End of training
Browse files- README.md +4 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +2877 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3.1
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: prm_gsm_2k_with_full_sol_mix_ref_redistribution_hf
|
|
@@ -15,7 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# prm_gsm_2k_with_full_sol_mix_ref_redistribution_hf
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on
|
|
|
|
|
|
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: prm_gsm_2k_with_full_sol_mix_ref_redistribution_hf
|
|
|
|
| 16 |
|
| 17 |
# prm_gsm_2k_with_full_sol_mix_ref_redistribution_hf
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the prm_conversations_prm_gsm_2k_with_full_sol_gsm_2k_with_full_sol_mix_ref_redistribution_hf dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.1029
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.998766954377312,
|
| 3 |
+
"eval_loss": 0.10287632793188095,
|
| 4 |
+
"eval_runtime": 5.3781,
|
| 5 |
+
"eval_samples_per_second": 48.902,
|
| 6 |
+
"eval_steps_per_second": 6.136,
|
| 7 |
+
"total_flos": 59480064983040.0,
|
| 8 |
+
"train_loss": 0.13387679483419584,
|
| 9 |
+
"train_runtime": 2011.9435,
|
| 10 |
+
"train_samples_per_second": 12.897,
|
| 11 |
+
"train_steps_per_second": 0.201
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.998766954377312,
|
| 3 |
+
"eval_loss": 0.10287632793188095,
|
| 4 |
+
"eval_runtime": 5.3781,
|
| 5 |
+
"eval_samples_per_second": 48.902,
|
| 6 |
+
"eval_steps_per_second": 6.136
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.998766954377312,
|
| 3 |
+
"total_flos": 59480064983040.0,
|
| 4 |
+
"train_loss": 0.13387679483419584,
|
| 5 |
+
"train_runtime": 2011.9435,
|
| 6 |
+
"train_samples_per_second": 12.897,
|
| 7 |
+
"train_steps_per_second": 0.201
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2877 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.998766954377312,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 405,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.002466091245376079,
|
| 13 |
+
"grad_norm": 5.861817070129502,
|
| 14 |
+
"learning_rate": 2.3809523809523811e-07,
|
| 15 |
+
"loss": 0.3009,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.004932182490752158,
|
| 20 |
+
"grad_norm": 5.454702130741054,
|
| 21 |
+
"learning_rate": 4.7619047619047623e-07,
|
| 22 |
+
"loss": 0.3018,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.007398273736128237,
|
| 27 |
+
"grad_norm": 5.409908938548702,
|
| 28 |
+
"learning_rate": 7.142857142857143e-07,
|
| 29 |
+
"loss": 0.3014,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.009864364981504316,
|
| 34 |
+
"grad_norm": 5.284268510657827,
|
| 35 |
+
"learning_rate": 9.523809523809525e-07,
|
| 36 |
+
"loss": 0.3253,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.012330456226880395,
|
| 41 |
+
"grad_norm": 3.9792737440158503,
|
| 42 |
+
"learning_rate": 1.1904761904761906e-06,
|
| 43 |
+
"loss": 0.3081,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.014796547472256474,
|
| 48 |
+
"grad_norm": 3.224749524944109,
|
| 49 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 50 |
+
"loss": 0.2982,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.01726263871763255,
|
| 55 |
+
"grad_norm": 3.066153179820133,
|
| 56 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 57 |
+
"loss": 0.2915,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.01972872996300863,
|
| 62 |
+
"grad_norm": 2.495046000956901,
|
| 63 |
+
"learning_rate": 1.904761904761905e-06,
|
| 64 |
+
"loss": 0.2586,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.02219482120838471,
|
| 69 |
+
"grad_norm": 2.504100758951496,
|
| 70 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 71 |
+
"loss": 0.2827,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.02466091245376079,
|
| 76 |
+
"grad_norm": 2.5327985350425806,
|
| 77 |
+
"learning_rate": 2.380952380952381e-06,
|
| 78 |
+
"loss": 0.2557,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.027127003699136867,
|
| 83 |
+
"grad_norm": 2.5784842750988526,
|
| 84 |
+
"learning_rate": 2.6190476190476192e-06,
|
| 85 |
+
"loss": 0.2529,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.029593094944512947,
|
| 90 |
+
"grad_norm": 2.305467632294948,
|
| 91 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 92 |
+
"loss": 0.2456,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.032059186189889025,
|
| 97 |
+
"grad_norm": 1.68538840427714,
|
| 98 |
+
"learning_rate": 3.0952380952380957e-06,
|
| 99 |
+
"loss": 0.2161,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0345252774352651,
|
| 104 |
+
"grad_norm": 2.1833177512501676,
|
| 105 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 106 |
+
"loss": 0.2223,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.036991368680641186,
|
| 111 |
+
"grad_norm": 2.7881875894363604,
|
| 112 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 113 |
+
"loss": 0.2307,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.03945745992601726,
|
| 118 |
+
"grad_norm": 1.840218395216151,
|
| 119 |
+
"learning_rate": 3.80952380952381e-06,
|
| 120 |
+
"loss": 0.218,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.04192355117139334,
|
| 125 |
+
"grad_norm": 1.6048231245050777,
|
| 126 |
+
"learning_rate": 4.047619047619048e-06,
|
| 127 |
+
"loss": 0.2024,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04438964241676942,
|
| 132 |
+
"grad_norm": 1.3536253435624086,
|
| 133 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 134 |
+
"loss": 0.2247,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.0468557336621455,
|
| 139 |
+
"grad_norm": 1.478284155369018,
|
| 140 |
+
"learning_rate": 4.523809523809524e-06,
|
| 141 |
+
"loss": 0.213,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.04932182490752158,
|
| 146 |
+
"grad_norm": 1.3524715062519435,
|
| 147 |
+
"learning_rate": 4.761904761904762e-06,
|
| 148 |
+
"loss": 0.1948,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.051787916152897656,
|
| 153 |
+
"grad_norm": 1.4067629033670692,
|
| 154 |
+
"learning_rate": 5e-06,
|
| 155 |
+
"loss": 0.2056,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.05425400739827373,
|
| 160 |
+
"grad_norm": 1.1881922454392644,
|
| 161 |
+
"learning_rate": 4.999916334793503e-06,
|
| 162 |
+
"loss": 0.2068,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.05672009864364982,
|
| 167 |
+
"grad_norm": 1.1590822007181945,
|
| 168 |
+
"learning_rate": 4.9996653447739054e-06,
|
| 169 |
+
"loss": 0.1917,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.059186189889025895,
|
| 174 |
+
"grad_norm": 1.505914442287729,
|
| 175 |
+
"learning_rate": 4.999247046740511e-06,
|
| 176 |
+
"loss": 0.1941,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.06165228113440197,
|
| 181 |
+
"grad_norm": 1.3782401676884288,
|
| 182 |
+
"learning_rate": 4.998661468690914e-06,
|
| 183 |
+
"loss": 0.2084,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.06411837237977805,
|
| 188 |
+
"grad_norm": 1.1590898938147485,
|
| 189 |
+
"learning_rate": 4.997908649819122e-06,
|
| 190 |
+
"loss": 0.207,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.06658446362515413,
|
| 195 |
+
"grad_norm": 1.079882983156062,
|
| 196 |
+
"learning_rate": 4.996988640512931e-06,
|
| 197 |
+
"loss": 0.1817,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.0690505548705302,
|
| 202 |
+
"grad_norm": 1.050596115616474,
|
| 203 |
+
"learning_rate": 4.995901502350556e-06,
|
| 204 |
+
"loss": 0.177,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.07151664611590629,
|
| 209 |
+
"grad_norm": 1.068725153791535,
|
| 210 |
+
"learning_rate": 4.994647308096509e-06,
|
| 211 |
+
"loss": 0.1778,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.07398273736128237,
|
| 216 |
+
"grad_norm": 1.0596573616643041,
|
| 217 |
+
"learning_rate": 4.993226141696726e-06,
|
| 218 |
+
"loss": 0.1681,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07644882860665844,
|
| 223 |
+
"grad_norm": 1.1753877395371195,
|
| 224 |
+
"learning_rate": 4.991638098272951e-06,
|
| 225 |
+
"loss": 0.1799,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.07891491985203453,
|
| 230 |
+
"grad_norm": 0.9897959119535461,
|
| 231 |
+
"learning_rate": 4.98988328411637e-06,
|
| 232 |
+
"loss": 0.1662,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.08138101109741061,
|
| 237 |
+
"grad_norm": 1.1598630094827584,
|
| 238 |
+
"learning_rate": 4.987961816680493e-06,
|
| 239 |
+
"loss": 0.1907,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.08384710234278668,
|
| 244 |
+
"grad_norm": 1.0031559822876062,
|
| 245 |
+
"learning_rate": 4.9858738245732955e-06,
|
| 246 |
+
"loss": 0.1867,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.08631319358816276,
|
| 251 |
+
"grad_norm": 0.9938289593496852,
|
| 252 |
+
"learning_rate": 4.983619447548611e-06,
|
| 253 |
+
"loss": 0.1781,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.08877928483353884,
|
| 258 |
+
"grad_norm": 0.9541253544509707,
|
| 259 |
+
"learning_rate": 4.981198836496776e-06,
|
| 260 |
+
"loss": 0.1771,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.09124537607891492,
|
| 265 |
+
"grad_norm": 0.9901594561902145,
|
| 266 |
+
"learning_rate": 4.978612153434527e-06,
|
| 267 |
+
"loss": 0.179,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.093711467324291,
|
| 272 |
+
"grad_norm": 1.0815334938640082,
|
| 273 |
+
"learning_rate": 4.9758595714941625e-06,
|
| 274 |
+
"loss": 0.1749,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.09617755856966707,
|
| 279 |
+
"grad_norm": 0.941987623128398,
|
| 280 |
+
"learning_rate": 4.972941274911953e-06,
|
| 281 |
+
"loss": 0.1685,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.09864364981504316,
|
| 286 |
+
"grad_norm": 0.9544612545429672,
|
| 287 |
+
"learning_rate": 4.969857459015807e-06,
|
| 288 |
+
"loss": 0.169,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.10110974106041924,
|
| 293 |
+
"grad_norm": 1.0059134522516986,
|
| 294 |
+
"learning_rate": 4.966608330212198e-06,
|
| 295 |
+
"loss": 0.1829,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.10357583230579531,
|
| 300 |
+
"grad_norm": 1.0576728568589142,
|
| 301 |
+
"learning_rate": 4.9631941059723535e-06,
|
| 302 |
+
"loss": 0.1656,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1060419235511714,
|
| 307 |
+
"grad_norm": 1.022560248491249,
|
| 308 |
+
"learning_rate": 4.959615014817694e-06,
|
| 309 |
+
"loss": 0.1659,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.10850801479654747,
|
| 314 |
+
"grad_norm": 1.0182845823122424,
|
| 315 |
+
"learning_rate": 4.95587129630454e-06,
|
| 316 |
+
"loss": 0.166,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.11097410604192355,
|
| 321 |
+
"grad_norm": 1.0281336322899663,
|
| 322 |
+
"learning_rate": 4.9519632010080765e-06,
|
| 323 |
+
"loss": 0.1761,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.11344019728729964,
|
| 328 |
+
"grad_norm": 1.051663791604757,
|
| 329 |
+
"learning_rate": 4.947890990505585e-06,
|
| 330 |
+
"loss": 0.1739,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.1159062885326757,
|
| 335 |
+
"grad_norm": 1.0867150445001044,
|
| 336 |
+
"learning_rate": 4.94365493735893e-06,
|
| 337 |
+
"loss": 0.1646,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.11837237977805179,
|
| 342 |
+
"grad_norm": 1.0161358760594283,
|
| 343 |
+
"learning_rate": 4.939255325096322e-06,
|
| 344 |
+
"loss": 0.1647,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.12083847102342787,
|
| 349 |
+
"grad_norm": 1.0643048734894447,
|
| 350 |
+
"learning_rate": 4.9346924481933345e-06,
|
| 351 |
+
"loss": 0.1732,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.12330456226880394,
|
| 356 |
+
"grad_norm": 0.9273473173688057,
|
| 357 |
+
"learning_rate": 4.9299666120531986e-06,
|
| 358 |
+
"loss": 0.1511,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.12577065351418001,
|
| 363 |
+
"grad_norm": 1.067677795114817,
|
| 364 |
+
"learning_rate": 4.925078132986361e-06,
|
| 365 |
+
"loss": 0.1619,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.1282367447595561,
|
| 370 |
+
"grad_norm": 1.0662297711232929,
|
| 371 |
+
"learning_rate": 4.920027338189307e-06,
|
| 372 |
+
"loss": 0.1656,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.13070283600493218,
|
| 377 |
+
"grad_norm": 1.0067512317098979,
|
| 378 |
+
"learning_rate": 4.914814565722671e-06,
|
| 379 |
+
"loss": 0.1634,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.13316892725030827,
|
| 384 |
+
"grad_norm": 1.0598968407249205,
|
| 385 |
+
"learning_rate": 4.9094401644886e-06,
|
| 386 |
+
"loss": 0.159,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13563501849568435,
|
| 391 |
+
"grad_norm": 0.997053808521598,
|
| 392 |
+
"learning_rate": 4.9039044942074055e-06,
|
| 393 |
+
"loss": 0.1562,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.1381011097410604,
|
| 398 |
+
"grad_norm": 1.0024366039477026,
|
| 399 |
+
"learning_rate": 4.898207925393485e-06,
|
| 400 |
+
"loss": 0.1619,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.1405672009864365,
|
| 405 |
+
"grad_norm": 1.0112435019512296,
|
| 406 |
+
"learning_rate": 4.8923508393305224e-06,
|
| 407 |
+
"loss": 0.1659,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.14303329223181258,
|
| 412 |
+
"grad_norm": 0.9940992706423503,
|
| 413 |
+
"learning_rate": 4.886333628045972e-06,
|
| 414 |
+
"loss": 0.1454,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.14549938347718866,
|
| 419 |
+
"grad_norm": 1.0232792895797451,
|
| 420 |
+
"learning_rate": 4.880156694284811e-06,
|
| 421 |
+
"loss": 0.1614,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.14796547472256474,
|
| 426 |
+
"grad_norm": 1.0589693172755839,
|
| 427 |
+
"learning_rate": 4.873820451482592e-06,
|
| 428 |
+
"loss": 0.1565,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.1504315659679408,
|
| 433 |
+
"grad_norm": 1.0651628059954148,
|
| 434 |
+
"learning_rate": 4.867325323737765e-06,
|
| 435 |
+
"loss": 0.1565,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.15289765721331688,
|
| 440 |
+
"grad_norm": 1.1602100272277833,
|
| 441 |
+
"learning_rate": 4.860671745783292e-06,
|
| 442 |
+
"loss": 0.1564,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.15536374845869297,
|
| 447 |
+
"grad_norm": 1.0157934500809114,
|
| 448 |
+
"learning_rate": 4.8538601629575525e-06,
|
| 449 |
+
"loss": 0.1525,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.15782983970406905,
|
| 454 |
+
"grad_norm": 1.0669325172009674,
|
| 455 |
+
"learning_rate": 4.846891031174533e-06,
|
| 456 |
+
"loss": 0.152,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.16029593094944514,
|
| 461 |
+
"grad_norm": 1.0787706336711722,
|
| 462 |
+
"learning_rate": 4.839764816893315e-06,
|
| 463 |
+
"loss": 0.1472,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.16276202219482122,
|
| 468 |
+
"grad_norm": 1.1141588294306453,
|
| 469 |
+
"learning_rate": 4.832481997086848e-06,
|
| 470 |
+
"loss": 0.1604,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.16522811344019728,
|
| 475 |
+
"grad_norm": 1.089583543896409,
|
| 476 |
+
"learning_rate": 4.82504305921003e-06,
|
| 477 |
+
"loss": 0.1615,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.16769420468557336,
|
| 482 |
+
"grad_norm": 1.2753507516006068,
|
| 483 |
+
"learning_rate": 4.817448501167083e-06,
|
| 484 |
+
"loss": 0.1583,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.17016029593094945,
|
| 489 |
+
"grad_norm": 1.1865381497470653,
|
| 490 |
+
"learning_rate": 4.809698831278217e-06,
|
| 491 |
+
"loss": 0.1613,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.17262638717632553,
|
| 496 |
+
"grad_norm": 1.1603795269742387,
|
| 497 |
+
"learning_rate": 4.801794568245619e-06,
|
| 498 |
+
"loss": 0.1553,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.17509247842170161,
|
| 503 |
+
"grad_norm": 1.0089518603728724,
|
| 504 |
+
"learning_rate": 4.793736241118729e-06,
|
| 505 |
+
"loss": 0.1487,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.17755856966707767,
|
| 510 |
+
"grad_norm": 1.097433021678051,
|
| 511 |
+
"learning_rate": 4.7855243892588275e-06,
|
| 512 |
+
"loss": 0.1652,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.18002466091245375,
|
| 517 |
+
"grad_norm": 1.4189381038349331,
|
| 518 |
+
"learning_rate": 4.77715956230294e-06,
|
| 519 |
+
"loss": 0.1679,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.18249075215782984,
|
| 524 |
+
"grad_norm": 1.1049506395029285,
|
| 525 |
+
"learning_rate": 4.768642320127047e-06,
|
| 526 |
+
"loss": 0.1586,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.18495684340320592,
|
| 531 |
+
"grad_norm": 1.1192651883282672,
|
| 532 |
+
"learning_rate": 4.759973232808609e-06,
|
| 533 |
+
"loss": 0.1671,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.187422934648582,
|
| 538 |
+
"grad_norm": 1.3491373705500664,
|
| 539 |
+
"learning_rate": 4.75115288058841e-06,
|
| 540 |
+
"loss": 0.1535,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.18988902589395806,
|
| 545 |
+
"grad_norm": 0.9948177751908293,
|
| 546 |
+
"learning_rate": 4.742181853831721e-06,
|
| 547 |
+
"loss": 0.162,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.19235511713933415,
|
| 552 |
+
"grad_norm": 1.464829663274484,
|
| 553 |
+
"learning_rate": 4.7330607529887885e-06,
|
| 554 |
+
"loss": 0.166,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.19482120838471023,
|
| 559 |
+
"grad_norm": 1.2400996455259394,
|
| 560 |
+
"learning_rate": 4.7237901885546405e-06,
|
| 561 |
+
"loss": 0.1611,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.19728729963008632,
|
| 566 |
+
"grad_norm": 0.958467279088388,
|
| 567 |
+
"learning_rate": 4.714370781028227e-06,
|
| 568 |
+
"loss": 0.151,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.1997533908754624,
|
| 573 |
+
"grad_norm": 1.038801453629039,
|
| 574 |
+
"learning_rate": 4.704803160870888e-06,
|
| 575 |
+
"loss": 0.1598,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.20221948212083848,
|
| 580 |
+
"grad_norm": 1.0427896424251202,
|
| 581 |
+
"learning_rate": 4.695087968464157e-06,
|
| 582 |
+
"loss": 0.1509,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.20468557336621454,
|
| 587 |
+
"grad_norm": 1.0365538149283298,
|
| 588 |
+
"learning_rate": 4.685225854066897e-06,
|
| 589 |
+
"loss": 0.1415,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.20715166461159062,
|
| 594 |
+
"grad_norm": 0.9863701011972386,
|
| 595 |
+
"learning_rate": 4.675217477771779e-06,
|
| 596 |
+
"loss": 0.1628,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2096177558569667,
|
| 601 |
+
"grad_norm": 0.9053867787577642,
|
| 602 |
+
"learning_rate": 4.665063509461098e-06,
|
| 603 |
+
"loss": 0.1477,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.2120838471023428,
|
| 608 |
+
"grad_norm": 1.0371758997277913,
|
| 609 |
+
"learning_rate": 4.6547646287619365e-06,
|
| 610 |
+
"loss": 0.1451,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.21454993834771888,
|
| 615 |
+
"grad_norm": 0.9996959873573208,
|
| 616 |
+
"learning_rate": 4.644321525000681e-06,
|
| 617 |
+
"loss": 0.1584,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.21701602959309493,
|
| 622 |
+
"grad_norm": 1.0419473584314491,
|
| 623 |
+
"learning_rate": 4.6337348971568765e-06,
|
| 624 |
+
"loss": 0.1556,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.21948212083847102,
|
| 629 |
+
"grad_norm": 1.176505030693245,
|
| 630 |
+
"learning_rate": 4.623005453816447e-06,
|
| 631 |
+
"loss": 0.1562,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.2219482120838471,
|
| 636 |
+
"grad_norm": 1.0895728685600041,
|
| 637 |
+
"learning_rate": 4.612133913124268e-06,
|
| 638 |
+
"loss": 0.1446,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.22441430332922319,
|
| 643 |
+
"grad_norm": 0.9577811245445861,
|
| 644 |
+
"learning_rate": 4.601121002736095e-06,
|
| 645 |
+
"loss": 0.1548,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.22688039457459927,
|
| 650 |
+
"grad_norm": 0.9627408268360163,
|
| 651 |
+
"learning_rate": 4.589967459769867e-06,
|
| 652 |
+
"loss": 0.1466,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.22934648581997533,
|
| 657 |
+
"grad_norm": 1.0947242286079646,
|
| 658 |
+
"learning_rate": 4.578674030756364e-06,
|
| 659 |
+
"loss": 0.1436,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.2318125770653514,
|
| 664 |
+
"grad_norm": 1.0146604735253648,
|
| 665 |
+
"learning_rate": 4.567241471589241e-06,
|
| 666 |
+
"loss": 0.1486,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.2342786683107275,
|
| 671 |
+
"grad_norm": 1.1123609539340134,
|
| 672 |
+
"learning_rate": 4.555670547474438e-06,
|
| 673 |
+
"loss": 0.1491,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.23674475955610358,
|
| 678 |
+
"grad_norm": 0.9874156166248013,
|
| 679 |
+
"learning_rate": 4.54396203287896e-06,
|
| 680 |
+
"loss": 0.1367,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.23921085080147966,
|
| 685 |
+
"grad_norm": 1.1084009571098752,
|
| 686 |
+
"learning_rate": 4.532116711479039e-06,
|
| 687 |
+
"loss": 0.1508,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.24167694204685575,
|
| 692 |
+
"grad_norm": 1.1351235449226345,
|
| 693 |
+
"learning_rate": 4.520135376107683e-06,
|
| 694 |
+
"loss": 0.1467,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.2441430332922318,
|
| 699 |
+
"grad_norm": 1.0536269384602388,
|
| 700 |
+
"learning_rate": 4.508018828701613e-06,
|
| 701 |
+
"loss": 0.1551,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.2466091245376079,
|
| 706 |
+
"grad_norm": 1.0359890676928114,
|
| 707 |
+
"learning_rate": 4.495767880247579e-06,
|
| 708 |
+
"loss": 0.1517,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.24907521578298397,
|
| 713 |
+
"grad_norm": 1.006489152226709,
|
| 714 |
+
"learning_rate": 4.4833833507280884e-06,
|
| 715 |
+
"loss": 0.1464,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.25154130702836003,
|
| 720 |
+
"grad_norm": 1.001144890936093,
|
| 721 |
+
"learning_rate": 4.470866069066516e-06,
|
| 722 |
+
"loss": 0.1477,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.25400739827373614,
|
| 727 |
+
"grad_norm": 0.8702870273664955,
|
| 728 |
+
"learning_rate": 4.4582168730716264e-06,
|
| 729 |
+
"loss": 0.1351,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.2564734895191122,
|
| 734 |
+
"grad_norm": 1.00145984830617,
|
| 735 |
+
"learning_rate": 4.445436609381495e-06,
|
| 736 |
+
"loss": 0.1382,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.2589395807644883,
|
| 741 |
+
"grad_norm": 1.0209833966786983,
|
| 742 |
+
"learning_rate": 4.432526133406843e-06,
|
| 743 |
+
"loss": 0.147,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.26140567200986436,
|
| 748 |
+
"grad_norm": 0.9926067510832088,
|
| 749 |
+
"learning_rate": 4.4194863092737805e-06,
|
| 750 |
+
"loss": 0.1469,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.2638717632552404,
|
| 755 |
+
"grad_norm": 1.0902897518646402,
|
| 756 |
+
"learning_rate": 4.406318009765971e-06,
|
| 757 |
+
"loss": 0.1375,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.26633785450061653,
|
| 762 |
+
"grad_norm": 1.2408315839259865,
|
| 763 |
+
"learning_rate": 4.393022116266212e-06,
|
| 764 |
+
"loss": 0.1485,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.2688039457459926,
|
| 769 |
+
"grad_norm": 1.0435133677408495,
|
| 770 |
+
"learning_rate": 4.379599518697444e-06,
|
| 771 |
+
"loss": 0.1484,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.2712700369913687,
|
| 776 |
+
"grad_norm": 1.0000775609037789,
|
| 777 |
+
"learning_rate": 4.366051115463185e-06,
|
| 778 |
+
"loss": 0.1459,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.27373612823674476,
|
| 783 |
+
"grad_norm": 1.1265421997244343,
|
| 784 |
+
"learning_rate": 4.352377813387398e-06,
|
| 785 |
+
"loss": 0.1504,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.2762022194821208,
|
| 790 |
+
"grad_norm": 1.0547968511485402,
|
| 791 |
+
"learning_rate": 4.338580527653797e-06,
|
| 792 |
+
"loss": 0.1438,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.2786683107274969,
|
| 797 |
+
"grad_norm": 0.9726420995283216,
|
| 798 |
+
"learning_rate": 4.324660181744589e-06,
|
| 799 |
+
"loss": 0.1381,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.281134401972873,
|
| 804 |
+
"grad_norm": 0.9070628149596835,
|
| 805 |
+
"learning_rate": 4.3106177073786684e-06,
|
| 806 |
+
"loss": 0.1405,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.2836004932182491,
|
| 811 |
+
"grad_norm": 0.9158260556696106,
|
| 812 |
+
"learning_rate": 4.296454044449245e-06,
|
| 813 |
+
"loss": 0.1359,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.28606658446362515,
|
| 818 |
+
"grad_norm": 0.9765082984745848,
|
| 819 |
+
"learning_rate": 4.28217014096095e-06,
|
| 820 |
+
"loss": 0.1525,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.2885326757090012,
|
| 825 |
+
"grad_norm": 0.9701123483287503,
|
| 826 |
+
"learning_rate": 4.267766952966369e-06,
|
| 827 |
+
"loss": 0.1453,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.2909987669543773,
|
| 832 |
+
"grad_norm": 0.9028970173158747,
|
| 833 |
+
"learning_rate": 4.253245444502061e-06,
|
| 834 |
+
"loss": 0.1427,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.2934648581997534,
|
| 839 |
+
"grad_norm": 0.870627710582214,
|
| 840 |
+
"learning_rate": 4.2386065875240295e-06,
|
| 841 |
+
"loss": 0.1373,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.2959309494451295,
|
| 846 |
+
"grad_norm": 0.9094737389101193,
|
| 847 |
+
"learning_rate": 4.223851361842668e-06,
|
| 848 |
+
"loss": 0.1426,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.29839704069050554,
|
| 853 |
+
"grad_norm": 0.9510536958264116,
|
| 854 |
+
"learning_rate": 4.2089807550571786e-06,
|
| 855 |
+
"loss": 0.1349,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.3008631319358816,
|
| 860 |
+
"grad_norm": 0.9977566163131177,
|
| 861 |
+
"learning_rate": 4.193995762489472e-06,
|
| 862 |
+
"loss": 0.1354,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.3033292231812577,
|
| 867 |
+
"grad_norm": 0.9883040970106479,
|
| 868 |
+
"learning_rate": 4.178897387117547e-06,
|
| 869 |
+
"loss": 0.1285,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.30579531442663377,
|
| 874 |
+
"grad_norm": 0.989227476749991,
|
| 875 |
+
"learning_rate": 4.163686639508356e-06,
|
| 876 |
+
"loss": 0.1295,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.3082614056720099,
|
| 881 |
+
"grad_norm": 1.030899401070542,
|
| 882 |
+
"learning_rate": 4.1483645377501726e-06,
|
| 883 |
+
"loss": 0.1405,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.31072749691738594,
|
| 888 |
+
"grad_norm": 0.9963592007255958,
|
| 889 |
+
"learning_rate": 4.1329321073844415e-06,
|
| 890 |
+
"loss": 0.1395,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.31319358816276205,
|
| 895 |
+
"grad_norm": 1.036677676843521,
|
| 896 |
+
"learning_rate": 4.117390381337144e-06,
|
| 897 |
+
"loss": 0.1482,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.3156596794081381,
|
| 902 |
+
"grad_norm": 1.0252893741480706,
|
| 903 |
+
"learning_rate": 4.1017403998496526e-06,
|
| 904 |
+
"loss": 0.1402,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.31812577065351416,
|
| 909 |
+
"grad_norm": 1.0175275526046348,
|
| 910 |
+
"learning_rate": 4.085983210409114e-06,
|
| 911 |
+
"loss": 0.1427,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.3205918618988903,
|
| 916 |
+
"grad_norm": 0.8721732883270084,
|
| 917 |
+
"learning_rate": 4.070119867678336e-06,
|
| 918 |
+
"loss": 0.1227,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.32305795314426633,
|
| 923 |
+
"grad_norm": 0.9724653410645763,
|
| 924 |
+
"learning_rate": 4.054151433425194e-06,
|
| 925 |
+
"loss": 0.1432,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.32552404438964244,
|
| 930 |
+
"grad_norm": 0.9883653901817202,
|
| 931 |
+
"learning_rate": 4.038078976451567e-06,
|
| 932 |
+
"loss": 0.1375,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.3279901356350185,
|
| 937 |
+
"grad_norm": 0.9516129708732718,
|
| 938 |
+
"learning_rate": 4.021903572521802e-06,
|
| 939 |
+
"loss": 0.1339,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.33045622688039455,
|
| 944 |
+
"grad_norm": 0.9809573251347133,
|
| 945 |
+
"learning_rate": 4.005626304290706e-06,
|
| 946 |
+
"loss": 0.1324,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.33292231812577067,
|
| 951 |
+
"grad_norm": 0.9480627599795061,
|
| 952 |
+
"learning_rate": 3.989248261231084e-06,
|
| 953 |
+
"loss": 0.1414,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.3353884093711467,
|
| 958 |
+
"grad_norm": 0.9924590629405446,
|
| 959 |
+
"learning_rate": 3.972770539560821e-06,
|
| 960 |
+
"loss": 0.1399,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.33785450061652283,
|
| 965 |
+
"grad_norm": 1.0290842479312958,
|
| 966 |
+
"learning_rate": 3.956194242169506e-06,
|
| 967 |
+
"loss": 0.1359,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.3403205918618989,
|
| 972 |
+
"grad_norm": 0.9576444844339111,
|
| 973 |
+
"learning_rate": 3.939520478544614e-06,
|
| 974 |
+
"loss": 0.1256,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.34278668310727495,
|
| 979 |
+
"grad_norm": 0.9461300326238143,
|
| 980 |
+
"learning_rate": 3.922750364697246e-06,
|
| 981 |
+
"loss": 0.1315,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.34525277435265106,
|
| 986 |
+
"grad_norm": 0.9011347978577009,
|
| 987 |
+
"learning_rate": 3.905885023087434e-06,
|
| 988 |
+
"loss": 0.1316,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.3477188655980271,
|
| 993 |
+
"grad_norm": 0.9715551507077147,
|
| 994 |
+
"learning_rate": 3.888925582549006e-06,
|
| 995 |
+
"loss": 0.1435,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.35018495684340323,
|
| 1000 |
+
"grad_norm": 0.9172796475982884,
|
| 1001 |
+
"learning_rate": 3.87187317821404e-06,
|
| 1002 |
+
"loss": 0.1281,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.3526510480887793,
|
| 1007 |
+
"grad_norm": 0.9140278725241636,
|
| 1008 |
+
"learning_rate": 3.8547289514368795e-06,
|
| 1009 |
+
"loss": 0.1344,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.35511713933415534,
|
| 1014 |
+
"grad_norm": 0.9939727532884871,
|
| 1015 |
+
"learning_rate": 3.8374940497177435e-06,
|
| 1016 |
+
"loss": 0.1413,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.35758323057953145,
|
| 1021 |
+
"grad_norm": 0.9302481655408,
|
| 1022 |
+
"learning_rate": 3.82016962662592e-06,
|
| 1023 |
+
"loss": 0.1313,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.3600493218249075,
|
| 1028 |
+
"grad_norm": 1.0029221726732422,
|
| 1029 |
+
"learning_rate": 3.802756841722559e-06,
|
| 1030 |
+
"loss": 0.1415,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.3625154130702836,
|
| 1035 |
+
"grad_norm": 0.9641843153968267,
|
| 1036 |
+
"learning_rate": 3.785256860483054e-06,
|
| 1037 |
+
"loss": 0.1326,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.3649815043156597,
|
| 1042 |
+
"grad_norm": 0.9460102984345635,
|
| 1043 |
+
"learning_rate": 3.767670854219043e-06,
|
| 1044 |
+
"loss": 0.1413,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.36744759556103573,
|
| 1049 |
+
"grad_norm": 0.8936869530293807,
|
| 1050 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 1051 |
+
"loss": 0.1265,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.36991368680641185,
|
| 1056 |
+
"grad_norm": 0.9790647067577934,
|
| 1057 |
+
"learning_rate": 3.7322454805744605e-06,
|
| 1058 |
+
"loss": 0.1299,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.3723797780517879,
|
| 1063 |
+
"grad_norm": 0.9954057916656672,
|
| 1064 |
+
"learning_rate": 3.7144084842908506e-06,
|
| 1065 |
+
"loss": 0.1307,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.374845869297164,
|
| 1070 |
+
"grad_norm": 0.8953071694198275,
|
| 1071 |
+
"learning_rate": 3.696490205017953e-06,
|
| 1072 |
+
"loss": 0.1278,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.37731196054254007,
|
| 1077 |
+
"grad_norm": 0.9078908530700724,
|
| 1078 |
+
"learning_rate": 3.6784918420649952e-06,
|
| 1079 |
+
"loss": 0.1333,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.3797780517879161,
|
| 1084 |
+
"grad_norm": 0.9702112005143135,
|
| 1085 |
+
"learning_rate": 3.660414600101379e-06,
|
| 1086 |
+
"loss": 0.1418,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.38224414303329224,
|
| 1091 |
+
"grad_norm": 0.8754729035657617,
|
| 1092 |
+
"learning_rate": 3.642259689076052e-06,
|
| 1093 |
+
"loss": 0.1303,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.3847102342786683,
|
| 1098 |
+
"grad_norm": 1.0019337448881533,
|
| 1099 |
+
"learning_rate": 3.624028324136517e-06,
|
| 1100 |
+
"loss": 0.1342,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.3871763255240444,
|
| 1105 |
+
"grad_norm": 0.9235881685046038,
|
| 1106 |
+
"learning_rate": 3.6057217255475034e-06,
|
| 1107 |
+
"loss": 0.133,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.38964241676942046,
|
| 1112 |
+
"grad_norm": 0.9037731915231353,
|
| 1113 |
+
"learning_rate": 3.587341118609293e-06,
|
| 1114 |
+
"loss": 0.1297,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.3921085080147966,
|
| 1119 |
+
"grad_norm": 0.9069564887617025,
|
| 1120 |
+
"learning_rate": 3.5688877335757055e-06,
|
| 1121 |
+
"loss": 0.1288,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.39457459926017263,
|
| 1126 |
+
"grad_norm": 0.9109051862203179,
|
| 1127 |
+
"learning_rate": 3.5503628055717566e-06,
|
| 1128 |
+
"loss": 0.1373,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.3970406905055487,
|
| 1133 |
+
"grad_norm": 0.9352261195746291,
|
| 1134 |
+
"learning_rate": 3.531767574510987e-06,
|
| 1135 |
+
"loss": 0.1346,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.3995067817509248,
|
| 1140 |
+
"grad_norm": 0.8715670799441637,
|
| 1141 |
+
"learning_rate": 3.5131032850124745e-06,
|
| 1142 |
+
"loss": 0.1354,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.40197287299630086,
|
| 1147 |
+
"grad_norm": 0.838780801432847,
|
| 1148 |
+
"learning_rate": 3.494371186317528e-06,
|
| 1149 |
+
"loss": 0.1239,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.40443896424167697,
|
| 1154 |
+
"grad_norm": 0.9741683793360403,
|
| 1155 |
+
"learning_rate": 3.4755725322060705e-06,
|
| 1156 |
+
"loss": 0.1359,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.406905055487053,
|
| 1161 |
+
"grad_norm": 0.872735557906736,
|
| 1162 |
+
"learning_rate": 3.4567085809127247e-06,
|
| 1163 |
+
"loss": 0.126,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.4093711467324291,
|
| 1168 |
+
"grad_norm": 1.0331640452763273,
|
| 1169 |
+
"learning_rate": 3.437780595042595e-06,
|
| 1170 |
+
"loss": 0.1381,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4118372379778052,
|
| 1175 |
+
"grad_norm": 1.044024517690311,
|
| 1176 |
+
"learning_rate": 3.4187898414867594e-06,
|
| 1177 |
+
"loss": 0.1468,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.41430332922318125,
|
| 1182 |
+
"grad_norm": 0.9084847723223093,
|
| 1183 |
+
"learning_rate": 3.399737591337471e-06,
|
| 1184 |
+
"loss": 0.1279,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.41676942046855736,
|
| 1189 |
+
"grad_norm": 0.9711336315434675,
|
| 1190 |
+
"learning_rate": 3.3806251198030843e-06,
|
| 1191 |
+
"loss": 0.1229,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.4192355117139334,
|
| 1196 |
+
"grad_norm": 0.923594854404584,
|
| 1197 |
+
"learning_rate": 3.3614537061227015e-06,
|
| 1198 |
+
"loss": 0.137,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.4217016029593095,
|
| 1203 |
+
"grad_norm": 0.9619577578045375,
|
| 1204 |
+
"learning_rate": 3.3422246334805504e-06,
|
| 1205 |
+
"loss": 0.1241,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.4241676942046856,
|
| 1210 |
+
"grad_norm": 1.0260727251018134,
|
| 1211 |
+
"learning_rate": 3.3229391889200974e-06,
|
| 1212 |
+
"loss": 0.1395,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.42663378545006164,
|
| 1217 |
+
"grad_norm": 0.9702238898813292,
|
| 1218 |
+
"learning_rate": 3.303598663257904e-06,
|
| 1219 |
+
"loss": 0.1204,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.42909987669543775,
|
| 1224 |
+
"grad_norm": 0.8785118494001238,
|
| 1225 |
+
"learning_rate": 3.2842043509972294e-06,
|
| 1226 |
+
"loss": 0.123,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.4315659679408138,
|
| 1231 |
+
"grad_norm": 0.9906512700651107,
|
| 1232 |
+
"learning_rate": 3.264757550241384e-06,
|
| 1233 |
+
"loss": 0.1305,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.43403205918618987,
|
| 1238 |
+
"grad_norm": 1.0291400265356092,
|
| 1239 |
+
"learning_rate": 3.2452595626068505e-06,
|
| 1240 |
+
"loss": 0.1295,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.436498150431566,
|
| 1245 |
+
"grad_norm": 0.8900049718639798,
|
| 1246 |
+
"learning_rate": 3.225711693136156e-06,
|
| 1247 |
+
"loss": 0.1215,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.43896424167694204,
|
| 1252 |
+
"grad_norm": 0.9165496528504126,
|
| 1253 |
+
"learning_rate": 3.2061152502105315e-06,
|
| 1254 |
+
"loss": 0.1379,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.44143033292231815,
|
| 1259 |
+
"grad_norm": 1.0237540222192676,
|
| 1260 |
+
"learning_rate": 3.1864715454623312e-06,
|
| 1261 |
+
"loss": 0.1314,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.4438964241676942,
|
| 1266 |
+
"grad_norm": 0.8797603130268353,
|
| 1267 |
+
"learning_rate": 3.1667818936872463e-06,
|
| 1268 |
+
"loss": 0.1343,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.44636251541307026,
|
| 1273 |
+
"grad_norm": 0.8695807338732953,
|
| 1274 |
+
"learning_rate": 3.147047612756302e-06,
|
| 1275 |
+
"loss": 0.1298,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.44882860665844637,
|
| 1280 |
+
"grad_norm": 0.9361087356106694,
|
| 1281 |
+
"learning_rate": 3.127270023527651e-06,
|
| 1282 |
+
"loss": 0.1262,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.45129469790382243,
|
| 1287 |
+
"grad_norm": 0.9958169849786923,
|
| 1288 |
+
"learning_rate": 3.10745044975816e-06,
|
| 1289 |
+
"loss": 0.1213,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.45376078914919854,
|
| 1294 |
+
"grad_norm": 0.9704071885308517,
|
| 1295 |
+
"learning_rate": 3.087590218014817e-06,
|
| 1296 |
+
"loss": 0.121,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.4562268803945746,
|
| 1301 |
+
"grad_norm": 0.9570400842235397,
|
| 1302 |
+
"learning_rate": 3.0676906575859335e-06,
|
| 1303 |
+
"loss": 0.1255,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.45869297163995065,
|
| 1308 |
+
"grad_norm": 0.930247862801058,
|
| 1309 |
+
"learning_rate": 3.0477531003921745e-06,
|
| 1310 |
+
"loss": 0.1238,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.46115906288532676,
|
| 1315 |
+
"grad_norm": 1.0359658689439766,
|
| 1316 |
+
"learning_rate": 3.027778880897413e-06,
|
| 1317 |
+
"loss": 0.125,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.4636251541307028,
|
| 1322 |
+
"grad_norm": 1.0188141528209287,
|
| 1323 |
+
"learning_rate": 3.0077693360194076e-06,
|
| 1324 |
+
"loss": 0.1248,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.46609124537607893,
|
| 1329 |
+
"grad_norm": 0.9196224508884366,
|
| 1330 |
+
"learning_rate": 2.9877258050403214e-06,
|
| 1331 |
+
"loss": 0.1287,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.468557336621455,
|
| 1336 |
+
"grad_norm": 0.940538541732393,
|
| 1337 |
+
"learning_rate": 2.9676496295170803e-06,
|
| 1338 |
+
"loss": 0.1205,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.47102342786683105,
|
| 1343 |
+
"grad_norm": 0.9372690050183837,
|
| 1344 |
+
"learning_rate": 2.947542153191583e-06,
|
| 1345 |
+
"loss": 0.121,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.47348951911220716,
|
| 1350 |
+
"grad_norm": 1.0443308791545658,
|
| 1351 |
+
"learning_rate": 2.9274047219007533e-06,
|
| 1352 |
+
"loss": 0.1383,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.4759556103575832,
|
| 1357 |
+
"grad_norm": 1.0418581231308712,
|
| 1358 |
+
"learning_rate": 2.9072386834864723e-06,
|
| 1359 |
+
"loss": 0.1211,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.4784217016029593,
|
| 1364 |
+
"grad_norm": 0.9381944135721112,
|
| 1365 |
+
"learning_rate": 2.8870453877053528e-06,
|
| 1366 |
+
"loss": 0.123,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.4808877928483354,
|
| 1371 |
+
"grad_norm": 0.9816359859051503,
|
| 1372 |
+
"learning_rate": 2.8668261861384045e-06,
|
| 1373 |
+
"loss": 0.1252,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.4833538840937115,
|
| 1378 |
+
"grad_norm": 0.9807964918939387,
|
| 1379 |
+
"learning_rate": 2.846582432100567e-06,
|
| 1380 |
+
"loss": 0.1261,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.48581997533908755,
|
| 1385 |
+
"grad_norm": 1.0136872010740587,
|
| 1386 |
+
"learning_rate": 2.82631548055013e-06,
|
| 1387 |
+
"loss": 0.1256,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.4882860665844636,
|
| 1392 |
+
"grad_norm": 0.979642819047493,
|
| 1393 |
+
"learning_rate": 2.806026687998041e-06,
|
| 1394 |
+
"loss": 0.1222,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.4907521578298397,
|
| 1399 |
+
"grad_norm": 0.90369525282672,
|
| 1400 |
+
"learning_rate": 2.7857174124171166e-06,
|
| 1401 |
+
"loss": 0.1195,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.4932182490752158,
|
| 1406 |
+
"grad_norm": 0.9344275828949675,
|
| 1407 |
+
"learning_rate": 2.7653890131511446e-06,
|
| 1408 |
+
"loss": 0.1176,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.4956843403205919,
|
| 1413 |
+
"grad_norm": 1.0620538204492203,
|
| 1414 |
+
"learning_rate": 2.7450428508239024e-06,
|
| 1415 |
+
"loss": 0.1271,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.49815043156596794,
|
| 1420 |
+
"grad_norm": 1.082860992255448,
|
| 1421 |
+
"learning_rate": 2.7246802872480876e-06,
|
| 1422 |
+
"loss": 0.1179,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.5006165228113441,
|
| 1427 |
+
"grad_norm": 0.9311025783963606,
|
| 1428 |
+
"learning_rate": 2.704302685334171e-06,
|
| 1429 |
+
"loss": 0.1134,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.5030826140567201,
|
| 1434 |
+
"grad_norm": 1.0144674814097432,
|
| 1435 |
+
"learning_rate": 2.683911408999169e-06,
|
| 1436 |
+
"loss": 0.1302,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.5055487053020962,
|
| 1441 |
+
"grad_norm": 0.984499419053219,
|
| 1442 |
+
"learning_rate": 2.663507823075358e-06,
|
| 1443 |
+
"loss": 0.1302,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.5080147965474723,
|
| 1448 |
+
"grad_norm": 0.9285583857769201,
|
| 1449 |
+
"learning_rate": 2.6430932932189223e-06,
|
| 1450 |
+
"loss": 0.1227,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.5104808877928483,
|
| 1455 |
+
"grad_norm": 1.0473296557535696,
|
| 1456 |
+
"learning_rate": 2.6226691858185454e-06,
|
| 1457 |
+
"loss": 0.13,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.5129469790382244,
|
| 1462 |
+
"grad_norm": 0.9697640283643515,
|
| 1463 |
+
"learning_rate": 2.602236867903959e-06,
|
| 1464 |
+
"loss": 0.1261,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.5154130702836005,
|
| 1469 |
+
"grad_norm": 0.9476280670717342,
|
| 1470 |
+
"learning_rate": 2.5817977070544408e-06,
|
| 1471 |
+
"loss": 0.121,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.5178791615289766,
|
| 1476 |
+
"grad_norm": 0.9186819945539865,
|
| 1477 |
+
"learning_rate": 2.561353071307281e-06,
|
| 1478 |
+
"loss": 0.1218,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.5203452527743526,
|
| 1483 |
+
"grad_norm": 0.9577733461167965,
|
| 1484 |
+
"learning_rate": 2.5409043290662173e-06,
|
| 1485 |
+
"loss": 0.1306,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.5228113440197287,
|
| 1490 |
+
"grad_norm": 1.0157228040259194,
|
| 1491 |
+
"learning_rate": 2.5204528490098436e-06,
|
| 1492 |
+
"loss": 0.13,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.5252774352651048,
|
| 1497 |
+
"grad_norm": 0.9321701271803269,
|
| 1498 |
+
"learning_rate": 2.5e-06,
|
| 1499 |
+
"loss": 0.1153,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.5277435265104808,
|
| 1504 |
+
"grad_norm": 0.9132233244869395,
|
| 1505 |
+
"learning_rate": 2.479547150990158e-06,
|
| 1506 |
+
"loss": 0.1108,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.530209617755857,
|
| 1511 |
+
"grad_norm": 0.9437953862136116,
|
| 1512 |
+
"learning_rate": 2.459095670933783e-06,
|
| 1513 |
+
"loss": 0.1196,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.5326757090012331,
|
| 1518 |
+
"grad_norm": 0.9403892441132017,
|
| 1519 |
+
"learning_rate": 2.43864692869272e-06,
|
| 1520 |
+
"loss": 0.1147,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.5351418002466092,
|
| 1525 |
+
"grad_norm": 0.9172193992658316,
|
| 1526 |
+
"learning_rate": 2.41820229294556e-06,
|
| 1527 |
+
"loss": 0.121,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.5376078914919852,
|
| 1532 |
+
"grad_norm": 0.9510458577076487,
|
| 1533 |
+
"learning_rate": 2.397763132096042e-06,
|
| 1534 |
+
"loss": 0.1162,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.5400739827373613,
|
| 1539 |
+
"grad_norm": 0.8990985185553175,
|
| 1540 |
+
"learning_rate": 2.377330814181455e-06,
|
| 1541 |
+
"loss": 0.1156,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.5425400739827374,
|
| 1546 |
+
"grad_norm": 0.9237885692961808,
|
| 1547 |
+
"learning_rate": 2.356906706781079e-06,
|
| 1548 |
+
"loss": 0.1181,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.5450061652281134,
|
| 1553 |
+
"grad_norm": 0.9652891485182359,
|
| 1554 |
+
"learning_rate": 2.3364921769246423e-06,
|
| 1555 |
+
"loss": 0.1202,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.5474722564734895,
|
| 1560 |
+
"grad_norm": 0.9656721335460944,
|
| 1561 |
+
"learning_rate": 2.3160885910008317e-06,
|
| 1562 |
+
"loss": 0.1185,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.5499383477188656,
|
| 1567 |
+
"grad_norm": 0.935775314239,
|
| 1568 |
+
"learning_rate": 2.2956973146658303e-06,
|
| 1569 |
+
"loss": 0.1246,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.5524044389642416,
|
| 1574 |
+
"grad_norm": 0.9151722771552518,
|
| 1575 |
+
"learning_rate": 2.275319712751913e-06,
|
| 1576 |
+
"loss": 0.1198,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.5548705302096177,
|
| 1581 |
+
"grad_norm": 0.862329148584726,
|
| 1582 |
+
"learning_rate": 2.2549571491760985e-06,
|
| 1583 |
+
"loss": 0.1117,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.5573366214549939,
|
| 1588 |
+
"grad_norm": 0.8864820454656598,
|
| 1589 |
+
"learning_rate": 2.234610986848857e-06,
|
| 1590 |
+
"loss": 0.1144,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.55980271270037,
|
| 1595 |
+
"grad_norm": 0.9173936154868577,
|
| 1596 |
+
"learning_rate": 2.2142825875828838e-06,
|
| 1597 |
+
"loss": 0.1145,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.562268803945746,
|
| 1602 |
+
"grad_norm": 0.9100026032735993,
|
| 1603 |
+
"learning_rate": 2.19397331200196e-06,
|
| 1604 |
+
"loss": 0.1228,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.5647348951911221,
|
| 1609 |
+
"grad_norm": 0.888826708665475,
|
| 1610 |
+
"learning_rate": 2.173684519449872e-06,
|
| 1611 |
+
"loss": 0.126,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.5672009864364982,
|
| 1616 |
+
"grad_norm": 0.9435384216190771,
|
| 1617 |
+
"learning_rate": 2.1534175678994332e-06,
|
| 1618 |
+
"loss": 0.1181,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.5696670776818742,
|
| 1623 |
+
"grad_norm": 0.8799070903740592,
|
| 1624 |
+
"learning_rate": 2.133173813861596e-06,
|
| 1625 |
+
"loss": 0.1115,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.5721331689272503,
|
| 1630 |
+
"grad_norm": 0.909561387709095,
|
| 1631 |
+
"learning_rate": 2.112954612294648e-06,
|
| 1632 |
+
"loss": 0.1171,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.5745992601726264,
|
| 1637 |
+
"grad_norm": 0.9058913889393377,
|
| 1638 |
+
"learning_rate": 2.0927613165135285e-06,
|
| 1639 |
+
"loss": 0.1063,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.5770653514180024,
|
| 1644 |
+
"grad_norm": 0.8718739640553517,
|
| 1645 |
+
"learning_rate": 2.072595278099247e-06,
|
| 1646 |
+
"loss": 0.1115,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.5795314426633785,
|
| 1651 |
+
"grad_norm": 0.9510179452907489,
|
| 1652 |
+
"learning_rate": 2.052457846808419e-06,
|
| 1653 |
+
"loss": 0.1306,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.5819975339087546,
|
| 1658 |
+
"grad_norm": 0.8493349373441297,
|
| 1659 |
+
"learning_rate": 2.03235037048292e-06,
|
| 1660 |
+
"loss": 0.1127,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.5844636251541308,
|
| 1665 |
+
"grad_norm": 0.9640524713085304,
|
| 1666 |
+
"learning_rate": 2.01227419495968e-06,
|
| 1667 |
+
"loss": 0.1159,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.5869297163995068,
|
| 1672 |
+
"grad_norm": 0.9143333804952902,
|
| 1673 |
+
"learning_rate": 1.9922306639805936e-06,
|
| 1674 |
+
"loss": 0.1182,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.5893958076448829,
|
| 1679 |
+
"grad_norm": 0.9340044615462166,
|
| 1680 |
+
"learning_rate": 1.972221119102587e-06,
|
| 1681 |
+
"loss": 0.1281,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.591861898890259,
|
| 1686 |
+
"grad_norm": 0.9381762414200482,
|
| 1687 |
+
"learning_rate": 1.952246899607826e-06,
|
| 1688 |
+
"loss": 0.112,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.594327990135635,
|
| 1693 |
+
"grad_norm": 0.9389267690841027,
|
| 1694 |
+
"learning_rate": 1.9323093424140673e-06,
|
| 1695 |
+
"loss": 0.121,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.5967940813810111,
|
| 1700 |
+
"grad_norm": 0.9082835200250261,
|
| 1701 |
+
"learning_rate": 1.9124097819851835e-06,
|
| 1702 |
+
"loss": 0.1139,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.5992601726263872,
|
| 1707 |
+
"grad_norm": 0.915785148531625,
|
| 1708 |
+
"learning_rate": 1.8925495502418407e-06,
|
| 1709 |
+
"loss": 0.1158,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.6017262638717632,
|
| 1714 |
+
"grad_norm": 0.9260002125597431,
|
| 1715 |
+
"learning_rate": 1.8727299764723507e-06,
|
| 1716 |
+
"loss": 0.1074,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.6041923551171393,
|
| 1721 |
+
"grad_norm": 0.8788204067020321,
|
| 1722 |
+
"learning_rate": 1.852952387243698e-06,
|
| 1723 |
+
"loss": 0.117,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.6066584463625154,
|
| 1728 |
+
"grad_norm": 0.997647923013629,
|
| 1729 |
+
"learning_rate": 1.8332181063127543e-06,
|
| 1730 |
+
"loss": 0.1211,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.6091245376078915,
|
| 1735 |
+
"grad_norm": 0.9033409199704276,
|
| 1736 |
+
"learning_rate": 1.81352845453767e-06,
|
| 1737 |
+
"loss": 0.1149,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.6115906288532675,
|
| 1742 |
+
"grad_norm": 1.1668894281909206,
|
| 1743 |
+
"learning_rate": 1.793884749789469e-06,
|
| 1744 |
+
"loss": 0.1261,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.6140567200986436,
|
| 1749 |
+
"grad_norm": 0.9130406639781518,
|
| 1750 |
+
"learning_rate": 1.7742883068638447e-06,
|
| 1751 |
+
"loss": 0.1174,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.6165228113440198,
|
| 1756 |
+
"grad_norm": 0.8931359516774195,
|
| 1757 |
+
"learning_rate": 1.754740437393151e-06,
|
| 1758 |
+
"loss": 0.1193,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.6189889025893958,
|
| 1763 |
+
"grad_norm": 1.0092939628316673,
|
| 1764 |
+
"learning_rate": 1.7352424497586163e-06,
|
| 1765 |
+
"loss": 0.1168,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.6214549938347719,
|
| 1770 |
+
"grad_norm": 0.9143194831704523,
|
| 1771 |
+
"learning_rate": 1.7157956490027716e-06,
|
| 1772 |
+
"loss": 0.1196,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.623921085080148,
|
| 1777 |
+
"grad_norm": 0.8843231634252613,
|
| 1778 |
+
"learning_rate": 1.6964013367420967e-06,
|
| 1779 |
+
"loss": 0.1061,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.6263871763255241,
|
| 1784 |
+
"grad_norm": 0.9473360944695854,
|
| 1785 |
+
"learning_rate": 1.6770608110799028e-06,
|
| 1786 |
+
"loss": 0.113,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.6288532675709001,
|
| 1791 |
+
"grad_norm": 0.9079034851376371,
|
| 1792 |
+
"learning_rate": 1.6577753665194502e-06,
|
| 1793 |
+
"loss": 0.1168,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.6313193588162762,
|
| 1798 |
+
"grad_norm": 0.8439649713802054,
|
| 1799 |
+
"learning_rate": 1.638546293877299e-06,
|
| 1800 |
+
"loss": 0.1153,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.6337854500616523,
|
| 1805 |
+
"grad_norm": 0.8962051447415076,
|
| 1806 |
+
"learning_rate": 1.6193748801969164e-06,
|
| 1807 |
+
"loss": 0.1137,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.6362515413070283,
|
| 1812 |
+
"grad_norm": 0.9120329709479187,
|
| 1813 |
+
"learning_rate": 1.6002624086625296e-06,
|
| 1814 |
+
"loss": 0.1207,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.6387176325524044,
|
| 1819 |
+
"grad_norm": 0.9770199026936922,
|
| 1820 |
+
"learning_rate": 1.5812101585132416e-06,
|
| 1821 |
+
"loss": 0.1281,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.6411837237977805,
|
| 1826 |
+
"grad_norm": 0.8835447672842621,
|
| 1827 |
+
"learning_rate": 1.5622194049574049e-06,
|
| 1828 |
+
"loss": 0.105,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.6436498150431565,
|
| 1833 |
+
"grad_norm": 0.9176836023087112,
|
| 1834 |
+
"learning_rate": 1.5432914190872757e-06,
|
| 1835 |
+
"loss": 0.1057,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.6461159062885327,
|
| 1840 |
+
"grad_norm": 0.9155694840686882,
|
| 1841 |
+
"learning_rate": 1.5244274677939307e-06,
|
| 1842 |
+
"loss": 0.1142,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.6485819975339088,
|
| 1847 |
+
"grad_norm": 0.8750355437796338,
|
| 1848 |
+
"learning_rate": 1.5056288136824725e-06,
|
| 1849 |
+
"loss": 0.1097,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.6510480887792849,
|
| 1854 |
+
"grad_norm": 0.920968579960804,
|
| 1855 |
+
"learning_rate": 1.4868967149875257e-06,
|
| 1856 |
+
"loss": 0.1128,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.6535141800246609,
|
| 1861 |
+
"grad_norm": 0.8967167103024075,
|
| 1862 |
+
"learning_rate": 1.4682324254890135e-06,
|
| 1863 |
+
"loss": 0.1105,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.655980271270037,
|
| 1868 |
+
"grad_norm": 0.8718837596518669,
|
| 1869 |
+
"learning_rate": 1.4496371944282443e-06,
|
| 1870 |
+
"loss": 0.1104,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.6584463625154131,
|
| 1875 |
+
"grad_norm": 0.9025049741344529,
|
| 1876 |
+
"learning_rate": 1.4311122664242955e-06,
|
| 1877 |
+
"loss": 0.116,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.6609124537607891,
|
| 1882 |
+
"grad_norm": 0.9843578797393775,
|
| 1883 |
+
"learning_rate": 1.4126588813907071e-06,
|
| 1884 |
+
"loss": 0.1113,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.6633785450061652,
|
| 1889 |
+
"grad_norm": 0.8724892376206785,
|
| 1890 |
+
"learning_rate": 1.3942782744524974e-06,
|
| 1891 |
+
"loss": 0.1149,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.6658446362515413,
|
| 1896 |
+
"grad_norm": 0.8721049500966537,
|
| 1897 |
+
"learning_rate": 1.3759716758634833e-06,
|
| 1898 |
+
"loss": 0.1153,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.6683107274969173,
|
| 1903 |
+
"grad_norm": 0.8726658460813645,
|
| 1904 |
+
"learning_rate": 1.3577403109239485e-06,
|
| 1905 |
+
"loss": 0.107,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.6707768187422934,
|
| 1910 |
+
"grad_norm": 0.9098670633340208,
|
| 1911 |
+
"learning_rate": 1.3395853998986208e-06,
|
| 1912 |
+
"loss": 0.1128,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.6732429099876696,
|
| 1917 |
+
"grad_norm": 0.9039753869528919,
|
| 1918 |
+
"learning_rate": 1.3215081579350058e-06,
|
| 1919 |
+
"loss": 0.1103,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.6757090012330457,
|
| 1924 |
+
"grad_norm": 0.989127513896241,
|
| 1925 |
+
"learning_rate": 1.3035097949820486e-06,
|
| 1926 |
+
"loss": 0.104,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.6781750924784217,
|
| 1931 |
+
"grad_norm": 0.877571530399145,
|
| 1932 |
+
"learning_rate": 1.2855915157091498e-06,
|
| 1933 |
+
"loss": 0.1219,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.6806411837237978,
|
| 1938 |
+
"grad_norm": 0.82634562341831,
|
| 1939 |
+
"learning_rate": 1.2677545194255403e-06,
|
| 1940 |
+
"loss": 0.1082,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.6831072749691739,
|
| 1945 |
+
"grad_norm": 0.8972600186401277,
|
| 1946 |
+
"learning_rate": 1.2500000000000007e-06,
|
| 1947 |
+
"loss": 0.0987,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.6855733662145499,
|
| 1952 |
+
"grad_norm": 0.8356440290676206,
|
| 1953 |
+
"learning_rate": 1.2323291457809574e-06,
|
| 1954 |
+
"loss": 0.1052,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.688039457459926,
|
| 1959 |
+
"grad_norm": 0.9368856966850118,
|
| 1960 |
+
"learning_rate": 1.214743139516946e-06,
|
| 1961 |
+
"loss": 0.1177,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.6905055487053021,
|
| 1966 |
+
"grad_norm": 0.914022985817142,
|
| 1967 |
+
"learning_rate": 1.197243158277442e-06,
|
| 1968 |
+
"loss": 0.1154,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.6929716399506781,
|
| 1973 |
+
"grad_norm": 0.8830620774383795,
|
| 1974 |
+
"learning_rate": 1.1798303733740801e-06,
|
| 1975 |
+
"loss": 0.1074,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.6954377311960542,
|
| 1980 |
+
"grad_norm": 0.9859448834926087,
|
| 1981 |
+
"learning_rate": 1.1625059502822575e-06,
|
| 1982 |
+
"loss": 0.1173,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.6979038224414303,
|
| 1987 |
+
"grad_norm": 0.9089769946191705,
|
| 1988 |
+
"learning_rate": 1.1452710485631216e-06,
|
| 1989 |
+
"loss": 0.1173,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.7003699136868065,
|
| 1994 |
+
"grad_norm": 0.8028092228945499,
|
| 1995 |
+
"learning_rate": 1.1281268217859606e-06,
|
| 1996 |
+
"loss": 0.097,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.7028360049321825,
|
| 2001 |
+
"grad_norm": 0.9277228824252272,
|
| 2002 |
+
"learning_rate": 1.1110744174509952e-06,
|
| 2003 |
+
"loss": 0.1116,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.7053020961775586,
|
| 2008 |
+
"grad_norm": 0.9117211175426764,
|
| 2009 |
+
"learning_rate": 1.0941149769125676e-06,
|
| 2010 |
+
"loss": 0.1055,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.7077681874229347,
|
| 2015 |
+
"grad_norm": 0.839627290488779,
|
| 2016 |
+
"learning_rate": 1.077249635302754e-06,
|
| 2017 |
+
"loss": 0.1082,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.7102342786683107,
|
| 2022 |
+
"grad_norm": 0.8255715103747457,
|
| 2023 |
+
"learning_rate": 1.0604795214553867e-06,
|
| 2024 |
+
"loss": 0.1098,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.7127003699136868,
|
| 2029 |
+
"grad_norm": 0.8647976705807752,
|
| 2030 |
+
"learning_rate": 1.043805757830495e-06,
|
| 2031 |
+
"loss": 0.1062,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.7151664611590629,
|
| 2036 |
+
"grad_norm": 0.9718746331555166,
|
| 2037 |
+
"learning_rate": 1.0272294604391792e-06,
|
| 2038 |
+
"loss": 0.1258,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.717632552404439,
|
| 2043 |
+
"grad_norm": 0.9126439380715431,
|
| 2044 |
+
"learning_rate": 1.0107517387689168e-06,
|
| 2045 |
+
"loss": 0.1123,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.720098643649815,
|
| 2050 |
+
"grad_norm": 0.8990104000241568,
|
| 2051 |
+
"learning_rate": 9.943736957092956e-07,
|
| 2052 |
+
"loss": 0.1129,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.7225647348951911,
|
| 2057 |
+
"grad_norm": 0.864712623403445,
|
| 2058 |
+
"learning_rate": 9.780964274781984e-07,
|
| 2059 |
+
"loss": 0.1081,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.7250308261405672,
|
| 2064 |
+
"grad_norm": 0.8399371395573897,
|
| 2065 |
+
"learning_rate": 9.619210235484333e-07,
|
| 2066 |
+
"loss": 0.1117,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.7274969173859432,
|
| 2071 |
+
"grad_norm": 0.8825066510711791,
|
| 2072 |
+
"learning_rate": 9.458485665748071e-07,
|
| 2073 |
+
"loss": 0.1118,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.7299630086313194,
|
| 2078 |
+
"grad_norm": 0.8668349547032418,
|
| 2079 |
+
"learning_rate": 9.298801323216647e-07,
|
| 2080 |
+
"loss": 0.1093,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.7324290998766955,
|
| 2085 |
+
"grad_norm": 0.82118592126248,
|
| 2086 |
+
"learning_rate": 9.140167895908867e-07,
|
| 2087 |
+
"loss": 0.107,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.7348951911220715,
|
| 2092 |
+
"grad_norm": 0.7927350936352466,
|
| 2093 |
+
"learning_rate": 8.982596001503488e-07,
|
| 2094 |
+
"loss": 0.1051,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.7373612823674476,
|
| 2099 |
+
"grad_norm": 0.8064491310959427,
|
| 2100 |
+
"learning_rate": 8.826096186628568e-07,
|
| 2101 |
+
"loss": 0.1026,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.7398273736128237,
|
| 2106 |
+
"grad_norm": 0.8263765829534724,
|
| 2107 |
+
"learning_rate": 8.670678926155588e-07,
|
| 2108 |
+
"loss": 0.0986,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.7422934648581998,
|
| 2113 |
+
"grad_norm": 0.8675207524067637,
|
| 2114 |
+
"learning_rate": 8.516354622498279e-07,
|
| 2115 |
+
"loss": 0.114,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.7447595561035758,
|
| 2120 |
+
"grad_norm": 0.8687561824637725,
|
| 2121 |
+
"learning_rate": 8.363133604916448e-07,
|
| 2122 |
+
"loss": 0.1076,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.7472256473489519,
|
| 2127 |
+
"grad_norm": 0.9924452227470472,
|
| 2128 |
+
"learning_rate": 8.21102612882454e-07,
|
| 2129 |
+
"loss": 0.1201,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.749691738594328,
|
| 2134 |
+
"grad_norm": 0.9072442283272004,
|
| 2135 |
+
"learning_rate": 8.060042375105285e-07,
|
| 2136 |
+
"loss": 0.1142,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.752157829839704,
|
| 2141 |
+
"grad_norm": 0.8431769690914566,
|
| 2142 |
+
"learning_rate": 7.910192449428216e-07,
|
| 2143 |
+
"loss": 0.1079,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.7546239210850801,
|
| 2148 |
+
"grad_norm": 0.8839768853643986,
|
| 2149 |
+
"learning_rate": 7.761486381573327e-07,
|
| 2150 |
+
"loss": 0.1095,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.7570900123304563,
|
| 2155 |
+
"grad_norm": 0.8290765629607486,
|
| 2156 |
+
"learning_rate": 7.613934124759712e-07,
|
| 2157 |
+
"loss": 0.1039,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.7595561035758323,
|
| 2162 |
+
"grad_norm": 0.8607203888248921,
|
| 2163 |
+
"learning_rate": 7.467545554979391e-07,
|
| 2164 |
+
"loss": 0.0985,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.7620221948212084,
|
| 2169 |
+
"grad_norm": 0.9599192361218681,
|
| 2170 |
+
"learning_rate": 7.322330470336314e-07,
|
| 2171 |
+
"loss": 0.109,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.7644882860665845,
|
| 2176 |
+
"grad_norm": 0.9002485432063522,
|
| 2177 |
+
"learning_rate": 7.178298590390509e-07,
|
| 2178 |
+
"loss": 0.1153,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.7669543773119606,
|
| 2183 |
+
"grad_norm": 0.8784168588017048,
|
| 2184 |
+
"learning_rate": 7.035459555507549e-07,
|
| 2185 |
+
"loss": 0.1115,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.7694204685573366,
|
| 2190 |
+
"grad_norm": 0.8382036560825521,
|
| 2191 |
+
"learning_rate": 6.893822926213331e-07,
|
| 2192 |
+
"loss": 0.1041,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.7718865598027127,
|
| 2197 |
+
"grad_norm": 0.9076876818793927,
|
| 2198 |
+
"learning_rate": 6.753398182554116e-07,
|
| 2199 |
+
"loss": 0.1081,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.7743526510480888,
|
| 2204 |
+
"grad_norm": 0.846644134454287,
|
| 2205 |
+
"learning_rate": 6.614194723462036e-07,
|
| 2206 |
+
"loss": 0.102,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.7768187422934648,
|
| 2211 |
+
"grad_norm": 0.9289192311205007,
|
| 2212 |
+
"learning_rate": 6.476221866126029e-07,
|
| 2213 |
+
"loss": 0.0963,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.7792848335388409,
|
| 2218 |
+
"grad_norm": 0.8420031544643374,
|
| 2219 |
+
"learning_rate": 6.339488845368156e-07,
|
| 2220 |
+
"loss": 0.1005,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.781750924784217,
|
| 2225 |
+
"grad_norm": 0.9485219609480249,
|
| 2226 |
+
"learning_rate": 6.204004813025569e-07,
|
| 2227 |
+
"loss": 0.1084,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.7842170160295932,
|
| 2232 |
+
"grad_norm": 0.8483090842060725,
|
| 2233 |
+
"learning_rate": 6.069778837337887e-07,
|
| 2234 |
+
"loss": 0.1104,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.7866831072749692,
|
| 2239 |
+
"grad_norm": 0.8562978222611917,
|
| 2240 |
+
"learning_rate": 5.936819902340299e-07,
|
| 2241 |
+
"loss": 0.1106,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.7891491985203453,
|
| 2246 |
+
"grad_norm": 0.8779656782917099,
|
| 2247 |
+
"learning_rate": 5.805136907262199e-07,
|
| 2248 |
+
"loss": 0.1049,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.7916152897657214,
|
| 2253 |
+
"grad_norm": 0.8218033180821755,
|
| 2254 |
+
"learning_rate": 5.674738665931575e-07,
|
| 2255 |
+
"loss": 0.0964,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.7940813810110974,
|
| 2260 |
+
"grad_norm": 0.9049032540417703,
|
| 2261 |
+
"learning_rate": 5.545633906185058e-07,
|
| 2262 |
+
"loss": 0.1032,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.7965474722564735,
|
| 2267 |
+
"grad_norm": 0.8814274256802208,
|
| 2268 |
+
"learning_rate": 5.417831269283741e-07,
|
| 2269 |
+
"loss": 0.1102,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.7990135635018496,
|
| 2274 |
+
"grad_norm": 0.9220220138008265,
|
| 2275 |
+
"learning_rate": 5.291339309334845e-07,
|
| 2276 |
+
"loss": 0.1066,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.8014796547472256,
|
| 2281 |
+
"grad_norm": 0.8484359082657426,
|
| 2282 |
+
"learning_rate": 5.166166492719124e-07,
|
| 2283 |
+
"loss": 0.1037,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.8039457459926017,
|
| 2288 |
+
"grad_norm": 0.841377944243339,
|
| 2289 |
+
"learning_rate": 5.042321197524214e-07,
|
| 2290 |
+
"loss": 0.1138,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.8064118372379778,
|
| 2295 |
+
"grad_norm": 0.8472048440145639,
|
| 2296 |
+
"learning_rate": 4.91981171298388e-07,
|
| 2297 |
+
"loss": 0.0971,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.8088779284833539,
|
| 2302 |
+
"grad_norm": 0.8772670558334653,
|
| 2303 |
+
"learning_rate": 4.798646238923177e-07,
|
| 2304 |
+
"loss": 0.1086,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.8113440197287299,
|
| 2309 |
+
"grad_norm": 0.8587072849352904,
|
| 2310 |
+
"learning_rate": 4.678832885209622e-07,
|
| 2311 |
+
"loss": 0.1043,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.813810110974106,
|
| 2316 |
+
"grad_norm": 0.9773397231933723,
|
| 2317 |
+
"learning_rate": 4.5603796712104104e-07,
|
| 2318 |
+
"loss": 0.1121,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.8162762022194822,
|
| 2323 |
+
"grad_norm": 0.8752855199312541,
|
| 2324 |
+
"learning_rate": 4.4432945252556285e-07,
|
| 2325 |
+
"loss": 0.1106,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.8187422934648582,
|
| 2330 |
+
"grad_norm": 0.8751816569743863,
|
| 2331 |
+
"learning_rate": 4.3275852841075965e-07,
|
| 2332 |
+
"loss": 0.0998,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.8212083847102343,
|
| 2337 |
+
"grad_norm": 0.8405865171479526,
|
| 2338 |
+
"learning_rate": 4.2132596924363666e-07,
|
| 2339 |
+
"loss": 0.097,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.8236744759556104,
|
| 2344 |
+
"grad_norm": 0.8593779024468443,
|
| 2345 |
+
"learning_rate": 4.100325402301333e-07,
|
| 2346 |
+
"loss": 0.1056,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.8261405672009864,
|
| 2351 |
+
"grad_norm": 0.8506369591795581,
|
| 2352 |
+
"learning_rate": 3.9887899726390455e-07,
|
| 2353 |
+
"loss": 0.1011,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.8286066584463625,
|
| 2358 |
+
"grad_norm": 0.8736867525942835,
|
| 2359 |
+
"learning_rate": 3.8786608687573226e-07,
|
| 2360 |
+
"loss": 0.1087,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.8310727496917386,
|
| 2365 |
+
"grad_norm": 0.8790912577375488,
|
| 2366 |
+
"learning_rate": 3.7699454618355306e-07,
|
| 2367 |
+
"loss": 0.107,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.8335388409371147,
|
| 2372 |
+
"grad_norm": 0.8862208021029517,
|
| 2373 |
+
"learning_rate": 3.662651028431238e-07,
|
| 2374 |
+
"loss": 0.1043,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.8360049321824907,
|
| 2379 |
+
"grad_norm": 0.8886677625295568,
|
| 2380 |
+
"learning_rate": 3.5567847499932004e-07,
|
| 2381 |
+
"loss": 0.109,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.8384710234278668,
|
| 2386 |
+
"grad_norm": 0.8417309037752911,
|
| 2387 |
+
"learning_rate": 3.452353712380646e-07,
|
| 2388 |
+
"loss": 0.111,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.840937114673243,
|
| 2393 |
+
"grad_norm": 0.8245160824964732,
|
| 2394 |
+
"learning_rate": 3.3493649053890325e-07,
|
| 2395 |
+
"loss": 0.109,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.843403205918619,
|
| 2400 |
+
"grad_norm": 0.8810761627851151,
|
| 2401 |
+
"learning_rate": 3.2478252222822166e-07,
|
| 2402 |
+
"loss": 0.1095,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.8458692971639951,
|
| 2407 |
+
"grad_norm": 0.829093497353401,
|
| 2408 |
+
"learning_rate": 3.147741459331033e-07,
|
| 2409 |
+
"loss": 0.1089,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.8483353884093712,
|
| 2414 |
+
"grad_norm": 0.8881173215737831,
|
| 2415 |
+
"learning_rate": 3.0491203153584337e-07,
|
| 2416 |
+
"loss": 0.0983,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.8508014796547472,
|
| 2421 |
+
"grad_norm": 0.8385259634302925,
|
| 2422 |
+
"learning_rate": 2.9519683912911267e-07,
|
| 2423 |
+
"loss": 0.1035,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.8532675709001233,
|
| 2428 |
+
"grad_norm": 0.888236041089151,
|
| 2429 |
+
"learning_rate": 2.856292189717741e-07,
|
| 2430 |
+
"loss": 0.1062,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.8557336621454994,
|
| 2435 |
+
"grad_norm": 0.8042541246282294,
|
| 2436 |
+
"learning_rate": 2.7620981144536014e-07,
|
| 2437 |
+
"loss": 0.1092,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.8581997533908755,
|
| 2442 |
+
"grad_norm": 0.8411923992042479,
|
| 2443 |
+
"learning_rate": 2.6693924701121196e-07,
|
| 2444 |
+
"loss": 0.1071,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.8606658446362515,
|
| 2449 |
+
"grad_norm": 0.8605859724185484,
|
| 2450 |
+
"learning_rate": 2.5781814616827936e-07,
|
| 2451 |
+
"loss": 0.1057,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.8631319358816276,
|
| 2456 |
+
"grad_norm": 0.8437134740054346,
|
| 2457 |
+
"learning_rate": 2.488471194115913e-07,
|
| 2458 |
+
"loss": 0.1029,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.8655980271270037,
|
| 2463 |
+
"grad_norm": 0.8593830934685845,
|
| 2464 |
+
"learning_rate": 2.400267671913917e-07,
|
| 2465 |
+
"loss": 0.1222,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.8680641183723797,
|
| 2470 |
+
"grad_norm": 0.891969275564653,
|
| 2471 |
+
"learning_rate": 2.313576798729536e-07,
|
| 2472 |
+
"loss": 0.1104,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.8705302096177558,
|
| 2477 |
+
"grad_norm": 0.8046940940350272,
|
| 2478 |
+
"learning_rate": 2.2284043769706026e-07,
|
| 2479 |
+
"loss": 0.1006,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.872996300863132,
|
| 2484 |
+
"grad_norm": 0.8634757581988033,
|
| 2485 |
+
"learning_rate": 2.144756107411733e-07,
|
| 2486 |
+
"loss": 0.0999,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.8754623921085081,
|
| 2491 |
+
"grad_norm": 0.8194977361422426,
|
| 2492 |
+
"learning_rate": 2.0626375888127187e-07,
|
| 2493 |
+
"loss": 0.1015,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.8779284833538841,
|
| 2498 |
+
"grad_norm": 0.8510293589337354,
|
| 2499 |
+
"learning_rate": 1.9820543175438077e-07,
|
| 2500 |
+
"loss": 0.1003,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.8803945745992602,
|
| 2505 |
+
"grad_norm": 0.8619662818910347,
|
| 2506 |
+
"learning_rate": 1.9030116872178317e-07,
|
| 2507 |
+
"loss": 0.1062,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.8828606658446363,
|
| 2512 |
+
"grad_norm": 0.906303197576803,
|
| 2513 |
+
"learning_rate": 1.82551498832918e-07,
|
| 2514 |
+
"loss": 0.11,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.8853267570900123,
|
| 2519 |
+
"grad_norm": 0.8691559587313747,
|
| 2520 |
+
"learning_rate": 1.7495694078996984e-07,
|
| 2521 |
+
"loss": 0.1061,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.8877928483353884,
|
| 2526 |
+
"grad_norm": 0.9322059801564989,
|
| 2527 |
+
"learning_rate": 1.675180029131529e-07,
|
| 2528 |
+
"loss": 0.1077,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.8902589395807645,
|
| 2533 |
+
"grad_norm": 0.9575331389552335,
|
| 2534 |
+
"learning_rate": 1.602351831066862e-07,
|
| 2535 |
+
"loss": 0.1055,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.8927250308261405,
|
| 2540 |
+
"grad_norm": 0.8858531694804256,
|
| 2541 |
+
"learning_rate": 1.531089688254672e-07,
|
| 2542 |
+
"loss": 0.1027,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.8951911220715166,
|
| 2547 |
+
"grad_norm": 0.8344537690883024,
|
| 2548 |
+
"learning_rate": 1.4613983704244826e-07,
|
| 2549 |
+
"loss": 0.1047,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.8976572133168927,
|
| 2554 |
+
"grad_norm": 0.8633024738476377,
|
| 2555 |
+
"learning_rate": 1.3932825421670876e-07,
|
| 2556 |
+
"loss": 0.1126,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.9001233045622689,
|
| 2561 |
+
"grad_norm": 0.876419400046166,
|
| 2562 |
+
"learning_rate": 1.3267467626223606e-07,
|
| 2563 |
+
"loss": 0.1044,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.9025893958076449,
|
| 2568 |
+
"grad_norm": 0.8775669698848996,
|
| 2569 |
+
"learning_rate": 1.261795485174083e-07,
|
| 2570 |
+
"loss": 0.1058,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.905055487053021,
|
| 2575 |
+
"grad_norm": 0.9370168003338669,
|
| 2576 |
+
"learning_rate": 1.1984330571518932e-07,
|
| 2577 |
+
"loss": 0.107,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.9075215782983971,
|
| 2582 |
+
"grad_norm": 0.8317921963443387,
|
| 2583 |
+
"learning_rate": 1.1366637195402875e-07,
|
| 2584 |
+
"loss": 0.1031,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.9099876695437731,
|
| 2589 |
+
"grad_norm": 0.8075257461164578,
|
| 2590 |
+
"learning_rate": 1.0764916066947795e-07,
|
| 2591 |
+
"loss": 0.1014,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.9124537607891492,
|
| 2596 |
+
"grad_norm": 0.8630691732068891,
|
| 2597 |
+
"learning_rate": 1.0179207460651596e-07,
|
| 2598 |
+
"loss": 0.1054,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.9149198520345253,
|
| 2603 |
+
"grad_norm": 0.8961350606343792,
|
| 2604 |
+
"learning_rate": 9.609550579259497e-08,
|
| 2605 |
+
"loss": 0.1124,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.9173859432799013,
|
| 2610 |
+
"grad_norm": 0.8210699330869444,
|
| 2611 |
+
"learning_rate": 9.055983551140041e-08,
|
| 2612 |
+
"loss": 0.0983,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.9198520345252774,
|
| 2617 |
+
"grad_norm": 0.8084461303198168,
|
| 2618 |
+
"learning_rate": 8.518543427732951e-08,
|
| 2619 |
+
"loss": 0.0942,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.9223181257706535,
|
| 2624 |
+
"grad_norm": 0.8418059270716391,
|
| 2625 |
+
"learning_rate": 7.99726618106933e-08,
|
| 2626 |
+
"loss": 0.1126,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.9247842170160296,
|
| 2631 |
+
"grad_norm": 0.7939196515199246,
|
| 2632 |
+
"learning_rate": 7.492186701364007e-08,
|
| 2633 |
+
"loss": 0.0997,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.9272503082614056,
|
| 2638 |
+
"grad_norm": 0.8000181224351651,
|
| 2639 |
+
"learning_rate": 7.003338794680153e-08,
|
| 2640 |
+
"loss": 0.1008,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.9297163995067818,
|
| 2645 |
+
"grad_norm": 0.8198821503882819,
|
| 2646 |
+
"learning_rate": 6.530755180666593e-08,
|
| 2647 |
+
"loss": 0.1044,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.9321824907521579,
|
| 2652 |
+
"grad_norm": 0.8514419848080834,
|
| 2653 |
+
"learning_rate": 6.074467490367886e-08,
|
| 2654 |
+
"loss": 0.1055,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.9346485819975339,
|
| 2659 |
+
"grad_norm": 0.8316812950570545,
|
| 2660 |
+
"learning_rate": 5.634506264107054e-08,
|
| 2661 |
+
"loss": 0.1012,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.93711467324291,
|
| 2666 |
+
"grad_norm": 0.8548476764668638,
|
| 2667 |
+
"learning_rate": 5.210900949441572e-08,
|
| 2668 |
+
"loss": 0.1146,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.9395807644882861,
|
| 2673 |
+
"grad_norm": 0.8145881326701868,
|
| 2674 |
+
"learning_rate": 4.8036798991923925e-08,
|
| 2675 |
+
"loss": 0.1055,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.9420468557336621,
|
| 2680 |
+
"grad_norm": 0.929041595564495,
|
| 2681 |
+
"learning_rate": 4.412870369546074e-08,
|
| 2682 |
+
"loss": 0.1119,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.9445129469790382,
|
| 2687 |
+
"grad_norm": 0.8695519382748947,
|
| 2688 |
+
"learning_rate": 4.038498518230627e-08,
|
| 2689 |
+
"loss": 0.1011,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.9469790382244143,
|
| 2694 |
+
"grad_norm": 0.8087903294955203,
|
| 2695 |
+
"learning_rate": 3.680589402764695e-08,
|
| 2696 |
+
"loss": 0.1078,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.9494451294697904,
|
| 2701 |
+
"grad_norm": 0.8319698621940769,
|
| 2702 |
+
"learning_rate": 3.339166978780256e-08,
|
| 2703 |
+
"loss": 0.1034,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.9519112207151664,
|
| 2708 |
+
"grad_norm": 0.8441945180499253,
|
| 2709 |
+
"learning_rate": 3.014254098419406e-08,
|
| 2710 |
+
"loss": 0.0989,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.9543773119605425,
|
| 2715 |
+
"grad_norm": 0.8503473881635062,
|
| 2716 |
+
"learning_rate": 2.7058725088047466e-08,
|
| 2717 |
+
"loss": 0.0914,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.9568434032059187,
|
| 2722 |
+
"grad_norm": 0.8072492930440737,
|
| 2723 |
+
"learning_rate": 2.414042850583803e-08,
|
| 2724 |
+
"loss": 0.1073,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.9593094944512947,
|
| 2729 |
+
"grad_norm": 0.8344562468004824,
|
| 2730 |
+
"learning_rate": 2.1387846565474047e-08,
|
| 2731 |
+
"loss": 0.1048,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.9617755856966708,
|
| 2736 |
+
"grad_norm": 0.8285661873749879,
|
| 2737 |
+
"learning_rate": 1.8801163503225083e-08,
|
| 2738 |
+
"loss": 0.1058,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.9642416769420469,
|
| 2743 |
+
"grad_norm": 0.8053362059723653,
|
| 2744 |
+
"learning_rate": 1.6380552451389088e-08,
|
| 2745 |
+
"loss": 0.1,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.966707768187423,
|
| 2750 |
+
"grad_norm": 0.7749438857969168,
|
| 2751 |
+
"learning_rate": 1.412617542670469e-08,
|
| 2752 |
+
"loss": 0.0838,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.969173859432799,
|
| 2757 |
+
"grad_norm": 0.8323595020609726,
|
| 2758 |
+
"learning_rate": 1.2038183319507957e-08,
|
| 2759 |
+
"loss": 0.096,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.9716399506781751,
|
| 2764 |
+
"grad_norm": 0.8837854393760676,
|
| 2765 |
+
"learning_rate": 1.0116715883630724e-08,
|
| 2766 |
+
"loss": 0.1027,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.9741060419235512,
|
| 2771 |
+
"grad_norm": 0.8034770429121565,
|
| 2772 |
+
"learning_rate": 8.361901727049204e-09,
|
| 2773 |
+
"loss": 0.1015,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.9765721331689272,
|
| 2778 |
+
"grad_norm": 0.8123780927603559,
|
| 2779 |
+
"learning_rate": 6.773858303274483e-09,
|
| 2780 |
+
"loss": 0.1016,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.9790382244143033,
|
| 2785 |
+
"grad_norm": 0.8370858433771333,
|
| 2786 |
+
"learning_rate": 5.352691903491303e-09,
|
| 2787 |
+
"loss": 0.0988,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.9815043156596794,
|
| 2792 |
+
"grad_norm": 0.8510284536072052,
|
| 2793 |
+
"learning_rate": 4.098497649443755e-09,
|
| 2794 |
+
"loss": 0.0992,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.9839704069050554,
|
| 2799 |
+
"grad_norm": 0.858653672988882,
|
| 2800 |
+
"learning_rate": 3.011359487068988e-09,
|
| 2801 |
+
"loss": 0.1071,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.9864364981504316,
|
| 2806 |
+
"grad_norm": 0.8155828639676944,
|
| 2807 |
+
"learning_rate": 2.091350180878082e-09,
|
| 2808 |
+
"loss": 0.1023,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.9889025893958077,
|
| 2813 |
+
"grad_norm": 0.8319551975813849,
|
| 2814 |
+
"learning_rate": 1.3385313090857888e-09,
|
| 2815 |
+
"loss": 0.1044,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 0.9913686806411838,
|
| 2820 |
+
"grad_norm": 0.8446197592668653,
|
| 2821 |
+
"learning_rate": 7.529532594893751e-10,
|
| 2822 |
+
"loss": 0.1026,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 0.9938347718865598,
|
| 2827 |
+
"grad_norm": 0.9071881714467116,
|
| 2828 |
+
"learning_rate": 3.346552260954927e-10,
|
| 2829 |
+
"loss": 0.1067,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 0.9963008631319359,
|
| 2834 |
+
"grad_norm": 0.807230030993721,
|
| 2835 |
+
"learning_rate": 8.366520649727427e-11,
|
| 2836 |
+
"loss": 0.1046,
|
| 2837 |
+
"step": 404
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 0.998766954377312,
|
| 2841 |
+
"grad_norm": 0.8671489581176017,
|
| 2842 |
+
"learning_rate": 0.0,
|
| 2843 |
+
"loss": 0.1063,
|
| 2844 |
+
"step": 405
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.998766954377312,
|
| 2848 |
+
"step": 405,
|
| 2849 |
+
"total_flos": 59480064983040.0,
|
| 2850 |
+
"train_loss": 0.13387679483419584,
|
| 2851 |
+
"train_runtime": 2011.9435,
|
| 2852 |
+
"train_samples_per_second": 12.897,
|
| 2853 |
+
"train_steps_per_second": 0.201
|
| 2854 |
+
}
|
| 2855 |
+
],
|
| 2856 |
+
"logging_steps": 1,
|
| 2857 |
+
"max_steps": 405,
|
| 2858 |
+
"num_input_tokens_seen": 0,
|
| 2859 |
+
"num_train_epochs": 1,
|
| 2860 |
+
"save_steps": 500,
|
| 2861 |
+
"stateful_callbacks": {
|
| 2862 |
+
"TrainerControl": {
|
| 2863 |
+
"args": {
|
| 2864 |
+
"should_epoch_stop": false,
|
| 2865 |
+
"should_evaluate": false,
|
| 2866 |
+
"should_log": false,
|
| 2867 |
+
"should_save": false,
|
| 2868 |
+
"should_training_stop": false
|
| 2869 |
+
},
|
| 2870 |
+
"attributes": {}
|
| 2871 |
+
}
|
| 2872 |
+
},
|
| 2873 |
+
"total_flos": 59480064983040.0,
|
| 2874 |
+
"train_batch_size": 1,
|
| 2875 |
+
"trial_name": null,
|
| 2876 |
+
"trial_params": null
|
| 2877 |
+
}
|
training_loss.png
ADDED
|