857142,
|
| 1505 |
+
"grad_norm": 0.26469874382019043,
|
| 1506 |
+
"learning_rate": 5.855229463068712e-08,
|
| 1507 |
+
"loss": 0.3451,
|
| 1508 |
+
"num_input_tokens_seen": 1451360,
|
| 1509 |
+
"step": 830
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 29.821428571428573,
|
| 1513 |
+
"grad_norm": 0.21112701296806335,
|
| 1514 |
+
"learning_rate": 1.742290655755707e-08,
|
| 1515 |
+
"loss": 0.3525,
|
| 1516 |
+
"num_input_tokens_seen": 1459808,
|
| 1517 |
+
"step": 835
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 30.0,
|
| 1521 |
+
"grad_norm": 0.2481113076210022,
|
| 1522 |
+
"learning_rate": 4.839969555581192e-10,
|
| 1523 |
+
"loss": 0.3344,
|
| 1524 |
+
"num_input_tokens_seen": 1468632,
|
| 1525 |
+
"step": 840
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 30.0,
|
| 1529 |
+
"eval_loss": 0.3523610234260559,
|
| 1530 |
+
"eval_runtime": 1.2257,
|
| 1531 |
+
"eval_samples_per_second": 90.557,
|
| 1532 |
+
"eval_steps_per_second": 11.422,
|
| 1533 |
+
"num_input_tokens_seen": 1468632,
|
| 1534 |
+
"step": 840
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 30.0,
|
| 1538 |
+
"num_input_tokens_seen": 1468632,
|
| 1539 |
+
"step": 840,
|
| 1540 |
+
"total_flos": 6.613183518533222e+16,
|
| 1541 |
+
"train_loss": 0.47382661629290806,
|
| 1542 |
+
"train_runtime": 311.6966,
|
| 1543 |
+
"train_samples_per_second": 42.638,
|
| 1544 |
+
"train_steps_per_second": 2.695
|
| 1545 |
+
}
|
| 1546 |
+
],
|
| 1547 |
+
"logging_steps": 5,
|
| 1548 |
+
"max_steps": 840,
|
| 1549 |
+
"num_input_tokens_seen": 1468632,
|
| 1550 |
+
"num_train_epochs": 30,
|
| 1551 |
+
"save_steps": 42,
|
| 1552 |
+
"stateful_callbacks": {
|
| 1553 |
+
"TrainerControl": {
|
| 1554 |
+
"args": {
|
| 1555 |
+
"should_epoch_stop": false,
|
| 1556 |
+
"should_evaluate": false,
|
| 1557 |
+
"should_log": false,
|
| 1558 |
+
"should_save": true,
|
| 1559 |
+
"should_training_stop": true
|
| 1560 |
+
},
|
| 1561 |
+
"attributes": {}
|
| 1562 |
+
}
|
| 1563 |
+
},
|
| 1564 |
+
"total_flos": 6.613183518533222e+16,
|
| 1565 |
+
"train_batch_size": 8,
|
| 1566 |
+
"trial_name": null,
|
| 1567 |
+
"trial_params": null
|
| 1568 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|