End of training
Browse files- README.md +14 -2
- all_results.json +16 -0
- eval_results.json +10 -0
- tokenizer.json +1 -6
- train_results.json +9 -0
- trainer_state.json +945 -0
README.md
CHANGED
|
@@ -3,11 +3,23 @@ license: other
|
|
| 3 |
base_model: Qwen/Qwen1.5-4B
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
- name: lmind_nq_train6000_eval6489_v1_doc_qa_v3_Qwen_Qwen1.5-4B_5e-4_lora2
|
| 10 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
library_name: peft
|
| 12 |
---
|
| 13 |
|
|
@@ -16,7 +28,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# lmind_nq_train6000_eval6489_v1_doc_qa_v3_Qwen_Qwen1.5-4B_5e-4_lora2
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [Qwen/Qwen1.5-4B](https://huggingface.co/Qwen/Qwen1.5-4B) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 2.2417
|
| 22 |
- Accuracy: 0.5610
|
|
|
|
| 3 |
base_model: Qwen/Qwen1.5-4B
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- tyzhu/lmind_nq_train6000_eval6489_v1_doc_qa_v3
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
| 10 |
model-index:
|
| 11 |
- name: lmind_nq_train6000_eval6489_v1_doc_qa_v3_Qwen_Qwen1.5-4B_5e-4_lora2
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Causal Language Modeling
|
| 15 |
+
type: text-generation
|
| 16 |
+
dataset:
|
| 17 |
+
name: tyzhu/lmind_nq_train6000_eval6489_v1_doc_qa_v3
|
| 18 |
+
type: tyzhu/lmind_nq_train6000_eval6489_v1_doc_qa_v3
|
| 19 |
+
metrics:
|
| 20 |
+
- name: Accuracy
|
| 21 |
+
type: accuracy
|
| 22 |
+
value: 0.560974358974359
|
| 23 |
library_name: peft
|
| 24 |
---
|
| 25 |
|
|
|
|
| 28 |
|
| 29 |
# lmind_nq_train6000_eval6489_v1_doc_qa_v3_Qwen_Qwen1.5-4B_5e-4_lora2
|
| 30 |
|
| 31 |
+
This model is a fine-tuned version of [Qwen/Qwen1.5-4B](https://huggingface.co/Qwen/Qwen1.5-4B) on the tyzhu/lmind_nq_train6000_eval6489_v1_doc_qa_v3 dataset.
|
| 32 |
It achieves the following results on the evaluation set:
|
| 33 |
- Loss: 2.2417
|
| 34 |
- Accuracy: 0.5610
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.560974358974359,
|
| 4 |
+
"eval_loss": 2.2417104244232178,
|
| 5 |
+
"eval_runtime": 5.9255,
|
| 6 |
+
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 84.381,
|
| 8 |
+
"eval_steps_per_second": 10.632,
|
| 9 |
+
"perplexity": 9.409411618867681,
|
| 10 |
+
"total_flos": 6.517631969856061e+17,
|
| 11 |
+
"train_loss": 0.6977483630405708,
|
| 12 |
+
"train_runtime": 24342.3236,
|
| 13 |
+
"train_samples": 16925,
|
| 14 |
+
"train_samples_per_second": 13.906,
|
| 15 |
+
"train_steps_per_second": 0.435
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.560974358974359,
|
| 4 |
+
"eval_loss": 2.2417104244232178,
|
| 5 |
+
"eval_runtime": 5.9255,
|
| 6 |
+
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 84.381,
|
| 8 |
+
"eval_steps_per_second": 10.632,
|
| 9 |
+
"perplexity": 9.409411618867681
|
| 10 |
+
}
|
tokenizer.json
CHANGED
|
@@ -1,11 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
| 4 |
-
"direction": "Right",
|
| 5 |
-
"max_length": 1024,
|
| 6 |
-
"strategy": "LongestFirst",
|
| 7 |
-
"stride": 0
|
| 8 |
-
},
|
| 9 |
"padding": null,
|
| 10 |
"added_tokens": [
|
| 11 |
{
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
"padding": null,
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 6.517631969856061e+17,
|
| 4 |
+
"train_loss": 0.6977483630405708,
|
| 5 |
+
"train_runtime": 24342.3236,
|
| 6 |
+
"train_samples": 16925,
|
| 7 |
+
"train_samples_per_second": 13.906,
|
| 8 |
+
"train_steps_per_second": 0.435
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,945 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 10580,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1890359168241966,
|
| 13 |
+
"grad_norm": 0.45108333230018616,
|
| 14 |
+
"learning_rate": 0.0005,
|
| 15 |
+
"loss": 1.9159,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.3780718336483932,
|
| 20 |
+
"grad_norm": 0.6339304447174072,
|
| 21 |
+
"learning_rate": 0.0005,
|
| 22 |
+
"loss": 1.8756,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.5671077504725898,
|
| 27 |
+
"grad_norm": 0.5402090549468994,
|
| 28 |
+
"learning_rate": 0.0005,
|
| 29 |
+
"loss": 1.8511,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.7561436672967864,
|
| 34 |
+
"grad_norm": 0.43492260575294495,
|
| 35 |
+
"learning_rate": 0.0005,
|
| 36 |
+
"loss": 1.8581,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.945179584120983,
|
| 41 |
+
"grad_norm": 0.5317010879516602,
|
| 42 |
+
"learning_rate": 0.0005,
|
| 43 |
+
"loss": 1.8583,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.0,
|
| 48 |
+
"eval_accuracy": 0.5726153846153846,
|
| 49 |
+
"eval_loss": 1.6375669240951538,
|
| 50 |
+
"eval_runtime": 5.9206,
|
| 51 |
+
"eval_samples_per_second": 84.451,
|
| 52 |
+
"eval_steps_per_second": 10.641,
|
| 53 |
+
"step": 529
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.1342155009451795,
|
| 57 |
+
"grad_norm": 0.5848847031593323,
|
| 58 |
+
"learning_rate": 0.0005,
|
| 59 |
+
"loss": 1.6655,
|
| 60 |
+
"step": 600
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.3232514177693762,
|
| 64 |
+
"grad_norm": 0.642599880695343,
|
| 65 |
+
"learning_rate": 0.0005,
|
| 66 |
+
"loss": 1.581,
|
| 67 |
+
"step": 700
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.5122873345935728,
|
| 71 |
+
"grad_norm": 0.5563580393791199,
|
| 72 |
+
"learning_rate": 0.0005,
|
| 73 |
+
"loss": 1.5945,
|
| 74 |
+
"step": 800
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.7013232514177694,
|
| 78 |
+
"grad_norm": 0.576246976852417,
|
| 79 |
+
"learning_rate": 0.0005,
|
| 80 |
+
"loss": 1.5962,
|
| 81 |
+
"step": 900
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.8903591682419658,
|
| 85 |
+
"grad_norm": 0.6526830196380615,
|
| 86 |
+
"learning_rate": 0.0005,
|
| 87 |
+
"loss": 1.6329,
|
| 88 |
+
"step": 1000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 2.0,
|
| 92 |
+
"eval_accuracy": 0.5712820512820512,
|
| 93 |
+
"eval_loss": 1.6881264448165894,
|
| 94 |
+
"eval_runtime": 6.002,
|
| 95 |
+
"eval_samples_per_second": 83.305,
|
| 96 |
+
"eval_steps_per_second": 10.496,
|
| 97 |
+
"step": 1058
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 2.0793950850661624,
|
| 101 |
+
"grad_norm": 0.614975094795227,
|
| 102 |
+
"learning_rate": 0.0005,
|
| 103 |
+
"loss": 1.4729,
|
| 104 |
+
"step": 1100
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.268431001890359,
|
| 108 |
+
"grad_norm": 0.7056101560592651,
|
| 109 |
+
"learning_rate": 0.0005,
|
| 110 |
+
"loss": 1.2845,
|
| 111 |
+
"step": 1200
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 2.4574669187145557,
|
| 115 |
+
"grad_norm": 0.8371734619140625,
|
| 116 |
+
"learning_rate": 0.0005,
|
| 117 |
+
"loss": 1.3214,
|
| 118 |
+
"step": 1300
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.6465028355387523,
|
| 122 |
+
"grad_norm": 0.8234328627586365,
|
| 123 |
+
"learning_rate": 0.0005,
|
| 124 |
+
"loss": 1.3326,
|
| 125 |
+
"step": 1400
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.835538752362949,
|
| 129 |
+
"grad_norm": 0.7623058557510376,
|
| 130 |
+
"learning_rate": 0.0005,
|
| 131 |
+
"loss": 1.3464,
|
| 132 |
+
"step": 1500
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 3.0,
|
| 136 |
+
"eval_accuracy": 0.5663076923076923,
|
| 137 |
+
"eval_loss": 1.8256189823150635,
|
| 138 |
+
"eval_runtime": 5.9355,
|
| 139 |
+
"eval_samples_per_second": 84.239,
|
| 140 |
+
"eval_steps_per_second": 10.614,
|
| 141 |
+
"step": 1587
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 3.0245746691871456,
|
| 145 |
+
"grad_norm": 0.8697965145111084,
|
| 146 |
+
"learning_rate": 0.0005,
|
| 147 |
+
"loss": 1.3199,
|
| 148 |
+
"step": 1600
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.213610586011342,
|
| 152 |
+
"grad_norm": 0.7469469308853149,
|
| 153 |
+
"learning_rate": 0.0005,
|
| 154 |
+
"loss": 1.0191,
|
| 155 |
+
"step": 1700
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.402646502835539,
|
| 159 |
+
"grad_norm": 0.8800496459007263,
|
| 160 |
+
"learning_rate": 0.0005,
|
| 161 |
+
"loss": 1.0606,
|
| 162 |
+
"step": 1800
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.5916824196597354,
|
| 166 |
+
"grad_norm": 0.7127139568328857,
|
| 167 |
+
"learning_rate": 0.0005,
|
| 168 |
+
"loss": 1.0988,
|
| 169 |
+
"step": 1900
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.780718336483932,
|
| 173 |
+
"grad_norm": 0.9053574204444885,
|
| 174 |
+
"learning_rate": 0.0005,
|
| 175 |
+
"loss": 1.1355,
|
| 176 |
+
"step": 2000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.9697542533081287,
|
| 180 |
+
"grad_norm": 0.9217483997344971,
|
| 181 |
+
"learning_rate": 0.0005,
|
| 182 |
+
"loss": 1.1624,
|
| 183 |
+
"step": 2100
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.5651794871794872,
|
| 188 |
+
"eval_loss": 1.9222664833068848,
|
| 189 |
+
"eval_runtime": 5.7781,
|
| 190 |
+
"eval_samples_per_second": 86.533,
|
| 191 |
+
"eval_steps_per_second": 10.903,
|
| 192 |
+
"step": 2116
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 4.158790170132325,
|
| 196 |
+
"grad_norm": 0.8502592444419861,
|
| 197 |
+
"learning_rate": 0.0005,
|
| 198 |
+
"loss": 0.8795,
|
| 199 |
+
"step": 2200
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 4.3478260869565215,
|
| 203 |
+
"grad_norm": 0.8051011562347412,
|
| 204 |
+
"learning_rate": 0.0005,
|
| 205 |
+
"loss": 0.8713,
|
| 206 |
+
"step": 2300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.536862003780718,
|
| 210 |
+
"grad_norm": 1.0111184120178223,
|
| 211 |
+
"learning_rate": 0.0005,
|
| 212 |
+
"loss": 0.8974,
|
| 213 |
+
"step": 2400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.725897920604915,
|
| 217 |
+
"grad_norm": 0.9804710745811462,
|
| 218 |
+
"learning_rate": 0.0005,
|
| 219 |
+
"loss": 0.9396,
|
| 220 |
+
"step": 2500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 4.914933837429111,
|
| 224 |
+
"grad_norm": 1.1629979610443115,
|
| 225 |
+
"learning_rate": 0.0005,
|
| 226 |
+
"loss": 0.964,
|
| 227 |
+
"step": 2600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 5.0,
|
| 231 |
+
"eval_accuracy": 0.5642564102564103,
|
| 232 |
+
"eval_loss": 1.9719713926315308,
|
| 233 |
+
"eval_runtime": 6.1201,
|
| 234 |
+
"eval_samples_per_second": 81.698,
|
| 235 |
+
"eval_steps_per_second": 10.294,
|
| 236 |
+
"step": 2645
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 5.103969754253308,
|
| 240 |
+
"grad_norm": 0.8145145177841187,
|
| 241 |
+
"learning_rate": 0.0005,
|
| 242 |
+
"loss": 0.8208,
|
| 243 |
+
"step": 2700
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 5.293005671077505,
|
| 247 |
+
"grad_norm": 0.7975667715072632,
|
| 248 |
+
"learning_rate": 0.0005,
|
| 249 |
+
"loss": 0.7022,
|
| 250 |
+
"step": 2800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 5.482041587901701,
|
| 254 |
+
"grad_norm": 0.8971696496009827,
|
| 255 |
+
"learning_rate": 0.0005,
|
| 256 |
+
"loss": 0.7435,
|
| 257 |
+
"step": 2900
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 5.671077504725898,
|
| 261 |
+
"grad_norm": 1.0234569311141968,
|
| 262 |
+
"learning_rate": 0.0005,
|
| 263 |
+
"loss": 0.7798,
|
| 264 |
+
"step": 3000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 5.8601134215500945,
|
| 268 |
+
"grad_norm": 1.0058211088180542,
|
| 269 |
+
"learning_rate": 0.0005,
|
| 270 |
+
"loss": 0.8117,
|
| 271 |
+
"step": 3100
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 6.0,
|
| 275 |
+
"eval_accuracy": 0.5647179487179487,
|
| 276 |
+
"eval_loss": 2.001645565032959,
|
| 277 |
+
"eval_runtime": 6.1282,
|
| 278 |
+
"eval_samples_per_second": 81.59,
|
| 279 |
+
"eval_steps_per_second": 10.28,
|
| 280 |
+
"step": 3174
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 6.049149338374291,
|
| 284 |
+
"grad_norm": 0.9231036901473999,
|
| 285 |
+
"learning_rate": 0.0005,
|
| 286 |
+
"loss": 0.7673,
|
| 287 |
+
"step": 3200
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 6.238185255198488,
|
| 291 |
+
"grad_norm": 0.9058907628059387,
|
| 292 |
+
"learning_rate": 0.0005,
|
| 293 |
+
"loss": 0.5828,
|
| 294 |
+
"step": 3300
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 6.427221172022684,
|
| 298 |
+
"grad_norm": 0.8709693551063538,
|
| 299 |
+
"learning_rate": 0.0005,
|
| 300 |
+
"loss": 0.6272,
|
| 301 |
+
"step": 3400
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 6.616257088846881,
|
| 305 |
+
"grad_norm": 0.9873590469360352,
|
| 306 |
+
"learning_rate": 0.0005,
|
| 307 |
+
"loss": 0.6599,
|
| 308 |
+
"step": 3500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 6.805293005671078,
|
| 312 |
+
"grad_norm": 0.9257177114486694,
|
| 313 |
+
"learning_rate": 0.0005,
|
| 314 |
+
"loss": 0.6918,
|
| 315 |
+
"step": 3600
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 6.994328922495274,
|
| 319 |
+
"grad_norm": 1.0284956693649292,
|
| 320 |
+
"learning_rate": 0.0005,
|
| 321 |
+
"loss": 0.7242,
|
| 322 |
+
"step": 3700
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 7.0,
|
| 326 |
+
"eval_accuracy": 0.5638974358974359,
|
| 327 |
+
"eval_loss": 2.0784664154052734,
|
| 328 |
+
"eval_runtime": 6.1613,
|
| 329 |
+
"eval_samples_per_second": 81.152,
|
| 330 |
+
"eval_steps_per_second": 10.225,
|
| 331 |
+
"step": 3703
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 7.183364839319471,
|
| 335 |
+
"grad_norm": 1.0036342144012451,
|
| 336 |
+
"learning_rate": 0.0005,
|
| 337 |
+
"loss": 0.5099,
|
| 338 |
+
"step": 3800
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 7.3724007561436675,
|
| 342 |
+
"grad_norm": 0.9070968627929688,
|
| 343 |
+
"learning_rate": 0.0005,
|
| 344 |
+
"loss": 0.541,
|
| 345 |
+
"step": 3900
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 7.561436672967864,
|
| 349 |
+
"grad_norm": 1.035317063331604,
|
| 350 |
+
"learning_rate": 0.0005,
|
| 351 |
+
"loss": 0.5681,
|
| 352 |
+
"step": 4000
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.750472589792061,
|
| 356 |
+
"grad_norm": 1.0337880849838257,
|
| 357 |
+
"learning_rate": 0.0005,
|
| 358 |
+
"loss": 0.5991,
|
| 359 |
+
"step": 4100
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 7.939508506616257,
|
| 363 |
+
"grad_norm": 0.9707123637199402,
|
| 364 |
+
"learning_rate": 0.0005,
|
| 365 |
+
"loss": 0.6381,
|
| 366 |
+
"step": 4200
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 8.0,
|
| 370 |
+
"eval_accuracy": 0.5644615384615385,
|
| 371 |
+
"eval_loss": 2.0954012870788574,
|
| 372 |
+
"eval_runtime": 5.9732,
|
| 373 |
+
"eval_samples_per_second": 83.708,
|
| 374 |
+
"eval_steps_per_second": 10.547,
|
| 375 |
+
"step": 4232
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 8.128544423440454,
|
| 379 |
+
"grad_norm": 0.9024428129196167,
|
| 380 |
+
"learning_rate": 0.0005,
|
| 381 |
+
"loss": 0.5113,
|
| 382 |
+
"step": 4300
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 8.31758034026465,
|
| 386 |
+
"grad_norm": 0.8392314314842224,
|
| 387 |
+
"learning_rate": 0.0005,
|
| 388 |
+
"loss": 0.4752,
|
| 389 |
+
"step": 4400
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 8.506616257088847,
|
| 393 |
+
"grad_norm": 0.9310020208358765,
|
| 394 |
+
"learning_rate": 0.0005,
|
| 395 |
+
"loss": 0.5088,
|
| 396 |
+
"step": 4500
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 8.695652173913043,
|
| 400 |
+
"grad_norm": 1.115989327430725,
|
| 401 |
+
"learning_rate": 0.0005,
|
| 402 |
+
"loss": 0.5358,
|
| 403 |
+
"step": 4600
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 8.88468809073724,
|
| 407 |
+
"grad_norm": 1.1282142400741577,
|
| 408 |
+
"learning_rate": 0.0005,
|
| 409 |
+
"loss": 0.573,
|
| 410 |
+
"step": 4700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 9.0,
|
| 414 |
+
"eval_accuracy": 0.5623076923076923,
|
| 415 |
+
"eval_loss": 2.106734275817871,
|
| 416 |
+
"eval_runtime": 5.9846,
|
| 417 |
+
"eval_samples_per_second": 83.548,
|
| 418 |
+
"eval_steps_per_second": 10.527,
|
| 419 |
+
"step": 4761
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 9.073724007561436,
|
| 423 |
+
"grad_norm": 0.7718724012374878,
|
| 424 |
+
"learning_rate": 0.0005,
|
| 425 |
+
"loss": 0.5165,
|
| 426 |
+
"step": 4800
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 9.262759924385634,
|
| 430 |
+
"grad_norm": 1.0482347011566162,
|
| 431 |
+
"learning_rate": 0.0005,
|
| 432 |
+
"loss": 0.4291,
|
| 433 |
+
"step": 4900
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 9.45179584120983,
|
| 437 |
+
"grad_norm": 1.1258679628372192,
|
| 438 |
+
"learning_rate": 0.0005,
|
| 439 |
+
"loss": 0.467,
|
| 440 |
+
"step": 5000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 9.640831758034027,
|
| 444 |
+
"grad_norm": 1.0340867042541504,
|
| 445 |
+
"learning_rate": 0.0005,
|
| 446 |
+
"loss": 0.4992,
|
| 447 |
+
"step": 5100
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 9.829867674858223,
|
| 451 |
+
"grad_norm": 0.9999015927314758,
|
| 452 |
+
"learning_rate": 0.0005,
|
| 453 |
+
"loss": 0.5269,
|
| 454 |
+
"step": 5200
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 10.0,
|
| 458 |
+
"eval_accuracy": 0.5646153846153846,
|
| 459 |
+
"eval_loss": 2.1355865001678467,
|
| 460 |
+
"eval_runtime": 5.6591,
|
| 461 |
+
"eval_samples_per_second": 88.353,
|
| 462 |
+
"eval_steps_per_second": 11.132,
|
| 463 |
+
"step": 5290
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 10.01890359168242,
|
| 467 |
+
"grad_norm": 0.8213431239128113,
|
| 468 |
+
"learning_rate": 0.0005,
|
| 469 |
+
"loss": 0.5325,
|
| 470 |
+
"step": 5300
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 10.207939508506616,
|
| 474 |
+
"grad_norm": 1.023992896080017,
|
| 475 |
+
"learning_rate": 0.0005,
|
| 476 |
+
"loss": 0.3966,
|
| 477 |
+
"step": 5400
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 10.396975425330814,
|
| 481 |
+
"grad_norm": 0.8211169242858887,
|
| 482 |
+
"learning_rate": 0.0005,
|
| 483 |
+
"loss": 0.4307,
|
| 484 |
+
"step": 5500
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 10.58601134215501,
|
| 488 |
+
"grad_norm": 0.9888399839401245,
|
| 489 |
+
"learning_rate": 0.0005,
|
| 490 |
+
"loss": 0.4604,
|
| 491 |
+
"step": 5600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 10.775047258979207,
|
| 495 |
+
"grad_norm": 0.9565384387969971,
|
| 496 |
+
"learning_rate": 0.0005,
|
| 497 |
+
"loss": 0.4888,
|
| 498 |
+
"step": 5700
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 10.964083175803403,
|
| 502 |
+
"grad_norm": 1.0463844537734985,
|
| 503 |
+
"learning_rate": 0.0005,
|
| 504 |
+
"loss": 0.5144,
|
| 505 |
+
"step": 5800
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 11.0,
|
| 509 |
+
"eval_accuracy": 0.5616410256410257,
|
| 510 |
+
"eval_loss": 2.195107936859131,
|
| 511 |
+
"eval_runtime": 5.722,
|
| 512 |
+
"eval_samples_per_second": 87.382,
|
| 513 |
+
"eval_steps_per_second": 11.01,
|
| 514 |
+
"step": 5819
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 11.1531190926276,
|
| 518 |
+
"grad_norm": 0.9159551858901978,
|
| 519 |
+
"learning_rate": 0.0005,
|
| 520 |
+
"loss": 0.4094,
|
| 521 |
+
"step": 5900
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 11.342155009451796,
|
| 525 |
+
"grad_norm": 0.9027566909790039,
|
| 526 |
+
"learning_rate": 0.0005,
|
| 527 |
+
"loss": 0.4083,
|
| 528 |
+
"step": 6000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 11.531190926275993,
|
| 532 |
+
"grad_norm": 0.9325262904167175,
|
| 533 |
+
"learning_rate": 0.0005,
|
| 534 |
+
"loss": 0.4339,
|
| 535 |
+
"step": 6100
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 11.720226843100189,
|
| 539 |
+
"grad_norm": 1.0173691511154175,
|
| 540 |
+
"learning_rate": 0.0005,
|
| 541 |
+
"loss": 0.464,
|
| 542 |
+
"step": 6200
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 11.909262759924385,
|
| 546 |
+
"grad_norm": 0.9742277264595032,
|
| 547 |
+
"learning_rate": 0.0005,
|
| 548 |
+
"loss": 0.4887,
|
| 549 |
+
"step": 6300
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 12.0,
|
| 553 |
+
"eval_accuracy": 0.5631282051282052,
|
| 554 |
+
"eval_loss": 2.1778857707977295,
|
| 555 |
+
"eval_runtime": 6.0538,
|
| 556 |
+
"eval_samples_per_second": 82.593,
|
| 557 |
+
"eval_steps_per_second": 10.407,
|
| 558 |
+
"step": 6348
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 12.098298676748582,
|
| 562 |
+
"grad_norm": 1.0347404479980469,
|
| 563 |
+
"learning_rate": 0.0005,
|
| 564 |
+
"loss": 0.438,
|
| 565 |
+
"step": 6400
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 12.287334593572778,
|
| 569 |
+
"grad_norm": 0.8733549118041992,
|
| 570 |
+
"learning_rate": 0.0005,
|
| 571 |
+
"loss": 0.3893,
|
| 572 |
+
"step": 6500
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 12.476370510396976,
|
| 576 |
+
"grad_norm": 0.8693263530731201,
|
| 577 |
+
"learning_rate": 0.0005,
|
| 578 |
+
"loss": 0.4142,
|
| 579 |
+
"step": 6600
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 12.665406427221171,
|
| 583 |
+
"grad_norm": 1.1597702503204346,
|
| 584 |
+
"learning_rate": 0.0005,
|
| 585 |
+
"loss": 0.4379,
|
| 586 |
+
"step": 6700
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 12.854442344045369,
|
| 590 |
+
"grad_norm": 1.018873929977417,
|
| 591 |
+
"learning_rate": 0.0005,
|
| 592 |
+
"loss": 0.4636,
|
| 593 |
+
"step": 6800
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 13.0,
|
| 597 |
+
"eval_accuracy": 0.561076923076923,
|
| 598 |
+
"eval_loss": 2.175729274749756,
|
| 599 |
+
"eval_runtime": 5.683,
|
| 600 |
+
"eval_samples_per_second": 87.981,
|
| 601 |
+
"eval_steps_per_second": 11.086,
|
| 602 |
+
"step": 6877
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 13.043478260869565,
|
| 606 |
+
"grad_norm": 0.8772552013397217,
|
| 607 |
+
"learning_rate": 0.0005,
|
| 608 |
+
"loss": 0.4552,
|
| 609 |
+
"step": 6900
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 13.232514177693762,
|
| 613 |
+
"grad_norm": 0.8089916110038757,
|
| 614 |
+
"learning_rate": 0.0005,
|
| 615 |
+
"loss": 0.3698,
|
| 616 |
+
"step": 7000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 13.421550094517958,
|
| 620 |
+
"grad_norm": 1.0659642219543457,
|
| 621 |
+
"learning_rate": 0.0005,
|
| 622 |
+
"loss": 0.3971,
|
| 623 |
+
"step": 7100
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 13.610586011342155,
|
| 627 |
+
"grad_norm": 1.2195113897323608,
|
| 628 |
+
"learning_rate": 0.0005,
|
| 629 |
+
"loss": 0.4213,
|
| 630 |
+
"step": 7200
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 13.799621928166351,
|
| 634 |
+
"grad_norm": 1.1026618480682373,
|
| 635 |
+
"learning_rate": 0.0005,
|
| 636 |
+
"loss": 0.4487,
|
| 637 |
+
"step": 7300
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 13.988657844990549,
|
| 641 |
+
"grad_norm": 1.120952844619751,
|
| 642 |
+
"learning_rate": 0.0005,
|
| 643 |
+
"loss": 0.467,
|
| 644 |
+
"step": 7400
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 14.0,
|
| 648 |
+
"eval_accuracy": 0.5624102564102564,
|
| 649 |
+
"eval_loss": 2.1781177520751953,
|
| 650 |
+
"eval_runtime": 5.7664,
|
| 651 |
+
"eval_samples_per_second": 86.709,
|
| 652 |
+
"eval_steps_per_second": 10.925,
|
| 653 |
+
"step": 7406
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 14.177693761814744,
|
| 657 |
+
"grad_norm": 0.9397806525230408,
|
| 658 |
+
"learning_rate": 0.0005,
|
| 659 |
+
"loss": 0.3685,
|
| 660 |
+
"step": 7500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 14.366729678638942,
|
| 664 |
+
"grad_norm": 0.8353122472763062,
|
| 665 |
+
"learning_rate": 0.0005,
|
| 666 |
+
"loss": 0.3828,
|
| 667 |
+
"step": 7600
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 14.555765595463138,
|
| 671 |
+
"grad_norm": 1.0425219535827637,
|
| 672 |
+
"learning_rate": 0.0005,
|
| 673 |
+
"loss": 0.4026,
|
| 674 |
+
"step": 7700
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 14.744801512287335,
|
| 678 |
+
"grad_norm": 1.0660319328308105,
|
| 679 |
+
"learning_rate": 0.0005,
|
| 680 |
+
"loss": 0.4287,
|
| 681 |
+
"step": 7800
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 14.93383742911153,
|
| 685 |
+
"grad_norm": 1.131705403327942,
|
| 686 |
+
"learning_rate": 0.0005,
|
| 687 |
+
"loss": 0.4613,
|
| 688 |
+
"step": 7900
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 15.0,
|
| 692 |
+
"eval_accuracy": 0.5611794871794872,
|
| 693 |
+
"eval_loss": 2.231194019317627,
|
| 694 |
+
"eval_runtime": 5.8033,
|
| 695 |
+
"eval_samples_per_second": 86.158,
|
| 696 |
+
"eval_steps_per_second": 10.856,
|
| 697 |
+
"step": 7935
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 15.122873345935728,
|
| 701 |
+
"grad_norm": 0.8532615303993225,
|
| 702 |
+
"learning_rate": 0.0005,
|
| 703 |
+
"loss": 0.3889,
|
| 704 |
+
"step": 8000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 15.311909262759924,
|
| 708 |
+
"grad_norm": 0.9682427048683167,
|
| 709 |
+
"learning_rate": 0.0005,
|
| 710 |
+
"loss": 0.3675,
|
| 711 |
+
"step": 8100
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 15.500945179584122,
|
| 715 |
+
"grad_norm": 0.9799471497535706,
|
| 716 |
+
"learning_rate": 0.0005,
|
| 717 |
+
"loss": 0.3968,
|
| 718 |
+
"step": 8200
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 15.689981096408317,
|
| 722 |
+
"grad_norm": 0.9278863072395325,
|
| 723 |
+
"learning_rate": 0.0005,
|
| 724 |
+
"loss": 0.4198,
|
| 725 |
+
"step": 8300
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 15.879017013232515,
|
| 729 |
+
"grad_norm": 1.0345618724822998,
|
| 730 |
+
"learning_rate": 0.0005,
|
| 731 |
+
"loss": 0.4405,
|
| 732 |
+
"step": 8400
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 16.0,
|
| 736 |
+
"eval_accuracy": 0.562923076923077,
|
| 737 |
+
"eval_loss": 2.1799755096435547,
|
| 738 |
+
"eval_runtime": 5.8871,
|
| 739 |
+
"eval_samples_per_second": 84.931,
|
| 740 |
+
"eval_steps_per_second": 10.701,
|
| 741 |
+
"step": 8464
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 16.068052930056712,
|
| 745 |
+
"grad_norm": 0.9081653356552124,
|
| 746 |
+
"learning_rate": 0.0005,
|
| 747 |
+
"loss": 0.4241,
|
| 748 |
+
"step": 8500
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 16.257088846880908,
|
| 752 |
+
"grad_norm": 0.8579433560371399,
|
| 753 |
+
"learning_rate": 0.0005,
|
| 754 |
+
"loss": 0.3634,
|
| 755 |
+
"step": 8600
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 16.446124763705104,
|
| 759 |
+
"grad_norm": 1.042361855506897,
|
| 760 |
+
"learning_rate": 0.0005,
|
| 761 |
+
"loss": 0.3837,
|
| 762 |
+
"step": 8700
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 16.6351606805293,
|
| 766 |
+
"grad_norm": 1.032463550567627,
|
| 767 |
+
"learning_rate": 0.0005,
|
| 768 |
+
"loss": 0.4071,
|
| 769 |
+
"step": 8800
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 16.8241965973535,
|
| 773 |
+
"grad_norm": 1.0003693103790283,
|
| 774 |
+
"learning_rate": 0.0005,
|
| 775 |
+
"loss": 0.4308,
|
| 776 |
+
"step": 8900
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 17.0,
|
| 780 |
+
"eval_accuracy": 0.5627692307692308,
|
| 781 |
+
"eval_loss": 2.1960062980651855,
|
| 782 |
+
"eval_runtime": 6.0711,
|
| 783 |
+
"eval_samples_per_second": 82.358,
|
| 784 |
+
"eval_steps_per_second": 10.377,
|
| 785 |
+
"step": 8993
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 17.013232514177695,
|
| 789 |
+
"grad_norm": 0.793249249458313,
|
| 790 |
+
"learning_rate": 0.0005,
|
| 791 |
+
"loss": 0.4471,
|
| 792 |
+
"step": 9000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 17.20226843100189,
|
| 796 |
+
"grad_norm": 0.9027940630912781,
|
| 797 |
+
"learning_rate": 0.0005,
|
| 798 |
+
"loss": 0.3499,
|
| 799 |
+
"step": 9100
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 17.391304347826086,
|
| 803 |
+
"grad_norm": 0.9174290299415588,
|
| 804 |
+
"learning_rate": 0.0005,
|
| 805 |
+
"loss": 0.3698,
|
| 806 |
+
"step": 9200
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 17.58034026465028,
|
| 810 |
+
"grad_norm": 1.135392665863037,
|
| 811 |
+
"learning_rate": 0.0005,
|
| 812 |
+
"loss": 0.4027,
|
| 813 |
+
"step": 9300
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 17.76937618147448,
|
| 817 |
+
"grad_norm": 0.9852614998817444,
|
| 818 |
+
"learning_rate": 0.0005,
|
| 819 |
+
"loss": 0.4179,
|
| 820 |
+
"step": 9400
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 17.958412098298677,
|
| 824 |
+
"grad_norm": 0.9487043619155884,
|
| 825 |
+
"learning_rate": 0.0005,
|
| 826 |
+
"loss": 0.4401,
|
| 827 |
+
"step": 9500
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 18.0,
|
| 831 |
+
"eval_accuracy": 0.560974358974359,
|
| 832 |
+
"eval_loss": 2.2354702949523926,
|
| 833 |
+
"eval_runtime": 5.7571,
|
| 834 |
+
"eval_samples_per_second": 86.85,
|
| 835 |
+
"eval_steps_per_second": 10.943,
|
| 836 |
+
"step": 9522
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 18.147448015122873,
|
| 840 |
+
"grad_norm": 0.7868739366531372,
|
| 841 |
+
"learning_rate": 0.0005,
|
| 842 |
+
"loss": 0.3637,
|
| 843 |
+
"step": 9600
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 18.33648393194707,
|
| 847 |
+
"grad_norm": 0.962329626083374,
|
| 848 |
+
"learning_rate": 0.0005,
|
| 849 |
+
"loss": 0.3642,
|
| 850 |
+
"step": 9700
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 18.525519848771268,
|
| 854 |
+
"grad_norm": 0.9594590663909912,
|
| 855 |
+
"learning_rate": 0.0005,
|
| 856 |
+
"loss": 0.3865,
|
| 857 |
+
"step": 9800
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 18.714555765595463,
|
| 861 |
+
"grad_norm": 1.0574795007705688,
|
| 862 |
+
"learning_rate": 0.0005,
|
| 863 |
+
"loss": 0.4081,
|
| 864 |
+
"step": 9900
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 18.90359168241966,
|
| 868 |
+
"grad_norm": 1.0566685199737549,
|
| 869 |
+
"learning_rate": 0.0005,
|
| 870 |
+
"loss": 0.4334,
|
| 871 |
+
"step": 10000
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 19.0,
|
| 875 |
+
"eval_accuracy": 0.5608205128205128,
|
| 876 |
+
"eval_loss": 2.2380332946777344,
|
| 877 |
+
"eval_runtime": 5.7808,
|
| 878 |
+
"eval_samples_per_second": 86.493,
|
| 879 |
+
"eval_steps_per_second": 10.898,
|
| 880 |
+
"step": 10051
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 19.092627599243855,
|
| 884 |
+
"grad_norm": 0.6632242798805237,
|
| 885 |
+
"learning_rate": 0.0005,
|
| 886 |
+
"loss": 0.393,
|
| 887 |
+
"step": 10100
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 19.281663516068054,
|
| 891 |
+
"grad_norm": 0.8786609768867493,
|
| 892 |
+
"learning_rate": 0.0005,
|
| 893 |
+
"loss": 0.3528,
|
| 894 |
+
"step": 10200
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 19.47069943289225,
|
| 898 |
+
"grad_norm": 1.0305085182189941,
|
| 899 |
+
"learning_rate": 0.0005,
|
| 900 |
+
"loss": 0.3752,
|
| 901 |
+
"step": 10300
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 19.659735349716446,
|
| 905 |
+
"grad_norm": 0.9528347849845886,
|
| 906 |
+
"learning_rate": 0.0005,
|
| 907 |
+
"loss": 0.4023,
|
| 908 |
+
"step": 10400
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 19.84877126654064,
|
| 912 |
+
"grad_norm": 0.9051763415336609,
|
| 913 |
+
"learning_rate": 0.0005,
|
| 914 |
+
"loss": 0.4218,
|
| 915 |
+
"step": 10500
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 20.0,
|
| 919 |
+
"eval_accuracy": 0.560974358974359,
|
| 920 |
+
"eval_loss": 2.2417104244232178,
|
| 921 |
+
"eval_runtime": 5.7479,
|
| 922 |
+
"eval_samples_per_second": 86.989,
|
| 923 |
+
"eval_steps_per_second": 10.961,
|
| 924 |
+
"step": 10580
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 20.0,
|
| 928 |
+
"step": 10580,
|
| 929 |
+
"total_flos": 6.517631969856061e+17,
|
| 930 |
+
"train_loss": 0.6977483630405708,
|
| 931 |
+
"train_runtime": 24342.3236,
|
| 932 |
+
"train_samples_per_second": 13.906,
|
| 933 |
+
"train_steps_per_second": 0.435
|
| 934 |
+
}
|
| 935 |
+
],
|
| 936 |
+
"logging_steps": 100,
|
| 937 |
+
"max_steps": 10580,
|
| 938 |
+
"num_input_tokens_seen": 0,
|
| 939 |
+
"num_train_epochs": 20,
|
| 940 |
+
"save_steps": 500,
|
| 941 |
+
"total_flos": 6.517631969856061e+17,
|
| 942 |
+
"train_batch_size": 1,
|
| 943 |
+
"trial_name": null,
|
| 944 |
+
"trial_params": null
|
| 945 |
+
}
|