End of training
Browse files
README.md
CHANGED
|
@@ -1,11 +1,23 @@
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
metrics:
|
| 5 |
- accuracy
|
| 6 |
model-index:
|
| 7 |
- name: smolm-autoreg-bpe-counterfactual_babylm_naans_new-1e-4
|
| 8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -13,7 +25,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# smolm-autoreg-bpe-counterfactual_babylm_naans_new-1e-4
|
| 15 |
|
| 16 |
-
This model was trained from scratch on
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
- Loss: 3.4162
|
| 19 |
- Accuracy: 0.4067
|
|
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
| 4 |
+
datasets:
|
| 5 |
+
- kanishka/counterfactual_babylm_naans_new
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
- name: smolm-autoreg-bpe-counterfactual_babylm_naans_new-1e-4
|
| 10 |
+
results:
|
| 11 |
+
- task:
|
| 12 |
+
name: Causal Language Modeling
|
| 13 |
+
type: text-generation
|
| 14 |
+
dataset:
|
| 15 |
+
name: kanishka/counterfactual_babylm_naans_new
|
| 16 |
+
type: kanishka/counterfactual_babylm_naans_new
|
| 17 |
+
metrics:
|
| 18 |
+
- name: Accuracy
|
| 19 |
+
type: accuracy
|
| 20 |
+
value: 0.4066641958384616
|
| 21 |
---
|
| 22 |
|
| 23 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 25 |
|
| 26 |
# smolm-autoreg-bpe-counterfactual_babylm_naans_new-1e-4
|
| 27 |
|
| 28 |
+
This model was trained from scratch on the kanishka/counterfactual_babylm_naans_new dataset.
|
| 29 |
It achieves the following results on the evaluation set:
|
| 30 |
- Loss: 3.4162
|
| 31 |
- Accuracy: 0.4067
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.4066641958384616,
|
| 4 |
+
"eval_loss": 3.41624116897583,
|
| 5 |
+
"eval_runtime": 153.8779,
|
| 6 |
+
"eval_samples": 57921,
|
| 7 |
+
"eval_samples_per_second": 376.409,
|
| 8 |
+
"eval_steps_per_second": 5.888,
|
| 9 |
+
"perplexity": 30.45472543989185,
|
| 10 |
+
"total_flos": 1.5669257538816e+18,
|
| 11 |
+
"train_loss": 3.1555080074711364,
|
| 12 |
+
"train_runtime": 80953.5697,
|
| 13 |
+
"train_samples": 595035,
|
| 14 |
+
"train_samples_per_second": 147.006,
|
| 15 |
+
"train_steps_per_second": 4.594
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.4066641958384616,
|
| 4 |
+
"eval_loss": 3.41624116897583,
|
| 5 |
+
"eval_runtime": 153.8779,
|
| 6 |
+
"eval_samples": 57921,
|
| 7 |
+
"eval_samples_per_second": 376.409,
|
| 8 |
+
"eval_steps_per_second": 5.888,
|
| 9 |
+
"perplexity": 30.45472543989185
|
| 10 |
+
}
|
runs/Jun04_09-08-08_phyl-ling-p01.la.utexas.edu/events.out.tfevents.1717591598.phyl-ling-p01.la.utexas.edu.147701.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0087a3ba01401b309f8ec4af8df5adb71b28febc09e99934878a156d3de5c629
|
| 3 |
+
size 417
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 1.5669257538816e+18,
|
| 4 |
+
"train_loss": 3.1555080074711364,
|
| 5 |
+
"train_runtime": 80953.5697,
|
| 6 |
+
"train_samples": 595035,
|
| 7 |
+
"train_samples_per_second": 147.006,
|
| 8 |
+
"train_steps_per_second": 4.594
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2807 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 371900,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05377789728421619,
|
| 13 |
+
"grad_norm": 1.0570567846298218,
|
| 14 |
+
"learning_rate": 3.125e-06,
|
| 15 |
+
"loss": 7.5477,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.10755579456843238,
|
| 20 |
+
"grad_norm": 0.6808657646179199,
|
| 21 |
+
"learning_rate": 6.25e-06,
|
| 22 |
+
"loss": 5.8123,
|
| 23 |
+
"step": 2000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.16133369185264856,
|
| 27 |
+
"grad_norm": 0.7959291934967041,
|
| 28 |
+
"learning_rate": 9.375000000000001e-06,
|
| 29 |
+
"loss": 5.3824,
|
| 30 |
+
"step": 3000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.21511158913686476,
|
| 34 |
+
"grad_norm": 0.8838356137275696,
|
| 35 |
+
"learning_rate": 1.25e-05,
|
| 36 |
+
"loss": 5.161,
|
| 37 |
+
"step": 4000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.26888948642108096,
|
| 41 |
+
"grad_norm": 1.0057505369186401,
|
| 42 |
+
"learning_rate": 1.5625e-05,
|
| 43 |
+
"loss": 5.0158,
|
| 44 |
+
"step": 5000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.32266738370529713,
|
| 48 |
+
"grad_norm": 0.9471587538719177,
|
| 49 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 50 |
+
"loss": 4.8697,
|
| 51 |
+
"step": 6000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.3764452809895133,
|
| 55 |
+
"grad_norm": 1.1075608730316162,
|
| 56 |
+
"learning_rate": 2.1875e-05,
|
| 57 |
+
"loss": 4.7547,
|
| 58 |
+
"step": 7000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.4302231782737295,
|
| 62 |
+
"grad_norm": 1.0996819734573364,
|
| 63 |
+
"learning_rate": 2.5e-05,
|
| 64 |
+
"loss": 4.6408,
|
| 65 |
+
"step": 8000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.4840010755579457,
|
| 69 |
+
"grad_norm": 1.1021990776062012,
|
| 70 |
+
"learning_rate": 2.8125000000000003e-05,
|
| 71 |
+
"loss": 4.5545,
|
| 72 |
+
"step": 9000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.5377789728421619,
|
| 76 |
+
"grad_norm": 1.1215450763702393,
|
| 77 |
+
"learning_rate": 3.125e-05,
|
| 78 |
+
"loss": 4.4789,
|
| 79 |
+
"step": 10000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.591556870126378,
|
| 83 |
+
"grad_norm": 1.0971699953079224,
|
| 84 |
+
"learning_rate": 3.4375e-05,
|
| 85 |
+
"loss": 4.4088,
|
| 86 |
+
"step": 11000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.6453347674105943,
|
| 90 |
+
"grad_norm": 1.058189868927002,
|
| 91 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 92 |
+
"loss": 4.345,
|
| 93 |
+
"step": 12000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.6991126646948105,
|
| 97 |
+
"grad_norm": 1.0064642429351807,
|
| 98 |
+
"learning_rate": 4.061875e-05,
|
| 99 |
+
"loss": 4.2886,
|
| 100 |
+
"step": 13000
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.7528905619790266,
|
| 104 |
+
"grad_norm": 1.0453424453735352,
|
| 105 |
+
"learning_rate": 4.374375e-05,
|
| 106 |
+
"loss": 4.2379,
|
| 107 |
+
"step": 14000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.8066684592632428,
|
| 111 |
+
"grad_norm": 1.0253971815109253,
|
| 112 |
+
"learning_rate": 4.686875e-05,
|
| 113 |
+
"loss": 4.1785,
|
| 114 |
+
"step": 15000
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.860446356547459,
|
| 118 |
+
"grad_norm": 1.0396300554275513,
|
| 119 |
+
"learning_rate": 4.9990625000000004e-05,
|
| 120 |
+
"loss": 4.1385,
|
| 121 |
+
"step": 16000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.9142242538316752,
|
| 125 |
+
"grad_norm": 0.9964418411254883,
|
| 126 |
+
"learning_rate": 5.3115625000000005e-05,
|
| 127 |
+
"loss": 4.0872,
|
| 128 |
+
"step": 17000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.9680021511158914,
|
| 132 |
+
"grad_norm": 1.0650404691696167,
|
| 133 |
+
"learning_rate": 5.6240625e-05,
|
| 134 |
+
"loss": 4.0514,
|
| 135 |
+
"step": 18000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.0,
|
| 139 |
+
"eval_accuracy": 0.3093366068112229,
|
| 140 |
+
"eval_loss": 4.236823558807373,
|
| 141 |
+
"eval_runtime": 152.4756,
|
| 142 |
+
"eval_samples_per_second": 379.871,
|
| 143 |
+
"eval_steps_per_second": 5.942,
|
| 144 |
+
"step": 18595
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.0217800484001076,
|
| 148 |
+
"grad_norm": 1.0185842514038086,
|
| 149 |
+
"learning_rate": 5.93625e-05,
|
| 150 |
+
"loss": 4.0053,
|
| 151 |
+
"step": 19000
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.0755579456843238,
|
| 155 |
+
"grad_norm": 0.972996175289154,
|
| 156 |
+
"learning_rate": 6.24875e-05,
|
| 157 |
+
"loss": 3.9601,
|
| 158 |
+
"step": 20000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.1293358429685398,
|
| 162 |
+
"grad_norm": 1.0843679904937744,
|
| 163 |
+
"learning_rate": 6.560625e-05,
|
| 164 |
+
"loss": 3.9279,
|
| 165 |
+
"step": 21000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.183113740252756,
|
| 169 |
+
"grad_norm": 1.0157690048217773,
|
| 170 |
+
"learning_rate": 6.873125000000001e-05,
|
| 171 |
+
"loss": 3.8922,
|
| 172 |
+
"step": 22000
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.2368916375369723,
|
| 176 |
+
"grad_norm": 0.9889029264450073,
|
| 177 |
+
"learning_rate": 7.185625e-05,
|
| 178 |
+
"loss": 3.8609,
|
| 179 |
+
"step": 23000
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.2906695348211885,
|
| 183 |
+
"grad_norm": 0.9810785055160522,
|
| 184 |
+
"learning_rate": 7.4978125e-05,
|
| 185 |
+
"loss": 3.8232,
|
| 186 |
+
"step": 24000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.3444474321054047,
|
| 190 |
+
"grad_norm": 0.952143669128418,
|
| 191 |
+
"learning_rate": 7.8103125e-05,
|
| 192 |
+
"loss": 3.7906,
|
| 193 |
+
"step": 25000
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.398225329389621,
|
| 197 |
+
"grad_norm": 0.9402492642402649,
|
| 198 |
+
"learning_rate": 8.1221875e-05,
|
| 199 |
+
"loss": 3.7712,
|
| 200 |
+
"step": 26000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.452003226673837,
|
| 204 |
+
"grad_norm": 0.9736520648002625,
|
| 205 |
+
"learning_rate": 8.434687500000001e-05,
|
| 206 |
+
"loss": 3.747,
|
| 207 |
+
"step": 27000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.5057811239580532,
|
| 211 |
+
"grad_norm": 0.8647023439407349,
|
| 212 |
+
"learning_rate": 8.7471875e-05,
|
| 213 |
+
"loss": 3.7225,
|
| 214 |
+
"step": 28000
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.5595590212422694,
|
| 218 |
+
"grad_norm": 0.9048119187355042,
|
| 219 |
+
"learning_rate": 9.059375e-05,
|
| 220 |
+
"loss": 3.6999,
|
| 221 |
+
"step": 29000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.6133369185264856,
|
| 225 |
+
"grad_norm": 0.8969938158988953,
|
| 226 |
+
"learning_rate": 9.371250000000001e-05,
|
| 227 |
+
"loss": 3.6886,
|
| 228 |
+
"step": 30000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.6671148158107019,
|
| 232 |
+
"grad_norm": 0.9778834581375122,
|
| 233 |
+
"learning_rate": 9.68375e-05,
|
| 234 |
+
"loss": 3.67,
|
| 235 |
+
"step": 31000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.7208927130949179,
|
| 239 |
+
"grad_norm": 0.8872840404510498,
|
| 240 |
+
"learning_rate": 9.99625e-05,
|
| 241 |
+
"loss": 3.6502,
|
| 242 |
+
"step": 32000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.7746706103791343,
|
| 246 |
+
"grad_norm": 0.8825569748878479,
|
| 247 |
+
"learning_rate": 9.970932627243307e-05,
|
| 248 |
+
"loss": 3.6309,
|
| 249 |
+
"step": 33000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.8284485076633503,
|
| 253 |
+
"grad_norm": 0.8813254237174988,
|
| 254 |
+
"learning_rate": 9.941512209473375e-05,
|
| 255 |
+
"loss": 3.6146,
|
| 256 |
+
"step": 34000
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.8822264049475665,
|
| 260 |
+
"grad_norm": 0.8341319561004639,
|
| 261 |
+
"learning_rate": 9.912091791703443e-05,
|
| 262 |
+
"loss": 3.5991,
|
| 263 |
+
"step": 35000
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.9360043022317828,
|
| 267 |
+
"grad_norm": 0.8238468170166016,
|
| 268 |
+
"learning_rate": 9.882700794351281e-05,
|
| 269 |
+
"loss": 3.581,
|
| 270 |
+
"step": 36000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.9897821995159988,
|
| 274 |
+
"grad_norm": 0.7880402207374573,
|
| 275 |
+
"learning_rate": 9.853309796999118e-05,
|
| 276 |
+
"loss": 3.5633,
|
| 277 |
+
"step": 37000
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0,
|
| 281 |
+
"eval_accuracy": 0.3636842131073876,
|
| 282 |
+
"eval_loss": 3.7424795627593994,
|
| 283 |
+
"eval_runtime": 153.4633,
|
| 284 |
+
"eval_samples_per_second": 377.426,
|
| 285 |
+
"eval_steps_per_second": 5.904,
|
| 286 |
+
"step": 37190
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 2.043560096800215,
|
| 290 |
+
"grad_norm": 0.8311464786529541,
|
| 291 |
+
"learning_rate": 9.823889379229185e-05,
|
| 292 |
+
"loss": 3.5262,
|
| 293 |
+
"step": 38000
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 2.0973379940844312,
|
| 297 |
+
"grad_norm": 0.8171642422676086,
|
| 298 |
+
"learning_rate": 9.794468961459253e-05,
|
| 299 |
+
"loss": 3.5117,
|
| 300 |
+
"step": 39000
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.1511158913686477,
|
| 304 |
+
"grad_norm": 0.8396981954574585,
|
| 305 |
+
"learning_rate": 9.765048543689321e-05,
|
| 306 |
+
"loss": 3.4998,
|
| 307 |
+
"step": 40000
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.2048937886528637,
|
| 311 |
+
"grad_norm": 0.7978439927101135,
|
| 312 |
+
"learning_rate": 9.735657546337158e-05,
|
| 313 |
+
"loss": 3.4945,
|
| 314 |
+
"step": 41000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.2586716859370797,
|
| 318 |
+
"grad_norm": 0.8739529252052307,
|
| 319 |
+
"learning_rate": 9.706237128567226e-05,
|
| 320 |
+
"loss": 3.486,
|
| 321 |
+
"step": 42000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.312449583221296,
|
| 325 |
+
"grad_norm": 0.7910430431365967,
|
| 326 |
+
"learning_rate": 9.676816710797295e-05,
|
| 327 |
+
"loss": 3.474,
|
| 328 |
+
"step": 43000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.366227480505512,
|
| 332 |
+
"grad_norm": 0.7927525639533997,
|
| 333 |
+
"learning_rate": 9.647396293027361e-05,
|
| 334 |
+
"loss": 3.4721,
|
| 335 |
+
"step": 44000
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 2.4200053777897286,
|
| 339 |
+
"grad_norm": 0.8172622919082642,
|
| 340 |
+
"learning_rate": 9.61797587525743e-05,
|
| 341 |
+
"loss": 3.4593,
|
| 342 |
+
"step": 45000
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.4737832750739446,
|
| 346 |
+
"grad_norm": 0.7903227210044861,
|
| 347 |
+
"learning_rate": 9.588584877905267e-05,
|
| 348 |
+
"loss": 3.456,
|
| 349 |
+
"step": 46000
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.527561172358161,
|
| 353 |
+
"grad_norm": 0.7282652854919434,
|
| 354 |
+
"learning_rate": 9.559164460135335e-05,
|
| 355 |
+
"loss": 3.4452,
|
| 356 |
+
"step": 47000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.581339069642377,
|
| 360 |
+
"grad_norm": 0.7477676272392273,
|
| 361 |
+
"learning_rate": 9.529744042365401e-05,
|
| 362 |
+
"loss": 3.4331,
|
| 363 |
+
"step": 48000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.635116966926593,
|
| 367 |
+
"grad_norm": 0.788345456123352,
|
| 368 |
+
"learning_rate": 9.50035304501324e-05,
|
| 369 |
+
"loss": 3.4329,
|
| 370 |
+
"step": 49000
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 2.6888948642108095,
|
| 374 |
+
"grad_norm": 0.7750670909881592,
|
| 375 |
+
"learning_rate": 9.470932627243308e-05,
|
| 376 |
+
"loss": 3.4215,
|
| 377 |
+
"step": 50000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.7426727614950255,
|
| 381 |
+
"grad_norm": 0.783518373966217,
|
| 382 |
+
"learning_rate": 9.441541629891145e-05,
|
| 383 |
+
"loss": 3.4109,
|
| 384 |
+
"step": 51000
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.796450658779242,
|
| 388 |
+
"grad_norm": 0.7716132998466492,
|
| 389 |
+
"learning_rate": 9.412121212121212e-05,
|
| 390 |
+
"loss": 3.408,
|
| 391 |
+
"step": 52000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.850228556063458,
|
| 395 |
+
"grad_norm": 0.7686366438865662,
|
| 396 |
+
"learning_rate": 9.382759635186819e-05,
|
| 397 |
+
"loss": 3.3991,
|
| 398 |
+
"step": 53000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.904006453347674,
|
| 402 |
+
"grad_norm": 0.7560417652130127,
|
| 403 |
+
"learning_rate": 9.353339217416887e-05,
|
| 404 |
+
"loss": 3.3948,
|
| 405 |
+
"step": 54000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.9577843506318904,
|
| 409 |
+
"grad_norm": 0.7553894519805908,
|
| 410 |
+
"learning_rate": 9.323918799646955e-05,
|
| 411 |
+
"loss": 3.3923,
|
| 412 |
+
"step": 55000
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 3.0,
|
| 416 |
+
"eval_accuracy": 0.3808668818127898,
|
| 417 |
+
"eval_loss": 3.571072816848755,
|
| 418 |
+
"eval_runtime": 153.4053,
|
| 419 |
+
"eval_samples_per_second": 377.568,
|
| 420 |
+
"eval_steps_per_second": 5.906,
|
| 421 |
+
"step": 55785
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.0115622479161064,
|
| 425 |
+
"grad_norm": 0.7793614864349365,
|
| 426 |
+
"learning_rate": 9.294498381877022e-05,
|
| 427 |
+
"loss": 3.3786,
|
| 428 |
+
"step": 56000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.065340145200323,
|
| 432 |
+
"grad_norm": 0.7647926807403564,
|
| 433 |
+
"learning_rate": 9.26510738452486e-05,
|
| 434 |
+
"loss": 3.3385,
|
| 435 |
+
"step": 57000
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.119118042484539,
|
| 439 |
+
"grad_norm": 0.7615726590156555,
|
| 440 |
+
"learning_rate": 9.235716387172698e-05,
|
| 441 |
+
"loss": 3.3351,
|
| 442 |
+
"step": 58000
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.172895939768755,
|
| 446 |
+
"grad_norm": 0.8117246627807617,
|
| 447 |
+
"learning_rate": 9.206295969402766e-05,
|
| 448 |
+
"loss": 3.3302,
|
| 449 |
+
"step": 59000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.2266738370529713,
|
| 453 |
+
"grad_norm": 0.7565757036209106,
|
| 454 |
+
"learning_rate": 9.176904972050604e-05,
|
| 455 |
+
"loss": 3.3332,
|
| 456 |
+
"step": 60000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.2804517343371873,
|
| 460 |
+
"grad_norm": 0.7616593241691589,
|
| 461 |
+
"learning_rate": 9.147484554280671e-05,
|
| 462 |
+
"loss": 3.3281,
|
| 463 |
+
"step": 61000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.3342296316214037,
|
| 467 |
+
"grad_norm": 0.748747706413269,
|
| 468 |
+
"learning_rate": 9.118064136510739e-05,
|
| 469 |
+
"loss": 3.3279,
|
| 470 |
+
"step": 62000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.3880075289056197,
|
| 474 |
+
"grad_norm": 0.7388283014297485,
|
| 475 |
+
"learning_rate": 9.088673139158576e-05,
|
| 476 |
+
"loss": 3.3238,
|
| 477 |
+
"step": 63000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.4417854261898357,
|
| 481 |
+
"grad_norm": 0.7752501964569092,
|
| 482 |
+
"learning_rate": 9.059252721388644e-05,
|
| 483 |
+
"loss": 3.3198,
|
| 484 |
+
"step": 64000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.495563323474052,
|
| 488 |
+
"grad_norm": 0.7556226253509521,
|
| 489 |
+
"learning_rate": 9.029832303618711e-05,
|
| 490 |
+
"loss": 3.3164,
|
| 491 |
+
"step": 65000
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.549341220758268,
|
| 495 |
+
"grad_norm": 0.749092698097229,
|
| 496 |
+
"learning_rate": 9.000411885848779e-05,
|
| 497 |
+
"loss": 3.3077,
|
| 498 |
+
"step": 66000
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.6031191180424846,
|
| 502 |
+
"grad_norm": 0.7285726070404053,
|
| 503 |
+
"learning_rate": 8.971020888496617e-05,
|
| 504 |
+
"loss": 3.3067,
|
| 505 |
+
"step": 67000
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.6568970153267006,
|
| 509 |
+
"grad_norm": 0.745343804359436,
|
| 510 |
+
"learning_rate": 8.941600470726684e-05,
|
| 511 |
+
"loss": 3.3077,
|
| 512 |
+
"step": 68000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.7106749126109166,
|
| 516 |
+
"grad_norm": 0.7441139221191406,
|
| 517 |
+
"learning_rate": 8.912209473374523e-05,
|
| 518 |
+
"loss": 3.304,
|
| 519 |
+
"step": 69000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.764452809895133,
|
| 523 |
+
"grad_norm": 0.7383400201797485,
|
| 524 |
+
"learning_rate": 8.88278905560459e-05,
|
| 525 |
+
"loss": 3.3051,
|
| 526 |
+
"step": 70000
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.8182307071793495,
|
| 530 |
+
"grad_norm": 0.7085037231445312,
|
| 531 |
+
"learning_rate": 8.853368637834658e-05,
|
| 532 |
+
"loss": 3.2976,
|
| 533 |
+
"step": 71000
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 3.8720086044635655,
|
| 537 |
+
"grad_norm": 0.7183367013931274,
|
| 538 |
+
"learning_rate": 8.823948220064724e-05,
|
| 539 |
+
"loss": 3.2941,
|
| 540 |
+
"step": 72000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 3.9257865017477815,
|
| 544 |
+
"grad_norm": 0.7206182479858398,
|
| 545 |
+
"learning_rate": 8.794527802294792e-05,
|
| 546 |
+
"loss": 3.2894,
|
| 547 |
+
"step": 73000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 3.979564399031998,
|
| 551 |
+
"grad_norm": 0.711100161075592,
|
| 552 |
+
"learning_rate": 8.76510738452486e-05,
|
| 553 |
+
"loss": 3.2852,
|
| 554 |
+
"step": 74000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 4.0,
|
| 558 |
+
"eval_accuracy": 0.3879775330641219,
|
| 559 |
+
"eval_loss": 3.514986038208008,
|
| 560 |
+
"eval_runtime": 153.4121,
|
| 561 |
+
"eval_samples_per_second": 377.552,
|
| 562 |
+
"eval_steps_per_second": 5.906,
|
| 563 |
+
"step": 74380
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.033342296316214,
|
| 567 |
+
"grad_norm": 0.7628651857376099,
|
| 568 |
+
"learning_rate": 8.735745807590468e-05,
|
| 569 |
+
"loss": 3.2569,
|
| 570 |
+
"step": 75000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 4.08712019360043,
|
| 574 |
+
"grad_norm": 0.7417529821395874,
|
| 575 |
+
"learning_rate": 8.706325389820536e-05,
|
| 576 |
+
"loss": 3.2338,
|
| 577 |
+
"step": 76000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 4.140898090884646,
|
| 581 |
+
"grad_norm": 0.7533174753189087,
|
| 582 |
+
"learning_rate": 8.676934392468373e-05,
|
| 583 |
+
"loss": 3.238,
|
| 584 |
+
"step": 77000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.1946759881688624,
|
| 588 |
+
"grad_norm": 0.7438268065452576,
|
| 589 |
+
"learning_rate": 8.647513974698441e-05,
|
| 590 |
+
"loss": 3.2377,
|
| 591 |
+
"step": 78000
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.2484538854530784,
|
| 595 |
+
"grad_norm": 0.7864522337913513,
|
| 596 |
+
"learning_rate": 8.61809355692851e-05,
|
| 597 |
+
"loss": 3.2328,
|
| 598 |
+
"step": 79000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.302231782737295,
|
| 602 |
+
"grad_norm": 0.7351215481758118,
|
| 603 |
+
"learning_rate": 8.588673139158576e-05,
|
| 604 |
+
"loss": 3.2363,
|
| 605 |
+
"step": 80000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 4.356009680021511,
|
| 609 |
+
"grad_norm": 0.7796077132225037,
|
| 610 |
+
"learning_rate": 8.559252721388644e-05,
|
| 611 |
+
"loss": 3.233,
|
| 612 |
+
"step": 81000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 4.409787577305727,
|
| 616 |
+
"grad_norm": 0.7376981973648071,
|
| 617 |
+
"learning_rate": 8.529861724036483e-05,
|
| 618 |
+
"loss": 3.2399,
|
| 619 |
+
"step": 82000
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 4.463565474589943,
|
| 623 |
+
"grad_norm": 0.7308441400527954,
|
| 624 |
+
"learning_rate": 8.50044130626655e-05,
|
| 625 |
+
"loss": 3.2347,
|
| 626 |
+
"step": 83000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.517343371874159,
|
| 630 |
+
"grad_norm": 0.7723622918128967,
|
| 631 |
+
"learning_rate": 8.471050308914386e-05,
|
| 632 |
+
"loss": 3.2318,
|
| 633 |
+
"step": 84000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.571121269158376,
|
| 637 |
+
"grad_norm": 0.7057169675827026,
|
| 638 |
+
"learning_rate": 8.441629891144455e-05,
|
| 639 |
+
"loss": 3.231,
|
| 640 |
+
"step": 85000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 4.624899166442592,
|
| 644 |
+
"grad_norm": 0.7399170398712158,
|
| 645 |
+
"learning_rate": 8.412238893792293e-05,
|
| 646 |
+
"loss": 3.2364,
|
| 647 |
+
"step": 86000
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 4.678677063726808,
|
| 651 |
+
"grad_norm": 0.7232281565666199,
|
| 652 |
+
"learning_rate": 8.38281847602236e-05,
|
| 653 |
+
"loss": 3.2289,
|
| 654 |
+
"step": 87000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 4.732454961011024,
|
| 658 |
+
"grad_norm": 0.7580758333206177,
|
| 659 |
+
"learning_rate": 8.353398058252428e-05,
|
| 660 |
+
"loss": 3.2274,
|
| 661 |
+
"step": 88000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 4.786232858295241,
|
| 665 |
+
"grad_norm": 0.7439358830451965,
|
| 666 |
+
"learning_rate": 8.324007060900265e-05,
|
| 667 |
+
"loss": 3.2295,
|
| 668 |
+
"step": 89000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.840010755579457,
|
| 672 |
+
"grad_norm": 0.7321400046348572,
|
| 673 |
+
"learning_rate": 8.294586643130333e-05,
|
| 674 |
+
"loss": 3.225,
|
| 675 |
+
"step": 90000
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 4.893788652863673,
|
| 679 |
+
"grad_norm": 0.7247775793075562,
|
| 680 |
+
"learning_rate": 8.2651662253604e-05,
|
| 681 |
+
"loss": 3.2252,
|
| 682 |
+
"step": 91000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 4.947566550147889,
|
| 686 |
+
"grad_norm": 0.7113615870475769,
|
| 687 |
+
"learning_rate": 8.235745807590468e-05,
|
| 688 |
+
"loss": 3.2225,
|
| 689 |
+
"step": 92000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 5.0,
|
| 693 |
+
"eval_accuracy": 0.39335161799947965,
|
| 694 |
+
"eval_loss": 3.447347640991211,
|
| 695 |
+
"eval_runtime": 153.9417,
|
| 696 |
+
"eval_samples_per_second": 376.253,
|
| 697 |
+
"eval_steps_per_second": 5.885,
|
| 698 |
+
"step": 92975
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 5.001344447432105,
|
| 702 |
+
"grad_norm": 0.7526562213897705,
|
| 703 |
+
"learning_rate": 8.206354810238306e-05,
|
| 704 |
+
"loss": 3.2241,
|
| 705 |
+
"step": 93000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 5.055122344716321,
|
| 709 |
+
"grad_norm": 0.7481445670127869,
|
| 710 |
+
"learning_rate": 8.176934392468374e-05,
|
| 711 |
+
"loss": 3.1671,
|
| 712 |
+
"step": 94000
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 5.108900242000538,
|
| 716 |
+
"grad_norm": 0.8045230507850647,
|
| 717 |
+
"learning_rate": 8.147513974698441e-05,
|
| 718 |
+
"loss": 3.1662,
|
| 719 |
+
"step": 95000
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 5.162678139284754,
|
| 723 |
+
"grad_norm": 0.754204273223877,
|
| 724 |
+
"learning_rate": 8.11809355692851e-05,
|
| 725 |
+
"loss": 3.173,
|
| 726 |
+
"step": 96000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 5.21645603656897,
|
| 730 |
+
"grad_norm": 0.7404423952102661,
|
| 731 |
+
"learning_rate": 8.088702559576346e-05,
|
| 732 |
+
"loss": 3.1749,
|
| 733 |
+
"step": 97000
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 5.270233933853186,
|
| 737 |
+
"grad_norm": 0.7419277429580688,
|
| 738 |
+
"learning_rate": 8.059282141806413e-05,
|
| 739 |
+
"loss": 3.1741,
|
| 740 |
+
"step": 98000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 5.324011831137403,
|
| 744 |
+
"grad_norm": 0.6956449747085571,
|
| 745 |
+
"learning_rate": 8.029891144454252e-05,
|
| 746 |
+
"loss": 3.1756,
|
| 747 |
+
"step": 99000
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 5.377789728421619,
|
| 751 |
+
"grad_norm": 0.7412118911743164,
|
| 752 |
+
"learning_rate": 8.00047072668432e-05,
|
| 753 |
+
"loss": 3.173,
|
| 754 |
+
"step": 100000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 5.431567625705835,
|
| 758 |
+
"grad_norm": 0.7704578638076782,
|
| 759 |
+
"learning_rate": 7.971050308914388e-05,
|
| 760 |
+
"loss": 3.1749,
|
| 761 |
+
"step": 101000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 5.485345522990051,
|
| 765 |
+
"grad_norm": 0.733038604259491,
|
| 766 |
+
"learning_rate": 7.941629891144455e-05,
|
| 767 |
+
"loss": 3.1731,
|
| 768 |
+
"step": 102000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 5.539123420274267,
|
| 772 |
+
"grad_norm": 0.7722159624099731,
|
| 773 |
+
"learning_rate": 7.912209473374523e-05,
|
| 774 |
+
"loss": 3.1728,
|
| 775 |
+
"step": 103000
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 5.592901317558484,
|
| 779 |
+
"grad_norm": 0.760040283203125,
|
| 780 |
+
"learning_rate": 7.88284789644013e-05,
|
| 781 |
+
"loss": 3.1734,
|
| 782 |
+
"step": 104000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 5.6466792148427,
|
| 786 |
+
"grad_norm": 0.7532190680503845,
|
| 787 |
+
"learning_rate": 7.853427478670198e-05,
|
| 788 |
+
"loss": 3.1766,
|
| 789 |
+
"step": 105000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 5.700457112126916,
|
| 793 |
+
"grad_norm": 0.7432074546813965,
|
| 794 |
+
"learning_rate": 7.824007060900265e-05,
|
| 795 |
+
"loss": 3.1757,
|
| 796 |
+
"step": 106000
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 5.754235009411132,
|
| 800 |
+
"grad_norm": 0.7651804089546204,
|
| 801 |
+
"learning_rate": 7.794586643130333e-05,
|
| 802 |
+
"loss": 3.173,
|
| 803 |
+
"step": 107000
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 5.808012906695348,
|
| 807 |
+
"grad_norm": 0.7277229428291321,
|
| 808 |
+
"learning_rate": 7.76519564577817e-05,
|
| 809 |
+
"loss": 3.1744,
|
| 810 |
+
"step": 108000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 5.861790803979565,
|
| 814 |
+
"grad_norm": 0.7225542664527893,
|
| 815 |
+
"learning_rate": 7.735775228008238e-05,
|
| 816 |
+
"loss": 3.174,
|
| 817 |
+
"step": 109000
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 5.915568701263781,
|
| 821 |
+
"grad_norm": 0.727301836013794,
|
| 822 |
+
"learning_rate": 7.706354810238305e-05,
|
| 823 |
+
"loss": 3.1716,
|
| 824 |
+
"step": 110000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 5.969346598547997,
|
| 828 |
+
"grad_norm": 0.6921839714050293,
|
| 829 |
+
"learning_rate": 7.676934392468373e-05,
|
| 830 |
+
"loss": 3.1717,
|
| 831 |
+
"step": 111000
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 6.0,
|
| 835 |
+
"eval_accuracy": 0.39688460674700016,
|
| 836 |
+
"eval_loss": 3.446603298187256,
|
| 837 |
+
"eval_runtime": 154.8201,
|
| 838 |
+
"eval_samples_per_second": 374.118,
|
| 839 |
+
"eval_steps_per_second": 5.852,
|
| 840 |
+
"step": 111570
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 6.023124495832213,
|
| 844 |
+
"grad_norm": 0.7526515126228333,
|
| 845 |
+
"learning_rate": 7.647543395116212e-05,
|
| 846 |
+
"loss": 3.1438,
|
| 847 |
+
"step": 112000
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 6.076902393116429,
|
| 851 |
+
"grad_norm": 0.7599433064460754,
|
| 852 |
+
"learning_rate": 7.61812297734628e-05,
|
| 853 |
+
"loss": 3.1177,
|
| 854 |
+
"step": 113000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 6.130680290400646,
|
| 858 |
+
"grad_norm": 0.7775200605392456,
|
| 859 |
+
"learning_rate": 7.588731979994115e-05,
|
| 860 |
+
"loss": 3.1142,
|
| 861 |
+
"step": 114000
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 6.184458187684862,
|
| 865 |
+
"grad_norm": 0.7804093360900879,
|
| 866 |
+
"learning_rate": 7.559311562224184e-05,
|
| 867 |
+
"loss": 3.1241,
|
| 868 |
+
"step": 115000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 6.238236084969078,
|
| 872 |
+
"grad_norm": 0.7484256029129028,
|
| 873 |
+
"learning_rate": 7.529891144454252e-05,
|
| 874 |
+
"loss": 3.1267,
|
| 875 |
+
"step": 116000
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 6.292013982253294,
|
| 879 |
+
"grad_norm": 0.7499005198478699,
|
| 880 |
+
"learning_rate": 7.50050014710209e-05,
|
| 881 |
+
"loss": 3.126,
|
| 882 |
+
"step": 117000
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 6.34579187953751,
|
| 886 |
+
"grad_norm": 0.7697141170501709,
|
| 887 |
+
"learning_rate": 7.471079729332157e-05,
|
| 888 |
+
"loss": 3.1272,
|
| 889 |
+
"step": 118000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 6.399569776821727,
|
| 893 |
+
"grad_norm": 0.7145485877990723,
|
| 894 |
+
"learning_rate": 7.441718152397764e-05,
|
| 895 |
+
"loss": 3.1242,
|
| 896 |
+
"step": 119000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 6.453347674105943,
|
| 900 |
+
"grad_norm": 0.754996120929718,
|
| 901 |
+
"learning_rate": 7.412297734627832e-05,
|
| 902 |
+
"loss": 3.1299,
|
| 903 |
+
"step": 120000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 6.507125571390159,
|
| 907 |
+
"grad_norm": 0.758781373500824,
|
| 908 |
+
"learning_rate": 7.3828773168579e-05,
|
| 909 |
+
"loss": 3.1322,
|
| 910 |
+
"step": 121000
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 6.560903468674375,
|
| 914 |
+
"grad_norm": 0.7344619035720825,
|
| 915 |
+
"learning_rate": 7.353456899087967e-05,
|
| 916 |
+
"loss": 3.1323,
|
| 917 |
+
"step": 122000
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 6.614681365958591,
|
| 921 |
+
"grad_norm": 0.7554389238357544,
|
| 922 |
+
"learning_rate": 7.324065901735804e-05,
|
| 923 |
+
"loss": 3.1284,
|
| 924 |
+
"step": 123000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 6.6684592632428075,
|
| 928 |
+
"grad_norm": 0.735532283782959,
|
| 929 |
+
"learning_rate": 7.294674904383643e-05,
|
| 930 |
+
"loss": 3.1293,
|
| 931 |
+
"step": 124000
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 6.7222371605270235,
|
| 935 |
+
"grad_norm": 0.7296610474586487,
|
| 936 |
+
"learning_rate": 7.265254486613711e-05,
|
| 937 |
+
"loss": 3.1297,
|
| 938 |
+
"step": 125000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 6.7760150578112395,
|
| 942 |
+
"grad_norm": 0.730621337890625,
|
| 943 |
+
"learning_rate": 7.235834068843778e-05,
|
| 944 |
+
"loss": 3.1307,
|
| 945 |
+
"step": 126000
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 6.8297929550954555,
|
| 949 |
+
"grad_norm": 0.7528539896011353,
|
| 950 |
+
"learning_rate": 7.206413651073846e-05,
|
| 951 |
+
"loss": 3.1275,
|
| 952 |
+
"step": 127000
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 6.8835708523796715,
|
| 956 |
+
"grad_norm": 0.745832085609436,
|
| 957 |
+
"learning_rate": 7.176993233303914e-05,
|
| 958 |
+
"loss": 3.1287,
|
| 959 |
+
"step": 128000
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 6.937348749663888,
|
| 963 |
+
"grad_norm": 0.7488901615142822,
|
| 964 |
+
"learning_rate": 7.147602235951751e-05,
|
| 965 |
+
"loss": 3.1291,
|
| 966 |
+
"step": 129000
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 6.991126646948104,
|
| 970 |
+
"grad_norm": 0.7614565491676331,
|
| 971 |
+
"learning_rate": 7.118181818181818e-05,
|
| 972 |
+
"loss": 3.128,
|
| 973 |
+
"step": 130000
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 7.0,
|
| 977 |
+
"eval_accuracy": 0.3993379685210575,
|
| 978 |
+
"eval_loss": 3.4202613830566406,
|
| 979 |
+
"eval_runtime": 154.6299,
|
| 980 |
+
"eval_samples_per_second": 374.578,
|
| 981 |
+
"eval_steps_per_second": 5.859,
|
| 982 |
+
"step": 130165
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 7.04490454423232,
|
| 986 |
+
"grad_norm": 0.7734451293945312,
|
| 987 |
+
"learning_rate": 7.088790820829656e-05,
|
| 988 |
+
"loss": 3.0802,
|
| 989 |
+
"step": 131000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 7.098682441516536,
|
| 993 |
+
"grad_norm": 0.7710923552513123,
|
| 994 |
+
"learning_rate": 7.059370403059724e-05,
|
| 995 |
+
"loss": 3.0728,
|
| 996 |
+
"step": 132000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 7.152460338800753,
|
| 1000 |
+
"grad_norm": 0.7468781471252441,
|
| 1001 |
+
"learning_rate": 7.029949985289791e-05,
|
| 1002 |
+
"loss": 3.0821,
|
| 1003 |
+
"step": 133000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 7.206238236084969,
|
| 1007 |
+
"grad_norm": 0.7747782468795776,
|
| 1008 |
+
"learning_rate": 7.000558987937629e-05,
|
| 1009 |
+
"loss": 3.0847,
|
| 1010 |
+
"step": 134000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 7.260016133369185,
|
| 1014 |
+
"grad_norm": 0.7802600264549255,
|
| 1015 |
+
"learning_rate": 6.971138570167697e-05,
|
| 1016 |
+
"loss": 3.0839,
|
| 1017 |
+
"step": 135000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 7.313794030653401,
|
| 1021 |
+
"grad_norm": 0.7671915292739868,
|
| 1022 |
+
"learning_rate": 6.941747572815534e-05,
|
| 1023 |
+
"loss": 3.087,
|
| 1024 |
+
"step": 136000
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 7.367571927937617,
|
| 1028 |
+
"grad_norm": 0.7691196799278259,
|
| 1029 |
+
"learning_rate": 6.912327155045601e-05,
|
| 1030 |
+
"loss": 3.0904,
|
| 1031 |
+
"step": 137000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 7.421349825221834,
|
| 1035 |
+
"grad_norm": 0.7774937748908997,
|
| 1036 |
+
"learning_rate": 6.88290673727567e-05,
|
| 1037 |
+
"loss": 3.0924,
|
| 1038 |
+
"step": 138000
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 7.47512772250605,
|
| 1042 |
+
"grad_norm": 0.7633153200149536,
|
| 1043 |
+
"learning_rate": 6.853515739923508e-05,
|
| 1044 |
+
"loss": 3.09,
|
| 1045 |
+
"step": 139000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 7.528905619790266,
|
| 1049 |
+
"grad_norm": 0.7772113084793091,
|
| 1050 |
+
"learning_rate": 6.824095322153576e-05,
|
| 1051 |
+
"loss": 3.0905,
|
| 1052 |
+
"step": 140000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 7.582683517074482,
|
| 1056 |
+
"grad_norm": 0.7359101176261902,
|
| 1057 |
+
"learning_rate": 6.794704324801413e-05,
|
| 1058 |
+
"loss": 3.094,
|
| 1059 |
+
"step": 141000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 7.636461414358699,
|
| 1063 |
+
"grad_norm": 0.7604731917381287,
|
| 1064 |
+
"learning_rate": 6.76531332744925e-05,
|
| 1065 |
+
"loss": 3.0936,
|
| 1066 |
+
"step": 142000
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 7.690239311642915,
|
| 1070 |
+
"grad_norm": 0.7610685229301453,
|
| 1071 |
+
"learning_rate": 6.735892909679318e-05,
|
| 1072 |
+
"loss": 3.0904,
|
| 1073 |
+
"step": 143000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 7.744017208927131,
|
| 1077 |
+
"grad_norm": 0.7654697299003601,
|
| 1078 |
+
"learning_rate": 6.706472491909386e-05,
|
| 1079 |
+
"loss": 3.0953,
|
| 1080 |
+
"step": 144000
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 7.797795106211347,
|
| 1084 |
+
"grad_norm": 0.7604765295982361,
|
| 1085 |
+
"learning_rate": 6.677052074139453e-05,
|
| 1086 |
+
"loss": 3.0934,
|
| 1087 |
+
"step": 145000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 7.851573003495563,
|
| 1091 |
+
"grad_norm": 0.7256069779396057,
|
| 1092 |
+
"learning_rate": 6.64766107678729e-05,
|
| 1093 |
+
"loss": 3.0945,
|
| 1094 |
+
"step": 146000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 7.90535090077978,
|
| 1098 |
+
"grad_norm": 0.7387049198150635,
|
| 1099 |
+
"learning_rate": 6.618240659017358e-05,
|
| 1100 |
+
"loss": 3.095,
|
| 1101 |
+
"step": 147000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 7.959128798063996,
|
| 1105 |
+
"grad_norm": 0.7363874316215515,
|
| 1106 |
+
"learning_rate": 6.588820241247426e-05,
|
| 1107 |
+
"loss": 3.0952,
|
| 1108 |
+
"step": 148000
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 8.0,
|
| 1112 |
+
"eval_accuracy": 0.4015226831501407,
|
| 1113 |
+
"eval_loss": 3.3998935222625732,
|
| 1114 |
+
"eval_runtime": 154.927,
|
| 1115 |
+
"eval_samples_per_second": 373.86,
|
| 1116 |
+
"eval_steps_per_second": 5.848,
|
| 1117 |
+
"step": 148760
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 8.012906695348212,
|
| 1121 |
+
"grad_norm": 0.7477943301200867,
|
| 1122 |
+
"learning_rate": 6.559429243895263e-05,
|
| 1123 |
+
"loss": 3.0746,
|
| 1124 |
+
"step": 149000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 8.066684592632429,
|
| 1128 |
+
"grad_norm": 0.7550663352012634,
|
| 1129 |
+
"learning_rate": 6.530008826125332e-05,
|
| 1130 |
+
"loss": 3.0386,
|
| 1131 |
+
"step": 150000
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 8.120462489916644,
|
| 1135 |
+
"grad_norm": 0.7995402812957764,
|
| 1136 |
+
"learning_rate": 6.500647249190939e-05,
|
| 1137 |
+
"loss": 3.0399,
|
| 1138 |
+
"step": 151000
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 8.17424038720086,
|
| 1142 |
+
"grad_norm": 0.7675312757492065,
|
| 1143 |
+
"learning_rate": 6.471226831421007e-05,
|
| 1144 |
+
"loss": 3.0491,
|
| 1145 |
+
"step": 152000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 8.228018284485076,
|
| 1149 |
+
"grad_norm": 0.7625625133514404,
|
| 1150 |
+
"learning_rate": 6.441806413651074e-05,
|
| 1151 |
+
"loss": 3.0487,
|
| 1152 |
+
"step": 153000
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 8.281796181769293,
|
| 1156 |
+
"grad_norm": 0.7804884314537048,
|
| 1157 |
+
"learning_rate": 6.412385995881142e-05,
|
| 1158 |
+
"loss": 3.0537,
|
| 1159 |
+
"step": 154000
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 8.33557407905351,
|
| 1163 |
+
"grad_norm": 0.7776561975479126,
|
| 1164 |
+
"learning_rate": 6.38296557811121e-05,
|
| 1165 |
+
"loss": 3.0546,
|
| 1166 |
+
"step": 155000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 8.389351976337725,
|
| 1170 |
+
"grad_norm": 0.7962175011634827,
|
| 1171 |
+
"learning_rate": 6.353604001176817e-05,
|
| 1172 |
+
"loss": 3.0582,
|
| 1173 |
+
"step": 156000
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 8.443129873621942,
|
| 1177 |
+
"grad_norm": 0.759337842464447,
|
| 1178 |
+
"learning_rate": 6.324183583406885e-05,
|
| 1179 |
+
"loss": 3.0575,
|
| 1180 |
+
"step": 157000
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 8.496907770906157,
|
| 1184 |
+
"grad_norm": 0.7623013257980347,
|
| 1185 |
+
"learning_rate": 6.294763165636952e-05,
|
| 1186 |
+
"loss": 3.0582,
|
| 1187 |
+
"step": 158000
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 8.550685668190374,
|
| 1191 |
+
"grad_norm": 0.7721423506736755,
|
| 1192 |
+
"learning_rate": 6.26534274786702e-05,
|
| 1193 |
+
"loss": 3.0607,
|
| 1194 |
+
"step": 159000
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 8.60446356547459,
|
| 1198 |
+
"grad_norm": 0.761391282081604,
|
| 1199 |
+
"learning_rate": 6.235922330097088e-05,
|
| 1200 |
+
"loss": 3.0586,
|
| 1201 |
+
"step": 160000
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 8.658241462758806,
|
| 1205 |
+
"grad_norm": 0.784119725227356,
|
| 1206 |
+
"learning_rate": 6.206531332744924e-05,
|
| 1207 |
+
"loss": 3.0575,
|
| 1208 |
+
"step": 161000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 8.712019360043023,
|
| 1212 |
+
"grad_norm": 0.7654731869697571,
|
| 1213 |
+
"learning_rate": 6.177110914974992e-05,
|
| 1214 |
+
"loss": 3.0626,
|
| 1215 |
+
"step": 162000
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 8.765797257327238,
|
| 1219 |
+
"grad_norm": 0.7638598084449768,
|
| 1220 |
+
"learning_rate": 6.14769049720506e-05,
|
| 1221 |
+
"loss": 3.0652,
|
| 1222 |
+
"step": 163000
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 8.819575154611455,
|
| 1226 |
+
"grad_norm": 0.756234347820282,
|
| 1227 |
+
"learning_rate": 6.118299499852899e-05,
|
| 1228 |
+
"loss": 3.0618,
|
| 1229 |
+
"step": 164000
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 8.873353051895672,
|
| 1233 |
+
"grad_norm": 0.7478641271591187,
|
| 1234 |
+
"learning_rate": 6.0888790820829656e-05,
|
| 1235 |
+
"loss": 3.0644,
|
| 1236 |
+
"step": 165000
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 8.927130949179887,
|
| 1240 |
+
"grad_norm": 0.7811470031738281,
|
| 1241 |
+
"learning_rate": 6.059458664313034e-05,
|
| 1242 |
+
"loss": 3.063,
|
| 1243 |
+
"step": 166000
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 8.980908846464104,
|
| 1247 |
+
"grad_norm": 0.7464701533317566,
|
| 1248 |
+
"learning_rate": 6.030097087378641e-05,
|
| 1249 |
+
"loss": 3.0633,
|
| 1250 |
+
"step": 167000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 9.0,
|
| 1254 |
+
"eval_accuracy": 0.40252921982759693,
|
| 1255 |
+
"eval_loss": 3.4023165702819824,
|
| 1256 |
+
"eval_runtime": 155.1829,
|
| 1257 |
+
"eval_samples_per_second": 373.243,
|
| 1258 |
+
"eval_steps_per_second": 5.838,
|
| 1259 |
+
"step": 167355
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 9.034686743748319,
|
| 1263 |
+
"grad_norm": 0.7871458530426025,
|
| 1264 |
+
"learning_rate": 6.000676669608709e-05,
|
| 1265 |
+
"loss": 3.025,
|
| 1266 |
+
"step": 168000
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 9.088464641032536,
|
| 1270 |
+
"grad_norm": 0.8074026107788086,
|
| 1271 |
+
"learning_rate": 5.971256251838776e-05,
|
| 1272 |
+
"loss": 3.0102,
|
| 1273 |
+
"step": 169000
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 9.142242538316752,
|
| 1277 |
+
"grad_norm": 0.8040223121643066,
|
| 1278 |
+
"learning_rate": 5.941835834068844e-05,
|
| 1279 |
+
"loss": 3.0151,
|
| 1280 |
+
"step": 170000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 9.196020435600968,
|
| 1284 |
+
"grad_norm": 0.8077701926231384,
|
| 1285 |
+
"learning_rate": 5.912444836716682e-05,
|
| 1286 |
+
"loss": 3.0174,
|
| 1287 |
+
"step": 171000
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 9.249798332885184,
|
| 1291 |
+
"grad_norm": 0.7964481115341187,
|
| 1292 |
+
"learning_rate": 5.88302441894675e-05,
|
| 1293 |
+
"loss": 3.0205,
|
| 1294 |
+
"step": 172000
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 9.3035762301694,
|
| 1298 |
+
"grad_norm": 0.7726272940635681,
|
| 1299 |
+
"learning_rate": 5.853604001176817e-05,
|
| 1300 |
+
"loss": 3.0249,
|
| 1301 |
+
"step": 173000
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 9.357354127453616,
|
| 1305 |
+
"grad_norm": 0.8204562067985535,
|
| 1306 |
+
"learning_rate": 5.8242130038246544e-05,
|
| 1307 |
+
"loss": 3.0274,
|
| 1308 |
+
"step": 174000
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 9.411132024737833,
|
| 1312 |
+
"grad_norm": 0.7974728345870972,
|
| 1313 |
+
"learning_rate": 5.7947925860547225e-05,
|
| 1314 |
+
"loss": 3.0311,
|
| 1315 |
+
"step": 175000
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 9.464909922022049,
|
| 1319 |
+
"grad_norm": 0.7713531851768494,
|
| 1320 |
+
"learning_rate": 5.76540158870256e-05,
|
| 1321 |
+
"loss": 3.0267,
|
| 1322 |
+
"step": 176000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 9.518687819306265,
|
| 1326 |
+
"grad_norm": 0.7875762581825256,
|
| 1327 |
+
"learning_rate": 5.735981170932627e-05,
|
| 1328 |
+
"loss": 3.0328,
|
| 1329 |
+
"step": 177000
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 9.57246571659048,
|
| 1333 |
+
"grad_norm": 0.7965496182441711,
|
| 1334 |
+
"learning_rate": 5.706560753162695e-05,
|
| 1335 |
+
"loss": 3.024,
|
| 1336 |
+
"step": 178000
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 9.626243613874697,
|
| 1340 |
+
"grad_norm": 0.776597797870636,
|
| 1341 |
+
"learning_rate": 5.677169755810533e-05,
|
| 1342 |
+
"loss": 3.0312,
|
| 1343 |
+
"step": 179000
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 9.680021511158914,
|
| 1347 |
+
"grad_norm": 0.794489324092865,
|
| 1348 |
+
"learning_rate": 5.6477493380406e-05,
|
| 1349 |
+
"loss": 3.0328,
|
| 1350 |
+
"step": 180000
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 9.73379940844313,
|
| 1354 |
+
"grad_norm": 0.7984843254089355,
|
| 1355 |
+
"learning_rate": 5.6183583406884374e-05,
|
| 1356 |
+
"loss": 3.0325,
|
| 1357 |
+
"step": 181000
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 9.787577305727346,
|
| 1361 |
+
"grad_norm": 0.7846671342849731,
|
| 1362 |
+
"learning_rate": 5.5889379229185055e-05,
|
| 1363 |
+
"loss": 3.0315,
|
| 1364 |
+
"step": 182000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 9.841355203011563,
|
| 1368 |
+
"grad_norm": 0.7539902925491333,
|
| 1369 |
+
"learning_rate": 5.559546925566343e-05,
|
| 1370 |
+
"loss": 3.0388,
|
| 1371 |
+
"step": 183000
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 9.895133100295778,
|
| 1375 |
+
"grad_norm": 0.7702674865722656,
|
| 1376 |
+
"learning_rate": 5.530155928214181e-05,
|
| 1377 |
+
"loss": 3.0381,
|
| 1378 |
+
"step": 184000
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 9.948910997579995,
|
| 1382 |
+
"grad_norm": 0.7680040001869202,
|
| 1383 |
+
"learning_rate": 5.5007355104442484e-05,
|
| 1384 |
+
"loss": 3.0408,
|
| 1385 |
+
"step": 185000
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 10.0,
|
| 1389 |
+
"eval_accuracy": 0.4035088850726976,
|
| 1390 |
+
"eval_loss": 3.402043104171753,
|
| 1391 |
+
"eval_runtime": 155.8026,
|
| 1392 |
+
"eval_samples_per_second": 371.759,
|
| 1393 |
+
"eval_steps_per_second": 5.815,
|
| 1394 |
+
"step": 185950
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 10.00268889486421,
|
| 1398 |
+
"grad_norm": 0.7649341821670532,
|
| 1399 |
+
"learning_rate": 5.471315092674316e-05,
|
| 1400 |
+
"loss": 3.0367,
|
| 1401 |
+
"step": 186000
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 10.056466792148427,
|
| 1405 |
+
"grad_norm": 0.8047515153884888,
|
| 1406 |
+
"learning_rate": 5.441894674904384e-05,
|
| 1407 |
+
"loss": 2.9796,
|
| 1408 |
+
"step": 187000
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 10.110244689432642,
|
| 1412 |
+
"grad_norm": 0.8032823204994202,
|
| 1413 |
+
"learning_rate": 5.4124742571344514e-05,
|
| 1414 |
+
"loss": 2.9902,
|
| 1415 |
+
"step": 188000
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 10.16402258671686,
|
| 1419 |
+
"grad_norm": 0.7732874155044556,
|
| 1420 |
+
"learning_rate": 5.383083259782289e-05,
|
| 1421 |
+
"loss": 2.9933,
|
| 1422 |
+
"step": 189000
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 10.217800484001076,
|
| 1426 |
+
"grad_norm": 0.8030092716217041,
|
| 1427 |
+
"learning_rate": 5.3536628420123566e-05,
|
| 1428 |
+
"loss": 2.9927,
|
| 1429 |
+
"step": 190000
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 10.271578381285291,
|
| 1433 |
+
"grad_norm": 0.8287734389305115,
|
| 1434 |
+
"learning_rate": 5.3242718446601943e-05,
|
| 1435 |
+
"loss": 2.9934,
|
| 1436 |
+
"step": 191000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 10.325356278569508,
|
| 1440 |
+
"grad_norm": 0.824225664138794,
|
| 1441 |
+
"learning_rate": 5.294851426890262e-05,
|
| 1442 |
+
"loss": 2.9984,
|
| 1443 |
+
"step": 192000
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 10.379134175853725,
|
| 1447 |
+
"grad_norm": 0.8080607056617737,
|
| 1448 |
+
"learning_rate": 5.2654604295380995e-05,
|
| 1449 |
+
"loss": 2.9986,
|
| 1450 |
+
"step": 193000
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 10.43291207313794,
|
| 1454 |
+
"grad_norm": 0.8220816254615784,
|
| 1455 |
+
"learning_rate": 5.2360400117681677e-05,
|
| 1456 |
+
"loss": 3.0034,
|
| 1457 |
+
"step": 194000
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 10.486689970422157,
|
| 1461 |
+
"grad_norm": 0.77823805809021,
|
| 1462 |
+
"learning_rate": 5.206619593998235e-05,
|
| 1463 |
+
"loss": 3.0036,
|
| 1464 |
+
"step": 195000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 10.540467867706372,
|
| 1468 |
+
"grad_norm": 0.8018755912780762,
|
| 1469 |
+
"learning_rate": 5.177228596646072e-05,
|
| 1470 |
+
"loss": 3.0067,
|
| 1471 |
+
"step": 196000
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 10.594245764990589,
|
| 1475 |
+
"grad_norm": 0.794617235660553,
|
| 1476 |
+
"learning_rate": 5.14783759929391e-05,
|
| 1477 |
+
"loss": 3.005,
|
| 1478 |
+
"step": 197000
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 10.648023662274806,
|
| 1482 |
+
"grad_norm": 0.7715270519256592,
|
| 1483 |
+
"learning_rate": 5.118417181523978e-05,
|
| 1484 |
+
"loss": 3.0108,
|
| 1485 |
+
"step": 198000
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 10.701801559559021,
|
| 1489 |
+
"grad_norm": 0.7753629684448242,
|
| 1490 |
+
"learning_rate": 5.089026184171816e-05,
|
| 1491 |
+
"loss": 3.0053,
|
| 1492 |
+
"step": 199000
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 10.755579456843238,
|
| 1496 |
+
"grad_norm": 0.8104413151741028,
|
| 1497 |
+
"learning_rate": 5.059605766401884e-05,
|
| 1498 |
+
"loss": 3.0096,
|
| 1499 |
+
"step": 200000
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 10.809357354127453,
|
| 1503 |
+
"grad_norm": 0.7920235395431519,
|
| 1504 |
+
"learning_rate": 5.0301853486319506e-05,
|
| 1505 |
+
"loss": 3.0137,
|
| 1506 |
+
"step": 201000
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 10.86313525141167,
|
| 1510 |
+
"grad_norm": 0.8009334802627563,
|
| 1511 |
+
"learning_rate": 5.000764930862019e-05,
|
| 1512 |
+
"loss": 3.0097,
|
| 1513 |
+
"step": 202000
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 10.916913148695887,
|
| 1517 |
+
"grad_norm": 0.7775697708129883,
|
| 1518 |
+
"learning_rate": 4.971373933509856e-05,
|
| 1519 |
+
"loss": 3.0141,
|
| 1520 |
+
"step": 203000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 10.970691045980102,
|
| 1524 |
+
"grad_norm": 0.7922327518463135,
|
| 1525 |
+
"learning_rate": 4.941953515739924e-05,
|
| 1526 |
+
"loss": 3.0104,
|
| 1527 |
+
"step": 204000
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 11.0,
|
| 1531 |
+
"eval_accuracy": 0.403707330600643,
|
| 1532 |
+
"eval_loss": 3.396641731262207,
|
| 1533 |
+
"eval_runtime": 153.5711,
|
| 1534 |
+
"eval_samples_per_second": 377.161,
|
| 1535 |
+
"eval_steps_per_second": 5.9,
|
| 1536 |
+
"step": 204545
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 11.024468943264319,
|
| 1540 |
+
"grad_norm": 0.8196681141853333,
|
| 1541 |
+
"learning_rate": 4.9125625183877617e-05,
|
| 1542 |
+
"loss": 2.9858,
|
| 1543 |
+
"step": 205000
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 11.078246840548534,
|
| 1547 |
+
"grad_norm": 0.8191550374031067,
|
| 1548 |
+
"learning_rate": 4.883142100617829e-05,
|
| 1549 |
+
"loss": 2.9621,
|
| 1550 |
+
"step": 206000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 11.13202473783275,
|
| 1554 |
+
"grad_norm": 0.844635009765625,
|
| 1555 |
+
"learning_rate": 4.8537216828478965e-05,
|
| 1556 |
+
"loss": 2.9632,
|
| 1557 |
+
"step": 207000
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 11.185802635116968,
|
| 1561 |
+
"grad_norm": 0.82164466381073,
|
| 1562 |
+
"learning_rate": 4.824330685495734e-05,
|
| 1563 |
+
"loss": 2.967,
|
| 1564 |
+
"step": 208000
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 11.239580532401183,
|
| 1568 |
+
"grad_norm": 0.7903093099594116,
|
| 1569 |
+
"learning_rate": 4.794910267725802e-05,
|
| 1570 |
+
"loss": 2.9702,
|
| 1571 |
+
"step": 209000
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 11.2933584296854,
|
| 1575 |
+
"grad_norm": 0.8418600559234619,
|
| 1576 |
+
"learning_rate": 4.7655192703736395e-05,
|
| 1577 |
+
"loss": 2.9721,
|
| 1578 |
+
"step": 210000
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 11.347136326969615,
|
| 1582 |
+
"grad_norm": 0.8335950374603271,
|
| 1583 |
+
"learning_rate": 4.736098852603707e-05,
|
| 1584 |
+
"loss": 2.9771,
|
| 1585 |
+
"step": 211000
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 11.400914224253832,
|
| 1589 |
+
"grad_norm": 0.8250709176063538,
|
| 1590 |
+
"learning_rate": 4.706678434833775e-05,
|
| 1591 |
+
"loss": 2.9783,
|
| 1592 |
+
"step": 212000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 11.454692121538049,
|
| 1596 |
+
"grad_norm": 0.7945151925086975,
|
| 1597 |
+
"learning_rate": 4.677287437481612e-05,
|
| 1598 |
+
"loss": 2.9819,
|
| 1599 |
+
"step": 213000
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 11.508470018822264,
|
| 1603 |
+
"grad_norm": 0.8292239904403687,
|
| 1604 |
+
"learning_rate": 4.64786701971168e-05,
|
| 1605 |
+
"loss": 2.9843,
|
| 1606 |
+
"step": 214000
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 11.56224791610648,
|
| 1610 |
+
"grad_norm": 0.8527119159698486,
|
| 1611 |
+
"learning_rate": 4.618476022359517e-05,
|
| 1612 |
+
"loss": 2.9849,
|
| 1613 |
+
"step": 215000
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 11.616025813390696,
|
| 1617 |
+
"grad_norm": 0.8189669847488403,
|
| 1618 |
+
"learning_rate": 4.5890556045895854e-05,
|
| 1619 |
+
"loss": 2.985,
|
| 1620 |
+
"step": 216000
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 11.669803710674913,
|
| 1624 |
+
"grad_norm": 0.8017052412033081,
|
| 1625 |
+
"learning_rate": 4.5596646072374224e-05,
|
| 1626 |
+
"loss": 2.9876,
|
| 1627 |
+
"step": 217000
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 11.72358160795913,
|
| 1631 |
+
"grad_norm": 0.7937666773796082,
|
| 1632 |
+
"learning_rate": 4.5302441894674906e-05,
|
| 1633 |
+
"loss": 2.9847,
|
| 1634 |
+
"step": 218000
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 11.777359505243345,
|
| 1638 |
+
"grad_norm": 0.7971929907798767,
|
| 1639 |
+
"learning_rate": 4.500853192115328e-05,
|
| 1640 |
+
"loss": 2.9815,
|
| 1641 |
+
"step": 219000
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 11.831137402527562,
|
| 1645 |
+
"grad_norm": 0.8196897506713867,
|
| 1646 |
+
"learning_rate": 4.471432774345396e-05,
|
| 1647 |
+
"loss": 2.9876,
|
| 1648 |
+
"step": 220000
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 11.884915299811777,
|
| 1652 |
+
"grad_norm": 0.8172261118888855,
|
| 1653 |
+
"learning_rate": 4.442012356575463e-05,
|
| 1654 |
+
"loss": 2.9883,
|
| 1655 |
+
"step": 221000
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 11.938693197095994,
|
| 1659 |
+
"grad_norm": 0.8116560578346252,
|
| 1660 |
+
"learning_rate": 4.412621359223301e-05,
|
| 1661 |
+
"loss": 2.991,
|
| 1662 |
+
"step": 222000
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 11.99247109438021,
|
| 1666 |
+
"grad_norm": 0.7963495254516602,
|
| 1667 |
+
"learning_rate": 4.3832009414533684e-05,
|
| 1668 |
+
"loss": 2.9874,
|
| 1669 |
+
"step": 223000
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 12.0,
|
| 1673 |
+
"eval_accuracy": 0.4045469285045907,
|
| 1674 |
+
"eval_loss": 3.3944478034973145,
|
| 1675 |
+
"eval_runtime": 153.6841,
|
| 1676 |
+
"eval_samples_per_second": 376.884,
|
| 1677 |
+
"eval_steps_per_second": 5.895,
|
| 1678 |
+
"step": 223140
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 12.046248991664426,
|
| 1682 |
+
"grad_norm": 0.8429515957832336,
|
| 1683 |
+
"learning_rate": 4.353809944101207e-05,
|
| 1684 |
+
"loss": 2.9386,
|
| 1685 |
+
"step": 224000
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 12.100026888948642,
|
| 1689 |
+
"grad_norm": 0.8465261459350586,
|
| 1690 |
+
"learning_rate": 4.324389526331274e-05,
|
| 1691 |
+
"loss": 2.9401,
|
| 1692 |
+
"step": 225000
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 12.153804786232858,
|
| 1696 |
+
"grad_norm": 0.8318625092506409,
|
| 1697 |
+
"learning_rate": 4.2949691085613417e-05,
|
| 1698 |
+
"loss": 2.9449,
|
| 1699 |
+
"step": 226000
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 12.207582683517074,
|
| 1703 |
+
"grad_norm": 0.8261258602142334,
|
| 1704 |
+
"learning_rate": 4.2655781112091794e-05,
|
| 1705 |
+
"loss": 2.9509,
|
| 1706 |
+
"step": 227000
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 12.261360580801291,
|
| 1710 |
+
"grad_norm": 0.8545381426811218,
|
| 1711 |
+
"learning_rate": 4.2361576934392475e-05,
|
| 1712 |
+
"loss": 2.9501,
|
| 1713 |
+
"step": 228000
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 12.315138478085506,
|
| 1717 |
+
"grad_norm": 0.8169071674346924,
|
| 1718 |
+
"learning_rate": 4.2067666960870846e-05,
|
| 1719 |
+
"loss": 2.9563,
|
| 1720 |
+
"step": 229000
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 12.368916375369723,
|
| 1724 |
+
"grad_norm": 0.8455543518066406,
|
| 1725 |
+
"learning_rate": 4.177346278317153e-05,
|
| 1726 |
+
"loss": 2.9537,
|
| 1727 |
+
"step": 230000
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 12.422694272653938,
|
| 1731 |
+
"grad_norm": 0.8388088941574097,
|
| 1732 |
+
"learning_rate": 4.14792586054722e-05,
|
| 1733 |
+
"loss": 2.9562,
|
| 1734 |
+
"step": 231000
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 12.476472169938155,
|
| 1738 |
+
"grad_norm": 0.8068108558654785,
|
| 1739 |
+
"learning_rate": 4.118534863195058e-05,
|
| 1740 |
+
"loss": 2.9604,
|
| 1741 |
+
"step": 232000
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 12.530250067222372,
|
| 1745 |
+
"grad_norm": 0.8662741184234619,
|
| 1746 |
+
"learning_rate": 4.089114445425125e-05,
|
| 1747 |
+
"loss": 2.9605,
|
| 1748 |
+
"step": 233000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 12.584027964506587,
|
| 1752 |
+
"grad_norm": 0.8282158374786377,
|
| 1753 |
+
"learning_rate": 4.0596940276551934e-05,
|
| 1754 |
+
"loss": 2.9633,
|
| 1755 |
+
"step": 234000
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 12.637805861790804,
|
| 1759 |
+
"grad_norm": 0.8375250697135925,
|
| 1760 |
+
"learning_rate": 4.0303030303030305e-05,
|
| 1761 |
+
"loss": 2.9611,
|
| 1762 |
+
"step": 235000
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 12.69158375907502,
|
| 1766 |
+
"grad_norm": 0.8239476084709167,
|
| 1767 |
+
"learning_rate": 4.0008826125330986e-05,
|
| 1768 |
+
"loss": 2.9615,
|
| 1769 |
+
"step": 236000
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 12.745361656359236,
|
| 1773 |
+
"grad_norm": 0.8005510568618774,
|
| 1774 |
+
"learning_rate": 3.9714916151809357e-05,
|
| 1775 |
+
"loss": 2.9652,
|
| 1776 |
+
"step": 237000
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 12.799139553643453,
|
| 1780 |
+
"grad_norm": 0.8203298449516296,
|
| 1781 |
+
"learning_rate": 3.942071197411004e-05,
|
| 1782 |
+
"loss": 2.9669,
|
| 1783 |
+
"step": 238000
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 12.852917450927668,
|
| 1787 |
+
"grad_norm": 0.8380211591720581,
|
| 1788 |
+
"learning_rate": 3.912680200058841e-05,
|
| 1789 |
+
"loss": 2.967,
|
| 1790 |
+
"step": 239000
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 12.906695348211885,
|
| 1794 |
+
"grad_norm": 0.7954049706459045,
|
| 1795 |
+
"learning_rate": 3.883259782288909e-05,
|
| 1796 |
+
"loss": 2.9692,
|
| 1797 |
+
"step": 240000
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 12.9604732454961,
|
| 1801 |
+
"grad_norm": 0.8414211273193359,
|
| 1802 |
+
"learning_rate": 3.853868784936746e-05,
|
| 1803 |
+
"loss": 2.9712,
|
| 1804 |
+
"step": 241000
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 13.0,
|
| 1808 |
+
"eval_accuracy": 0.405673580484676,
|
| 1809 |
+
"eval_loss": 3.388187885284424,
|
| 1810 |
+
"eval_runtime": 154.3608,
|
| 1811 |
+
"eval_samples_per_second": 375.231,
|
| 1812 |
+
"eval_steps_per_second": 5.869,
|
| 1813 |
+
"step": 241735
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 13.014251142780317,
|
| 1817 |
+
"grad_norm": 0.8339217901229858,
|
| 1818 |
+
"learning_rate": 3.824448367166814e-05,
|
| 1819 |
+
"loss": 2.9554,
|
| 1820 |
+
"step": 242000
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 13.068029040064534,
|
| 1824 |
+
"grad_norm": 0.8675106763839722,
|
| 1825 |
+
"learning_rate": 3.7950279493968816e-05,
|
| 1826 |
+
"loss": 2.9241,
|
| 1827 |
+
"step": 243000
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 13.12180693734875,
|
| 1831 |
+
"grad_norm": 0.8612968921661377,
|
| 1832 |
+
"learning_rate": 3.765636952044719e-05,
|
| 1833 |
+
"loss": 2.9263,
|
| 1834 |
+
"step": 244000
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 13.175584834632966,
|
| 1838 |
+
"grad_norm": 0.841973602771759,
|
| 1839 |
+
"learning_rate": 3.736216534274787e-05,
|
| 1840 |
+
"loss": 2.9269,
|
| 1841 |
+
"step": 245000
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 13.229362731917181,
|
| 1845 |
+
"grad_norm": 0.83591628074646,
|
| 1846 |
+
"learning_rate": 3.7068255369226245e-05,
|
| 1847 |
+
"loss": 2.9287,
|
| 1848 |
+
"step": 246000
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 13.283140629201398,
|
| 1852 |
+
"grad_norm": 0.837383508682251,
|
| 1853 |
+
"learning_rate": 3.677405119152692e-05,
|
| 1854 |
+
"loss": 2.9326,
|
| 1855 |
+
"step": 247000
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 13.336918526485615,
|
| 1859 |
+
"grad_norm": 0.8568441867828369,
|
| 1860 |
+
"learning_rate": 3.64798470138276e-05,
|
| 1861 |
+
"loss": 2.9311,
|
| 1862 |
+
"step": 248000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 13.39069642376983,
|
| 1866 |
+
"grad_norm": 0.8351905345916748,
|
| 1867 |
+
"learning_rate": 3.618593704030597e-05,
|
| 1868 |
+
"loss": 2.9315,
|
| 1869 |
+
"step": 249000
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 13.444474321054047,
|
| 1873 |
+
"grad_norm": 0.8454658389091492,
|
| 1874 |
+
"learning_rate": 3.589173286260665e-05,
|
| 1875 |
+
"loss": 2.9354,
|
| 1876 |
+
"step": 250000
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 13.498252218338264,
|
| 1880 |
+
"grad_norm": 0.8542545437812805,
|
| 1881 |
+
"learning_rate": 3.559782288908502e-05,
|
| 1882 |
+
"loss": 2.9378,
|
| 1883 |
+
"step": 251000
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 13.552030115622479,
|
| 1887 |
+
"grad_norm": 0.8623389601707458,
|
| 1888 |
+
"learning_rate": 3.5303618711385704e-05,
|
| 1889 |
+
"loss": 2.9449,
|
| 1890 |
+
"step": 252000
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 13.605808012906696,
|
| 1894 |
+
"grad_norm": 0.8494634628295898,
|
| 1895 |
+
"learning_rate": 3.500941453368638e-05,
|
| 1896 |
+
"loss": 2.9448,
|
| 1897 |
+
"step": 253000
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 13.659585910190911,
|
| 1901 |
+
"grad_norm": 0.8554257750511169,
|
| 1902 |
+
"learning_rate": 3.471521035598705e-05,
|
| 1903 |
+
"loss": 2.9418,
|
| 1904 |
+
"step": 254000
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 13.713363807475128,
|
| 1908 |
+
"grad_norm": 0.8486588597297668,
|
| 1909 |
+
"learning_rate": 3.442159458664313e-05,
|
| 1910 |
+
"loss": 2.9472,
|
| 1911 |
+
"step": 255000
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 13.767141704759345,
|
| 1915 |
+
"grad_norm": 0.8385838270187378,
|
| 1916 |
+
"learning_rate": 3.412739040894381e-05,
|
| 1917 |
+
"loss": 2.9474,
|
| 1918 |
+
"step": 256000
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 13.82091960204356,
|
| 1922 |
+
"grad_norm": 0.820898711681366,
|
| 1923 |
+
"learning_rate": 3.383318623124448e-05,
|
| 1924 |
+
"loss": 2.9506,
|
| 1925 |
+
"step": 257000
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 13.874697499327777,
|
| 1929 |
+
"grad_norm": 0.8409541845321655,
|
| 1930 |
+
"learning_rate": 3.3538982053545157e-05,
|
| 1931 |
+
"loss": 2.9461,
|
| 1932 |
+
"step": 258000
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 13.928475396611992,
|
| 1936 |
+
"grad_norm": 0.8366772532463074,
|
| 1937 |
+
"learning_rate": 3.3245072080023534e-05,
|
| 1938 |
+
"loss": 2.9454,
|
| 1939 |
+
"step": 259000
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 13.982253293896209,
|
| 1943 |
+
"grad_norm": 0.8429400324821472,
|
| 1944 |
+
"learning_rate": 3.2950867902324215e-05,
|
| 1945 |
+
"loss": 2.9451,
|
| 1946 |
+
"step": 260000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 14.0,
|
| 1950 |
+
"eval_accuracy": 0.4058326593642206,
|
| 1951 |
+
"eval_loss": 3.3960049152374268,
|
| 1952 |
+
"eval_runtime": 153.9735,
|
| 1953 |
+
"eval_samples_per_second": 376.175,
|
| 1954 |
+
"eval_steps_per_second": 5.884,
|
| 1955 |
+
"step": 260330
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 14.036031191180426,
|
| 1959 |
+
"grad_norm": 0.8864750266075134,
|
| 1960 |
+
"learning_rate": 3.265666372462489e-05,
|
| 1961 |
+
"loss": 2.9184,
|
| 1962 |
+
"step": 261000
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 14.08980908846464,
|
| 1966 |
+
"grad_norm": 0.8641127347946167,
|
| 1967 |
+
"learning_rate": 3.236275375110327e-05,
|
| 1968 |
+
"loss": 2.9024,
|
| 1969 |
+
"step": 262000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 14.143586985748858,
|
| 1973 |
+
"grad_norm": 0.8877108097076416,
|
| 1974 |
+
"learning_rate": 3.206854957340394e-05,
|
| 1975 |
+
"loss": 2.9038,
|
| 1976 |
+
"step": 263000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 14.197364883033073,
|
| 1980 |
+
"grad_norm": 0.8459917902946472,
|
| 1981 |
+
"learning_rate": 3.1774345395704616e-05,
|
| 1982 |
+
"loss": 2.9074,
|
| 1983 |
+
"step": 264000
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 14.25114278031729,
|
| 1987 |
+
"grad_norm": 0.8829137086868286,
|
| 1988 |
+
"learning_rate": 3.14801412180053e-05,
|
| 1989 |
+
"loss": 2.9145,
|
| 1990 |
+
"step": 265000
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 14.304920677601507,
|
| 1994 |
+
"grad_norm": 0.8695775270462036,
|
| 1995 |
+
"learning_rate": 3.1186231244483674e-05,
|
| 1996 |
+
"loss": 2.9148,
|
| 1997 |
+
"step": 266000
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 14.358698574885722,
|
| 2001 |
+
"grad_norm": 0.8519364595413208,
|
| 2002 |
+
"learning_rate": 3.089202706678435e-05,
|
| 2003 |
+
"loss": 2.9146,
|
| 2004 |
+
"step": 267000
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 14.412476472169939,
|
| 2008 |
+
"grad_norm": 0.8653022050857544,
|
| 2009 |
+
"learning_rate": 3.059782288908502e-05,
|
| 2010 |
+
"loss": 2.9208,
|
| 2011 |
+
"step": 268000
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 14.466254369454154,
|
| 2015 |
+
"grad_norm": 0.8624246716499329,
|
| 2016 |
+
"learning_rate": 3.0303912915563404e-05,
|
| 2017 |
+
"loss": 2.918,
|
| 2018 |
+
"step": 269000
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 14.52003226673837,
|
| 2022 |
+
"grad_norm": 0.8843507170677185,
|
| 2023 |
+
"learning_rate": 3.000970873786408e-05,
|
| 2024 |
+
"loss": 2.9218,
|
| 2025 |
+
"step": 270000
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 14.573810164022587,
|
| 2029 |
+
"grad_norm": 0.8987634778022766,
|
| 2030 |
+
"learning_rate": 2.9715504560164753e-05,
|
| 2031 |
+
"loss": 2.9245,
|
| 2032 |
+
"step": 271000
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 14.627588061306803,
|
| 2036 |
+
"grad_norm": 0.8603494763374329,
|
| 2037 |
+
"learning_rate": 2.942159458664313e-05,
|
| 2038 |
+
"loss": 2.9243,
|
| 2039 |
+
"step": 272000
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 14.68136595859102,
|
| 2043 |
+
"grad_norm": 0.8882373571395874,
|
| 2044 |
+
"learning_rate": 2.912739040894381e-05,
|
| 2045 |
+
"loss": 2.9259,
|
| 2046 |
+
"step": 273000
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 14.735143855875235,
|
| 2050 |
+
"grad_norm": 0.8474394083023071,
|
| 2051 |
+
"learning_rate": 2.8833186231244486e-05,
|
| 2052 |
+
"loss": 2.9261,
|
| 2053 |
+
"step": 274000
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 14.788921753159451,
|
| 2057 |
+
"grad_norm": 0.8681209683418274,
|
| 2058 |
+
"learning_rate": 2.8539276257722863e-05,
|
| 2059 |
+
"loss": 2.9286,
|
| 2060 |
+
"step": 275000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 14.842699650443668,
|
| 2064 |
+
"grad_norm": 0.8800134658813477,
|
| 2065 |
+
"learning_rate": 2.8245072080023538e-05,
|
| 2066 |
+
"loss": 2.9283,
|
| 2067 |
+
"step": 276000
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 14.896477547727883,
|
| 2071 |
+
"grad_norm": 0.8928600549697876,
|
| 2072 |
+
"learning_rate": 2.7950867902324212e-05,
|
| 2073 |
+
"loss": 2.9311,
|
| 2074 |
+
"step": 277000
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 14.9502554450121,
|
| 2078 |
+
"grad_norm": 0.8456066846847534,
|
| 2079 |
+
"learning_rate": 2.765695792880259e-05,
|
| 2080 |
+
"loss": 2.9277,
|
| 2081 |
+
"step": 278000
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 15.0,
|
| 2085 |
+
"eval_accuracy": 0.4060881395073405,
|
| 2086 |
+
"eval_loss": 3.403709888458252,
|
| 2087 |
+
"eval_runtime": 153.7682,
|
| 2088 |
+
"eval_samples_per_second": 376.677,
|
| 2089 |
+
"eval_steps_per_second": 5.892,
|
| 2090 |
+
"step": 278925
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 15.004033342296315,
|
| 2094 |
+
"grad_norm": 0.8958522081375122,
|
| 2095 |
+
"learning_rate": 2.7362753751103264e-05,
|
| 2096 |
+
"loss": 2.9261,
|
| 2097 |
+
"step": 279000
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 15.057811239580532,
|
| 2101 |
+
"grad_norm": 0.8942525386810303,
|
| 2102 |
+
"learning_rate": 2.706884377758164e-05,
|
| 2103 |
+
"loss": 2.8847,
|
| 2104 |
+
"step": 280000
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 15.11158913686475,
|
| 2108 |
+
"grad_norm": 0.924169659614563,
|
| 2109 |
+
"learning_rate": 2.677463959988232e-05,
|
| 2110 |
+
"loss": 2.8916,
|
| 2111 |
+
"step": 281000
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 15.165367034148964,
|
| 2115 |
+
"grad_norm": 0.8767936825752258,
|
| 2116 |
+
"learning_rate": 2.6480435422182997e-05,
|
| 2117 |
+
"loss": 2.8917,
|
| 2118 |
+
"step": 282000
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 15.219144931433181,
|
| 2122 |
+
"grad_norm": 0.8910636901855469,
|
| 2123 |
+
"learning_rate": 2.618623124448367e-05,
|
| 2124 |
+
"loss": 2.8955,
|
| 2125 |
+
"step": 283000
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 15.272922828717396,
|
| 2129 |
+
"grad_norm": 0.893880307674408,
|
| 2130 |
+
"learning_rate": 2.589232127096205e-05,
|
| 2131 |
+
"loss": 2.8964,
|
| 2132 |
+
"step": 284000
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 15.326700726001613,
|
| 2136 |
+
"grad_norm": 0.8897101879119873,
|
| 2137 |
+
"learning_rate": 2.5598411297440423e-05,
|
| 2138 |
+
"loss": 2.9002,
|
| 2139 |
+
"step": 285000
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 15.38047862328583,
|
| 2143 |
+
"grad_norm": 0.8648462295532227,
|
| 2144 |
+
"learning_rate": 2.5304207119741104e-05,
|
| 2145 |
+
"loss": 2.9034,
|
| 2146 |
+
"step": 286000
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 15.434256520570045,
|
| 2150 |
+
"grad_norm": 0.8728475570678711,
|
| 2151 |
+
"learning_rate": 2.501000294204178e-05,
|
| 2152 |
+
"loss": 2.9022,
|
| 2153 |
+
"step": 287000
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 15.488034417854262,
|
| 2157 |
+
"grad_norm": 0.8891308903694153,
|
| 2158 |
+
"learning_rate": 2.4716092968520156e-05,
|
| 2159 |
+
"loss": 2.9033,
|
| 2160 |
+
"step": 288000
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 15.541812315138477,
|
| 2164 |
+
"grad_norm": 0.8900574445724487,
|
| 2165 |
+
"learning_rate": 2.4421888790820833e-05,
|
| 2166 |
+
"loss": 2.9037,
|
| 2167 |
+
"step": 289000
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 15.595590212422694,
|
| 2171 |
+
"grad_norm": 0.9271774291992188,
|
| 2172 |
+
"learning_rate": 2.4127684613121508e-05,
|
| 2173 |
+
"loss": 2.9059,
|
| 2174 |
+
"step": 290000
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 15.649368109706911,
|
| 2178 |
+
"grad_norm": 0.8639081716537476,
|
| 2179 |
+
"learning_rate": 2.3833774639599885e-05,
|
| 2180 |
+
"loss": 2.9027,
|
| 2181 |
+
"step": 291000
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 15.703146006991126,
|
| 2185 |
+
"grad_norm": 0.916532576084137,
|
| 2186 |
+
"learning_rate": 2.353957046190056e-05,
|
| 2187 |
+
"loss": 2.9057,
|
| 2188 |
+
"step": 292000
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 15.756923904275343,
|
| 2192 |
+
"grad_norm": 0.8904699087142944,
|
| 2193 |
+
"learning_rate": 2.3245366284201238e-05,
|
| 2194 |
+
"loss": 2.9099,
|
| 2195 |
+
"step": 293000
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 15.81070180155956,
|
| 2199 |
+
"grad_norm": 0.9009742736816406,
|
| 2200 |
+
"learning_rate": 2.2951162106501915e-05,
|
| 2201 |
+
"loss": 2.909,
|
| 2202 |
+
"step": 294000
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 15.864479698843775,
|
| 2206 |
+
"grad_norm": 0.9066869020462036,
|
| 2207 |
+
"learning_rate": 2.265695792880259e-05,
|
| 2208 |
+
"loss": 2.9111,
|
| 2209 |
+
"step": 295000
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 15.918257596127992,
|
| 2213 |
+
"grad_norm": 0.8992923498153687,
|
| 2214 |
+
"learning_rate": 2.2362753751103264e-05,
|
| 2215 |
+
"loss": 2.9099,
|
| 2216 |
+
"step": 296000
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 15.972035493412207,
|
| 2220 |
+
"grad_norm": 0.8612625002861023,
|
| 2221 |
+
"learning_rate": 2.2068843777581645e-05,
|
| 2222 |
+
"loss": 2.9085,
|
| 2223 |
+
"step": 297000
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 16.0,
|
| 2227 |
+
"eval_accuracy": 0.40623620109961933,
|
| 2228 |
+
"eval_loss": 3.4047741889953613,
|
| 2229 |
+
"eval_runtime": 153.8203,
|
| 2230 |
+
"eval_samples_per_second": 376.55,
|
| 2231 |
+
"eval_steps_per_second": 5.89,
|
| 2232 |
+
"step": 297520
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 16.025813390696424,
|
| 2236 |
+
"grad_norm": 0.917313277721405,
|
| 2237 |
+
"learning_rate": 2.177493380406002e-05,
|
| 2238 |
+
"loss": 2.8922,
|
| 2239 |
+
"step": 298000
|
| 2240 |
+
},
|
| 2241 |
+
{
|
| 2242 |
+
"epoch": 16.07959128798064,
|
| 2243 |
+
"grad_norm": 0.8836639523506165,
|
| 2244 |
+
"learning_rate": 2.1480729626360697e-05,
|
| 2245 |
+
"loss": 2.8792,
|
| 2246 |
+
"step": 299000
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 16.133369185264858,
|
| 2250 |
+
"grad_norm": 0.9119729995727539,
|
| 2251 |
+
"learning_rate": 2.118681965283907e-05,
|
| 2252 |
+
"loss": 2.8809,
|
| 2253 |
+
"step": 300000
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 16.18714708254907,
|
| 2257 |
+
"grad_norm": 0.9336438179016113,
|
| 2258 |
+
"learning_rate": 2.089261547513975e-05,
|
| 2259 |
+
"loss": 2.8775,
|
| 2260 |
+
"step": 301000
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 16.240924979833288,
|
| 2264 |
+
"grad_norm": 0.8615949153900146,
|
| 2265 |
+
"learning_rate": 2.0598411297440426e-05,
|
| 2266 |
+
"loss": 2.8817,
|
| 2267 |
+
"step": 302000
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 16.294702877117505,
|
| 2271 |
+
"grad_norm": 0.8850204944610596,
|
| 2272 |
+
"learning_rate": 2.03042071197411e-05,
|
| 2273 |
+
"loss": 2.8831,
|
| 2274 |
+
"step": 303000
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 16.34848077440172,
|
| 2278 |
+
"grad_norm": 0.912068784236908,
|
| 2279 |
+
"learning_rate": 2.0010297146219478e-05,
|
| 2280 |
+
"loss": 2.8818,
|
| 2281 |
+
"step": 304000
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 16.40225867168594,
|
| 2285 |
+
"grad_norm": 0.9005885124206543,
|
| 2286 |
+
"learning_rate": 1.9716092968520153e-05,
|
| 2287 |
+
"loss": 2.8822,
|
| 2288 |
+
"step": 305000
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 16.456036568970152,
|
| 2292 |
+
"grad_norm": 0.911561131477356,
|
| 2293 |
+
"learning_rate": 1.942188879082083e-05,
|
| 2294 |
+
"loss": 2.8862,
|
| 2295 |
+
"step": 306000
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 16.50981446625437,
|
| 2299 |
+
"grad_norm": 0.8996069431304932,
|
| 2300 |
+
"learning_rate": 1.9127978817299204e-05,
|
| 2301 |
+
"loss": 2.8877,
|
| 2302 |
+
"step": 307000
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 16.563592363538586,
|
| 2306 |
+
"grad_norm": 0.8976821899414062,
|
| 2307 |
+
"learning_rate": 1.8833774639599882e-05,
|
| 2308 |
+
"loss": 2.8872,
|
| 2309 |
+
"step": 308000
|
| 2310 |
+
},
|
| 2311 |
+
{
|
| 2312 |
+
"epoch": 16.617370260822803,
|
| 2313 |
+
"grad_norm": 0.8714480996131897,
|
| 2314 |
+
"learning_rate": 1.853957046190056e-05,
|
| 2315 |
+
"loss": 2.8852,
|
| 2316 |
+
"step": 309000
|
| 2317 |
+
},
|
| 2318 |
+
{
|
| 2319 |
+
"epoch": 16.67114815810702,
|
| 2320 |
+
"grad_norm": 0.9162812232971191,
|
| 2321 |
+
"learning_rate": 1.8245660488378934e-05,
|
| 2322 |
+
"loss": 2.8921,
|
| 2323 |
+
"step": 310000
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"epoch": 16.724926055391233,
|
| 2327 |
+
"grad_norm": 0.928915798664093,
|
| 2328 |
+
"learning_rate": 1.795175051485731e-05,
|
| 2329 |
+
"loss": 2.8913,
|
| 2330 |
+
"step": 311000
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 16.77870395267545,
|
| 2334 |
+
"grad_norm": 0.9134919047355652,
|
| 2335 |
+
"learning_rate": 1.7657546337157986e-05,
|
| 2336 |
+
"loss": 2.89,
|
| 2337 |
+
"step": 312000
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 16.832481849959667,
|
| 2341 |
+
"grad_norm": 0.9167847633361816,
|
| 2342 |
+
"learning_rate": 1.7363636363636366e-05,
|
| 2343 |
+
"loss": 2.8924,
|
| 2344 |
+
"step": 313000
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 16.886259747243884,
|
| 2348 |
+
"grad_norm": 0.9279534816741943,
|
| 2349 |
+
"learning_rate": 1.706943218593704e-05,
|
| 2350 |
+
"loss": 2.8927,
|
| 2351 |
+
"step": 314000
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 16.9400376445281,
|
| 2355 |
+
"grad_norm": 0.9238570928573608,
|
| 2356 |
+
"learning_rate": 1.677522800823772e-05,
|
| 2357 |
+
"loss": 2.8915,
|
| 2358 |
+
"step": 315000
|
| 2359 |
+
},
|
| 2360 |
+
{
|
| 2361 |
+
"epoch": 16.993815541812314,
|
| 2362 |
+
"grad_norm": 0.9144686460494995,
|
| 2363 |
+
"learning_rate": 1.6481023830538393e-05,
|
| 2364 |
+
"loss": 2.8914,
|
| 2365 |
+
"step": 316000
|
| 2366 |
+
},
|
| 2367 |
+
{
|
| 2368 |
+
"epoch": 17.0,
|
| 2369 |
+
"eval_accuracy": 0.4061480628014933,
|
| 2370 |
+
"eval_loss": 3.4033284187316895,
|
| 2371 |
+
"eval_runtime": 153.9786,
|
| 2372 |
+
"eval_samples_per_second": 376.163,
|
| 2373 |
+
"eval_steps_per_second": 5.884,
|
| 2374 |
+
"step": 316115
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 17.04759343909653,
|
| 2378 |
+
"grad_norm": 0.928446352481842,
|
| 2379 |
+
"learning_rate": 1.618681965283907e-05,
|
| 2380 |
+
"loss": 2.8629,
|
| 2381 |
+
"step": 317000
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 17.101371336380748,
|
| 2385 |
+
"grad_norm": 0.8849624395370483,
|
| 2386 |
+
"learning_rate": 1.589290967931745e-05,
|
| 2387 |
+
"loss": 2.8618,
|
| 2388 |
+
"step": 318000
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 17.155149233664964,
|
| 2392 |
+
"grad_norm": 0.9349420666694641,
|
| 2393 |
+
"learning_rate": 1.5598705501618123e-05,
|
| 2394 |
+
"loss": 2.8652,
|
| 2395 |
+
"step": 319000
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 17.20892713094918,
|
| 2399 |
+
"grad_norm": 0.931307315826416,
|
| 2400 |
+
"learning_rate": 1.53045013239188e-05,
|
| 2401 |
+
"loss": 2.8672,
|
| 2402 |
+
"step": 320000
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 17.262705028233395,
|
| 2406 |
+
"grad_norm": 0.9236338138580322,
|
| 2407 |
+
"learning_rate": 1.5010591350397176e-05,
|
| 2408 |
+
"loss": 2.8657,
|
| 2409 |
+
"step": 321000
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 17.31648292551761,
|
| 2413 |
+
"grad_norm": 0.9281163215637207,
|
| 2414 |
+
"learning_rate": 1.4716387172697854e-05,
|
| 2415 |
+
"loss": 2.8689,
|
| 2416 |
+
"step": 322000
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 17.37026082280183,
|
| 2420 |
+
"grad_norm": 0.9277437329292297,
|
| 2421 |
+
"learning_rate": 1.442247719917623e-05,
|
| 2422 |
+
"loss": 2.872,
|
| 2423 |
+
"step": 323000
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 17.424038720086045,
|
| 2427 |
+
"grad_norm": 0.9395264387130737,
|
| 2428 |
+
"learning_rate": 1.4128273021476906e-05,
|
| 2429 |
+
"loss": 2.8692,
|
| 2430 |
+
"step": 324000
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 17.477816617370262,
|
| 2434 |
+
"grad_norm": 0.9242090582847595,
|
| 2435 |
+
"learning_rate": 1.3834363047955282e-05,
|
| 2436 |
+
"loss": 2.8718,
|
| 2437 |
+
"step": 325000
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 17.531594514654476,
|
| 2441 |
+
"grad_norm": 0.9287353157997131,
|
| 2442 |
+
"learning_rate": 1.354015887025596e-05,
|
| 2443 |
+
"loss": 2.8684,
|
| 2444 |
+
"step": 326000
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 17.585372411938692,
|
| 2448 |
+
"grad_norm": 0.9129466414451599,
|
| 2449 |
+
"learning_rate": 1.3245954692556634e-05,
|
| 2450 |
+
"loss": 2.8724,
|
| 2451 |
+
"step": 327000
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 17.63915030922291,
|
| 2455 |
+
"grad_norm": 0.9155105352401733,
|
| 2456 |
+
"learning_rate": 1.2952044719035011e-05,
|
| 2457 |
+
"loss": 2.8735,
|
| 2458 |
+
"step": 328000
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 17.692928206507126,
|
| 2462 |
+
"grad_norm": 0.9174969792366028,
|
| 2463 |
+
"learning_rate": 1.2657840541335689e-05,
|
| 2464 |
+
"loss": 2.8747,
|
| 2465 |
+
"step": 329000
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 17.746706103791343,
|
| 2469 |
+
"grad_norm": 0.9102948307991028,
|
| 2470 |
+
"learning_rate": 1.2363930567814063e-05,
|
| 2471 |
+
"loss": 2.8736,
|
| 2472 |
+
"step": 330000
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 17.800484001075557,
|
| 2476 |
+
"grad_norm": 0.9251583218574524,
|
| 2477 |
+
"learning_rate": 1.2069726390114739e-05,
|
| 2478 |
+
"loss": 2.8737,
|
| 2479 |
+
"step": 331000
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 17.854261898359773,
|
| 2483 |
+
"grad_norm": 0.9378359913825989,
|
| 2484 |
+
"learning_rate": 1.1775522212415417e-05,
|
| 2485 |
+
"loss": 2.875,
|
| 2486 |
+
"step": 332000
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 17.90803979564399,
|
| 2490 |
+
"grad_norm": 0.9483941197395325,
|
| 2491 |
+
"learning_rate": 1.1481318034716093e-05,
|
| 2492 |
+
"loss": 2.8769,
|
| 2493 |
+
"step": 333000
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 17.961817692928207,
|
| 2497 |
+
"grad_norm": 0.9456818699836731,
|
| 2498 |
+
"learning_rate": 1.1187408061194469e-05,
|
| 2499 |
+
"loss": 2.8772,
|
| 2500 |
+
"step": 334000
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 18.0,
|
| 2504 |
+
"eval_accuracy": 0.40659050093522664,
|
| 2505 |
+
"eval_loss": 3.4093759059906006,
|
| 2506 |
+
"eval_runtime": 154.1524,
|
| 2507 |
+
"eval_samples_per_second": 375.739,
|
| 2508 |
+
"eval_steps_per_second": 5.877,
|
| 2509 |
+
"step": 334710
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 18.015595590212424,
|
| 2513 |
+
"grad_norm": 0.9420192837715149,
|
| 2514 |
+
"learning_rate": 1.0893203883495145e-05,
|
| 2515 |
+
"loss": 2.8719,
|
| 2516 |
+
"step": 335000
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 18.069373487496637,
|
| 2520 |
+
"grad_norm": 0.9638044238090515,
|
| 2521 |
+
"learning_rate": 1.0599293909973522e-05,
|
| 2522 |
+
"loss": 2.8533,
|
| 2523 |
+
"step": 336000
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 18.123151384780854,
|
| 2527 |
+
"grad_norm": 0.9174690246582031,
|
| 2528 |
+
"learning_rate": 1.03050897322742e-05,
|
| 2529 |
+
"loss": 2.8528,
|
| 2530 |
+
"step": 337000
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 18.17692928206507,
|
| 2534 |
+
"grad_norm": 0.9203698039054871,
|
| 2535 |
+
"learning_rate": 1.0010885554574875e-05,
|
| 2536 |
+
"loss": 2.8505,
|
| 2537 |
+
"step": 338000
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 18.230707179349288,
|
| 2541 |
+
"grad_norm": 0.9802068471908569,
|
| 2542 |
+
"learning_rate": 9.716975581053252e-06,
|
| 2543 |
+
"loss": 2.8494,
|
| 2544 |
+
"step": 339000
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 18.284485076633505,
|
| 2548 |
+
"grad_norm": 0.9169878959655762,
|
| 2549 |
+
"learning_rate": 9.422771403353928e-06,
|
| 2550 |
+
"loss": 2.8528,
|
| 2551 |
+
"step": 340000
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 18.33826297391772,
|
| 2555 |
+
"grad_norm": 0.9182587265968323,
|
| 2556 |
+
"learning_rate": 9.128861429832304e-06,
|
| 2557 |
+
"loss": 2.8588,
|
| 2558 |
+
"step": 341000
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 18.392040871201935,
|
| 2562 |
+
"grad_norm": 0.9429558515548706,
|
| 2563 |
+
"learning_rate": 8.83465725213298e-06,
|
| 2564 |
+
"loss": 2.8548,
|
| 2565 |
+
"step": 342000
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 18.445818768486152,
|
| 2569 |
+
"grad_norm": 0.943413496017456,
|
| 2570 |
+
"learning_rate": 8.540747278611357e-06,
|
| 2571 |
+
"loss": 2.8587,
|
| 2572 |
+
"step": 343000
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 18.49959666577037,
|
| 2576 |
+
"grad_norm": 0.952652096748352,
|
| 2577 |
+
"learning_rate": 8.246543100912033e-06,
|
| 2578 |
+
"loss": 2.8602,
|
| 2579 |
+
"step": 344000
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 18.553374563054586,
|
| 2583 |
+
"grad_norm": 0.9689589142799377,
|
| 2584 |
+
"learning_rate": 7.952633127390409e-06,
|
| 2585 |
+
"loss": 2.8597,
|
| 2586 |
+
"step": 345000
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 18.6071524603388,
|
| 2590 |
+
"grad_norm": 0.9562444686889648,
|
| 2591 |
+
"learning_rate": 7.658428949691087e-06,
|
| 2592 |
+
"loss": 2.8557,
|
| 2593 |
+
"step": 346000
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 18.660930357623016,
|
| 2597 |
+
"grad_norm": 0.9256601333618164,
|
| 2598 |
+
"learning_rate": 7.364224771991762e-06,
|
| 2599 |
+
"loss": 2.8595,
|
| 2600 |
+
"step": 347000
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 18.714708254907233,
|
| 2604 |
+
"grad_norm": 0.9548230767250061,
|
| 2605 |
+
"learning_rate": 7.07002059429244e-06,
|
| 2606 |
+
"loss": 2.8578,
|
| 2607 |
+
"step": 348000
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 18.76848615219145,
|
| 2611 |
+
"grad_norm": 0.9354435801506042,
|
| 2612 |
+
"learning_rate": 6.776110620770815e-06,
|
| 2613 |
+
"loss": 2.8583,
|
| 2614 |
+
"step": 349000
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 18.822264049475667,
|
| 2618 |
+
"grad_norm": 0.9104819893836975,
|
| 2619 |
+
"learning_rate": 6.4819064430714925e-06,
|
| 2620 |
+
"loss": 2.859,
|
| 2621 |
+
"step": 350000
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 18.87604194675988,
|
| 2625 |
+
"grad_norm": 0.9418288469314575,
|
| 2626 |
+
"learning_rate": 6.187702265372169e-06,
|
| 2627 |
+
"loss": 2.8582,
|
| 2628 |
+
"step": 351000
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 18.929819844044097,
|
| 2632 |
+
"grad_norm": 0.9473483562469482,
|
| 2633 |
+
"learning_rate": 5.893498087672846e-06,
|
| 2634 |
+
"loss": 2.8597,
|
| 2635 |
+
"step": 352000
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 18.983597741328314,
|
| 2639 |
+
"grad_norm": 0.9198222756385803,
|
| 2640 |
+
"learning_rate": 5.599588114151221e-06,
|
| 2641 |
+
"loss": 2.8635,
|
| 2642 |
+
"step": 353000
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 19.0,
|
| 2646 |
+
"eval_accuracy": 0.40674695985011655,
|
| 2647 |
+
"eval_loss": 3.4112322330474854,
|
| 2648 |
+
"eval_runtime": 153.9746,
|
| 2649 |
+
"eval_samples_per_second": 376.173,
|
| 2650 |
+
"eval_steps_per_second": 5.884,
|
| 2651 |
+
"step": 353305
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 19.03737563861253,
|
| 2655 |
+
"grad_norm": 0.9766749739646912,
|
| 2656 |
+
"learning_rate": 5.305678140629597e-06,
|
| 2657 |
+
"loss": 2.8485,
|
| 2658 |
+
"step": 354000
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 19.091153535896748,
|
| 2662 |
+
"grad_norm": 0.933742880821228,
|
| 2663 |
+
"learning_rate": 5.011473962930274e-06,
|
| 2664 |
+
"loss": 2.8427,
|
| 2665 |
+
"step": 355000
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 19.14493143318096,
|
| 2669 |
+
"grad_norm": 0.9536868333816528,
|
| 2670 |
+
"learning_rate": 4.71726978523095e-06,
|
| 2671 |
+
"loss": 2.8435,
|
| 2672 |
+
"step": 356000
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 19.198709330465178,
|
| 2676 |
+
"grad_norm": 0.9543789625167847,
|
| 2677 |
+
"learning_rate": 4.423065607531627e-06,
|
| 2678 |
+
"loss": 2.8441,
|
| 2679 |
+
"step": 357000
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 19.252487227749395,
|
| 2683 |
+
"grad_norm": 0.965768039226532,
|
| 2684 |
+
"learning_rate": 4.1291556340100035e-06,
|
| 2685 |
+
"loss": 2.843,
|
| 2686 |
+
"step": 358000
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 19.30626512503361,
|
| 2690 |
+
"grad_norm": 0.949124276638031,
|
| 2691 |
+
"learning_rate": 3.83495145631068e-06,
|
| 2692 |
+
"loss": 2.8449,
|
| 2693 |
+
"step": 359000
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 19.36004302231783,
|
| 2697 |
+
"grad_norm": 0.9277152419090271,
|
| 2698 |
+
"learning_rate": 3.5410414827890556e-06,
|
| 2699 |
+
"loss": 2.8438,
|
| 2700 |
+
"step": 360000
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 19.413820919602042,
|
| 2704 |
+
"grad_norm": 0.9696019887924194,
|
| 2705 |
+
"learning_rate": 3.2468373050897322e-06,
|
| 2706 |
+
"loss": 2.8438,
|
| 2707 |
+
"step": 361000
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 19.46759881688626,
|
| 2711 |
+
"grad_norm": 0.9465562701225281,
|
| 2712 |
+
"learning_rate": 2.9529273315681087e-06,
|
| 2713 |
+
"loss": 2.8402,
|
| 2714 |
+
"step": 362000
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 19.521376714170476,
|
| 2718 |
+
"grad_norm": 0.9613474607467651,
|
| 2719 |
+
"learning_rate": 2.6590173580464843e-06,
|
| 2720 |
+
"loss": 2.8449,
|
| 2721 |
+
"step": 363000
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 19.575154611454693,
|
| 2725 |
+
"grad_norm": 0.9396346211433411,
|
| 2726 |
+
"learning_rate": 2.364813180347161e-06,
|
| 2727 |
+
"loss": 2.8459,
|
| 2728 |
+
"step": 364000
|
| 2729 |
+
},
|
| 2730 |
+
{
|
| 2731 |
+
"epoch": 19.62893250873891,
|
| 2732 |
+
"grad_norm": 0.9625892043113708,
|
| 2733 |
+
"learning_rate": 2.0706090026478374e-06,
|
| 2734 |
+
"loss": 2.8453,
|
| 2735 |
+
"step": 365000
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 19.682710406023123,
|
| 2739 |
+
"grad_norm": 0.9505961537361145,
|
| 2740 |
+
"learning_rate": 1.7766990291262137e-06,
|
| 2741 |
+
"loss": 2.8473,
|
| 2742 |
+
"step": 366000
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 19.73648830330734,
|
| 2746 |
+
"grad_norm": 0.9687027335166931,
|
| 2747 |
+
"learning_rate": 1.4824948514268903e-06,
|
| 2748 |
+
"loss": 2.8413,
|
| 2749 |
+
"step": 367000
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 19.790266200591557,
|
| 2753 |
+
"grad_norm": 0.9597449898719788,
|
| 2754 |
+
"learning_rate": 1.188290673727567e-06,
|
| 2755 |
+
"loss": 2.8471,
|
| 2756 |
+
"step": 368000
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 19.844044097875774,
|
| 2760 |
+
"grad_norm": 0.9320659041404724,
|
| 2761 |
+
"learning_rate": 8.94380700205943e-07,
|
| 2762 |
+
"loss": 2.8465,
|
| 2763 |
+
"step": 369000
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 19.89782199515999,
|
| 2767 |
+
"grad_norm": 0.959924578666687,
|
| 2768 |
+
"learning_rate": 6.001765225066196e-07,
|
| 2769 |
+
"loss": 2.8409,
|
| 2770 |
+
"step": 370000
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 19.951599892444204,
|
| 2774 |
+
"grad_norm": 0.9497726559638977,
|
| 2775 |
+
"learning_rate": 3.0597234480729626e-07,
|
| 2776 |
+
"loss": 2.8506,
|
| 2777 |
+
"step": 371000
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 20.0,
|
| 2781 |
+
"eval_accuracy": 0.4066641958384616,
|
| 2782 |
+
"eval_loss": 3.41624116897583,
|
| 2783 |
+
"eval_runtime": 153.8342,
|
| 2784 |
+
"eval_samples_per_second": 376.516,
|
| 2785 |
+
"eval_steps_per_second": 5.889,
|
| 2786 |
+
"step": 371900
|
| 2787 |
+
},
|
| 2788 |
+
{
|
| 2789 |
+
"epoch": 20.0,
|
| 2790 |
+
"step": 371900,
|
| 2791 |
+
"total_flos": 1.5669257538816e+18,
|
| 2792 |
+
"train_loss": 3.1555080074711364,
|
| 2793 |
+
"train_runtime": 80953.5697,
|
| 2794 |
+
"train_samples_per_second": 147.006,
|
| 2795 |
+
"train_steps_per_second": 4.594
|
| 2796 |
+
}
|
| 2797 |
+
],
|
| 2798 |
+
"logging_steps": 1000,
|
| 2799 |
+
"max_steps": 371900,
|
| 2800 |
+
"num_input_tokens_seen": 0,
|
| 2801 |
+
"num_train_epochs": 20,
|
| 2802 |
+
"save_steps": 5000,
|
| 2803 |
+
"total_flos": 1.5669257538816e+18,
|
| 2804 |
+
"train_batch_size": 32,
|
| 2805 |
+
"trial_name": null,
|
| 2806 |
+
"trial_params": null
|
| 2807 |
+
}
|