End of training
Browse files- README.md +14 -2
- all_results.json +15 -0
- eval_results.json +10 -0
- tokenizer.json +1 -6
- train_results.json +8 -0
- trainer_state.json +838 -0
README.md
CHANGED
|
@@ -3,11 +3,23 @@ license: llama2
|
|
| 3 |
base_model: meta-llama/Llama-2-7b-hf
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
- name: lmind_nq_train6000_eval6489_v1_doc_qa_v3_meta-llama_Llama-2-7b-hf_3e-5_lora2
|
| 10 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -15,7 +27,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# lmind_nq_train6000_eval6489_v1_doc_qa_v3_meta-llama_Llama-2-7b-hf_3e-5_lora2
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 2.0825
|
| 21 |
- Accuracy: 0.5966
|
|
|
|
| 3 |
base_model: meta-llama/Llama-2-7b-hf
|
| 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_meta-llama_Llama-2-7b-hf_3e-5_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: 0.5965641025641025
|
| 23 |
---
|
| 24 |
|
| 25 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 27 |
|
| 28 |
# lmind_nq_train6000_eval6489_v1_doc_qa_v3_meta-llama_Llama-2-7b-hf_3e-5_lora2
|
| 29 |
|
| 30 |
+
This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the tyzhu/lmind_nq_train6000_eval6489_v1_doc_qa_v3 dataset.
|
| 31 |
It achieves the following results on the evaluation set:
|
| 32 |
- Loss: 2.0825
|
| 33 |
- Accuracy: 0.5966
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.5965641025641025,
|
| 4 |
+
"eval_loss": 2.0825321674346924,
|
| 5 |
+
"eval_runtime": 4.4652,
|
| 6 |
+
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 111.976,
|
| 8 |
+
"eval_steps_per_second": 14.109,
|
| 9 |
+
"perplexity": 8.024763253301586,
|
| 10 |
+
"train_loss": 1.0529077490245913,
|
| 11 |
+
"train_runtime": 7366.0235,
|
| 12 |
+
"train_samples": 16925,
|
| 13 |
+
"train_samples_per_second": 45.954,
|
| 14 |
+
"train_steps_per_second": 1.436
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.5965641025641025,
|
| 4 |
+
"eval_loss": 2.0825321674346924,
|
| 5 |
+
"eval_runtime": 4.4652,
|
| 6 |
+
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 111.976,
|
| 8 |
+
"eval_steps_per_second": 14.109,
|
| 9 |
+
"perplexity": 8.024763253301586
|
| 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 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"train_loss": 1.0529077490245913,
|
| 4 |
+
"train_runtime": 7366.0235,
|
| 5 |
+
"train_samples": 16925,
|
| 6 |
+
"train_samples_per_second": 45.954,
|
| 7 |
+
"train_steps_per_second": 1.436
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,838 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 0.19,
|
| 13 |
+
"learning_rate": 3e-05,
|
| 14 |
+
"loss": 1.7194,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.38,
|
| 19 |
+
"learning_rate": 3e-05,
|
| 20 |
+
"loss": 1.4369,
|
| 21 |
+
"step": 200
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.57,
|
| 25 |
+
"learning_rate": 3e-05,
|
| 26 |
+
"loss": 1.4257,
|
| 27 |
+
"step": 300
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.76,
|
| 31 |
+
"learning_rate": 3e-05,
|
| 32 |
+
"loss": 1.3947,
|
| 33 |
+
"step": 400
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.95,
|
| 37 |
+
"learning_rate": 3e-05,
|
| 38 |
+
"loss": 1.3948,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.0,
|
| 43 |
+
"eval_accuracy": 0.6132307692307692,
|
| 44 |
+
"eval_loss": 1.3087366819381714,
|
| 45 |
+
"eval_runtime": 4.3528,
|
| 46 |
+
"eval_samples_per_second": 114.87,
|
| 47 |
+
"eval_steps_per_second": 14.474,
|
| 48 |
+
"step": 529
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.13,
|
| 52 |
+
"learning_rate": 3e-05,
|
| 53 |
+
"loss": 1.3927,
|
| 54 |
+
"step": 600
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 1.32,
|
| 58 |
+
"learning_rate": 3e-05,
|
| 59 |
+
"loss": 1.3644,
|
| 60 |
+
"step": 700
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.51,
|
| 64 |
+
"learning_rate": 3e-05,
|
| 65 |
+
"loss": 1.3791,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.7,
|
| 70 |
+
"learning_rate": 3e-05,
|
| 71 |
+
"loss": 1.3627,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.89,
|
| 76 |
+
"learning_rate": 3e-05,
|
| 77 |
+
"loss": 1.3789,
|
| 78 |
+
"step": 1000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.0,
|
| 82 |
+
"eval_accuracy": 0.6146153846153846,
|
| 83 |
+
"eval_loss": 1.289727807044983,
|
| 84 |
+
"eval_runtime": 4.1348,
|
| 85 |
+
"eval_samples_per_second": 120.924,
|
| 86 |
+
"eval_steps_per_second": 15.236,
|
| 87 |
+
"step": 1058
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 2.08,
|
| 91 |
+
"learning_rate": 3e-05,
|
| 92 |
+
"loss": 1.3605,
|
| 93 |
+
"step": 1100
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 2.27,
|
| 97 |
+
"learning_rate": 3e-05,
|
| 98 |
+
"loss": 1.3306,
|
| 99 |
+
"step": 1200
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 2.46,
|
| 103 |
+
"learning_rate": 3e-05,
|
| 104 |
+
"loss": 1.3318,
|
| 105 |
+
"step": 1300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 2.65,
|
| 109 |
+
"learning_rate": 3e-05,
|
| 110 |
+
"loss": 1.3491,
|
| 111 |
+
"step": 1400
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 2.84,
|
| 115 |
+
"learning_rate": 3e-05,
|
| 116 |
+
"loss": 1.3259,
|
| 117 |
+
"step": 1500
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 3.0,
|
| 121 |
+
"eval_accuracy": 0.6178974358974358,
|
| 122 |
+
"eval_loss": 1.2849316596984863,
|
| 123 |
+
"eval_runtime": 4.2318,
|
| 124 |
+
"eval_samples_per_second": 118.153,
|
| 125 |
+
"eval_steps_per_second": 14.887,
|
| 126 |
+
"step": 1587
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 3.02,
|
| 130 |
+
"learning_rate": 3e-05,
|
| 131 |
+
"loss": 1.3255,
|
| 132 |
+
"step": 1600
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 3.21,
|
| 136 |
+
"learning_rate": 3e-05,
|
| 137 |
+
"loss": 1.28,
|
| 138 |
+
"step": 1700
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.4,
|
| 142 |
+
"learning_rate": 3e-05,
|
| 143 |
+
"loss": 1.2995,
|
| 144 |
+
"step": 1800
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 3.59,
|
| 148 |
+
"learning_rate": 3e-05,
|
| 149 |
+
"loss": 1.2991,
|
| 150 |
+
"step": 1900
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 3.78,
|
| 154 |
+
"learning_rate": 3e-05,
|
| 155 |
+
"loss": 1.3024,
|
| 156 |
+
"step": 2000
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.97,
|
| 160 |
+
"learning_rate": 3e-05,
|
| 161 |
+
"loss": 1.2853,
|
| 162 |
+
"step": 2100
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 4.0,
|
| 166 |
+
"eval_accuracy": 0.615948717948718,
|
| 167 |
+
"eval_loss": 1.316924810409546,
|
| 168 |
+
"eval_runtime": 4.1288,
|
| 169 |
+
"eval_samples_per_second": 121.101,
|
| 170 |
+
"eval_steps_per_second": 15.259,
|
| 171 |
+
"step": 2116
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 4.16,
|
| 175 |
+
"learning_rate": 3e-05,
|
| 176 |
+
"loss": 1.2575,
|
| 177 |
+
"step": 2200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 4.35,
|
| 181 |
+
"learning_rate": 3e-05,
|
| 182 |
+
"loss": 1.2261,
|
| 183 |
+
"step": 2300
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.54,
|
| 187 |
+
"learning_rate": 3e-05,
|
| 188 |
+
"loss": 1.2536,
|
| 189 |
+
"step": 2400
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 4.73,
|
| 193 |
+
"learning_rate": 3e-05,
|
| 194 |
+
"loss": 1.2376,
|
| 195 |
+
"step": 2500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 4.91,
|
| 199 |
+
"learning_rate": 3e-05,
|
| 200 |
+
"loss": 1.2556,
|
| 201 |
+
"step": 2600
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 5.0,
|
| 205 |
+
"eval_accuracy": 0.6131794871794872,
|
| 206 |
+
"eval_loss": 1.3531945943832397,
|
| 207 |
+
"eval_runtime": 4.4339,
|
| 208 |
+
"eval_samples_per_second": 112.769,
|
| 209 |
+
"eval_steps_per_second": 14.209,
|
| 210 |
+
"step": 2645
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 5.1,
|
| 214 |
+
"learning_rate": 3e-05,
|
| 215 |
+
"loss": 1.2195,
|
| 216 |
+
"step": 2700
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.29,
|
| 220 |
+
"learning_rate": 3e-05,
|
| 221 |
+
"loss": 1.2003,
|
| 222 |
+
"step": 2800
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 5.48,
|
| 226 |
+
"learning_rate": 3e-05,
|
| 227 |
+
"loss": 1.1864,
|
| 228 |
+
"step": 2900
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 5.67,
|
| 232 |
+
"learning_rate": 3e-05,
|
| 233 |
+
"loss": 1.2231,
|
| 234 |
+
"step": 3000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 5.86,
|
| 238 |
+
"learning_rate": 3e-05,
|
| 239 |
+
"loss": 1.1972,
|
| 240 |
+
"step": 3100
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 6.0,
|
| 244 |
+
"eval_accuracy": 0.6125641025641025,
|
| 245 |
+
"eval_loss": 1.4135066270828247,
|
| 246 |
+
"eval_runtime": 4.9566,
|
| 247 |
+
"eval_samples_per_second": 100.875,
|
| 248 |
+
"eval_steps_per_second": 12.71,
|
| 249 |
+
"step": 3174
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 6.05,
|
| 253 |
+
"learning_rate": 3e-05,
|
| 254 |
+
"loss": 1.1889,
|
| 255 |
+
"step": 3200
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 6.24,
|
| 259 |
+
"learning_rate": 3e-05,
|
| 260 |
+
"loss": 1.1471,
|
| 261 |
+
"step": 3300
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 6.43,
|
| 265 |
+
"learning_rate": 3e-05,
|
| 266 |
+
"loss": 1.1509,
|
| 267 |
+
"step": 3400
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.62,
|
| 271 |
+
"learning_rate": 3e-05,
|
| 272 |
+
"loss": 1.1654,
|
| 273 |
+
"step": 3500
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 6.81,
|
| 277 |
+
"learning_rate": 3e-05,
|
| 278 |
+
"loss": 1.1716,
|
| 279 |
+
"step": 3600
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 6.99,
|
| 283 |
+
"learning_rate": 3e-05,
|
| 284 |
+
"loss": 1.1839,
|
| 285 |
+
"step": 3700
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 7.0,
|
| 289 |
+
"eval_accuracy": 0.6081025641025641,
|
| 290 |
+
"eval_loss": 1.500714898109436,
|
| 291 |
+
"eval_runtime": 4.8786,
|
| 292 |
+
"eval_samples_per_second": 102.488,
|
| 293 |
+
"eval_steps_per_second": 12.914,
|
| 294 |
+
"step": 3703
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 7.18,
|
| 298 |
+
"learning_rate": 3e-05,
|
| 299 |
+
"loss": 1.1059,
|
| 300 |
+
"step": 3800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 7.37,
|
| 304 |
+
"learning_rate": 3e-05,
|
| 305 |
+
"loss": 1.1146,
|
| 306 |
+
"step": 3900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 7.56,
|
| 310 |
+
"learning_rate": 3e-05,
|
| 311 |
+
"loss": 1.1121,
|
| 312 |
+
"step": 4000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 7.75,
|
| 316 |
+
"learning_rate": 3e-05,
|
| 317 |
+
"loss": 1.1286,
|
| 318 |
+
"step": 4100
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 7.94,
|
| 322 |
+
"learning_rate": 3e-05,
|
| 323 |
+
"loss": 1.1334,
|
| 324 |
+
"step": 4200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 8.0,
|
| 328 |
+
"eval_accuracy": 0.6074358974358974,
|
| 329 |
+
"eval_loss": 1.524243950843811,
|
| 330 |
+
"eval_runtime": 4.9049,
|
| 331 |
+
"eval_samples_per_second": 101.94,
|
| 332 |
+
"eval_steps_per_second": 12.844,
|
| 333 |
+
"step": 4232
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 8.13,
|
| 337 |
+
"learning_rate": 3e-05,
|
| 338 |
+
"loss": 1.1072,
|
| 339 |
+
"step": 4300
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 8.32,
|
| 343 |
+
"learning_rate": 3e-05,
|
| 344 |
+
"loss": 1.0626,
|
| 345 |
+
"step": 4400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 8.51,
|
| 349 |
+
"learning_rate": 3e-05,
|
| 350 |
+
"loss": 1.0842,
|
| 351 |
+
"step": 4500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 8.7,
|
| 355 |
+
"learning_rate": 3e-05,
|
| 356 |
+
"loss": 1.0812,
|
| 357 |
+
"step": 4600
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 8.88,
|
| 361 |
+
"learning_rate": 3e-05,
|
| 362 |
+
"loss": 1.0966,
|
| 363 |
+
"step": 4700
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 9.0,
|
| 367 |
+
"eval_accuracy": 0.5802564102564103,
|
| 368 |
+
"eval_loss": 1.6106736660003662,
|
| 369 |
+
"eval_runtime": 4.8671,
|
| 370 |
+
"eval_samples_per_second": 102.731,
|
| 371 |
+
"eval_steps_per_second": 12.944,
|
| 372 |
+
"step": 4761
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 9.07,
|
| 376 |
+
"learning_rate": 3e-05,
|
| 377 |
+
"loss": 1.0755,
|
| 378 |
+
"step": 4800
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 9.26,
|
| 382 |
+
"learning_rate": 3e-05,
|
| 383 |
+
"loss": 1.0189,
|
| 384 |
+
"step": 4900
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 9.45,
|
| 388 |
+
"learning_rate": 3e-05,
|
| 389 |
+
"loss": 1.0565,
|
| 390 |
+
"step": 5000
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 9.64,
|
| 394 |
+
"learning_rate": 3e-05,
|
| 395 |
+
"loss": 1.0635,
|
| 396 |
+
"step": 5100
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 9.83,
|
| 400 |
+
"learning_rate": 3e-05,
|
| 401 |
+
"loss": 1.0485,
|
| 402 |
+
"step": 5200
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 10.0,
|
| 406 |
+
"eval_accuracy": 0.6049230769230769,
|
| 407 |
+
"eval_loss": 1.6749216318130493,
|
| 408 |
+
"eval_runtime": 4.8537,
|
| 409 |
+
"eval_samples_per_second": 103.013,
|
| 410 |
+
"eval_steps_per_second": 12.98,
|
| 411 |
+
"step": 5290
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 10.02,
|
| 415 |
+
"learning_rate": 3e-05,
|
| 416 |
+
"loss": 1.0585,
|
| 417 |
+
"step": 5300
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 10.21,
|
| 421 |
+
"learning_rate": 3e-05,
|
| 422 |
+
"loss": 1.0098,
|
| 423 |
+
"step": 5400
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 10.4,
|
| 427 |
+
"learning_rate": 3e-05,
|
| 428 |
+
"loss": 0.9935,
|
| 429 |
+
"step": 5500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 10.59,
|
| 433 |
+
"learning_rate": 3e-05,
|
| 434 |
+
"loss": 1.0178,
|
| 435 |
+
"step": 5600
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 10.78,
|
| 439 |
+
"learning_rate": 3e-05,
|
| 440 |
+
"loss": 1.0217,
|
| 441 |
+
"step": 5700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 10.96,
|
| 445 |
+
"learning_rate": 3e-05,
|
| 446 |
+
"loss": 1.021,
|
| 447 |
+
"step": 5800
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 11.0,
|
| 451 |
+
"eval_accuracy": 0.6014871794871794,
|
| 452 |
+
"eval_loss": 1.7323671579360962,
|
| 453 |
+
"eval_runtime": 4.2401,
|
| 454 |
+
"eval_samples_per_second": 117.922,
|
| 455 |
+
"eval_steps_per_second": 14.858,
|
| 456 |
+
"step": 5819
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 11.15,
|
| 460 |
+
"learning_rate": 3e-05,
|
| 461 |
+
"loss": 0.9788,
|
| 462 |
+
"step": 5900
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 11.34,
|
| 466 |
+
"learning_rate": 3e-05,
|
| 467 |
+
"loss": 0.9799,
|
| 468 |
+
"step": 6000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 11.53,
|
| 472 |
+
"learning_rate": 3e-05,
|
| 473 |
+
"loss": 0.9749,
|
| 474 |
+
"step": 6100
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 11.72,
|
| 478 |
+
"learning_rate": 3e-05,
|
| 479 |
+
"loss": 0.981,
|
| 480 |
+
"step": 6200
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 11.91,
|
| 484 |
+
"learning_rate": 3e-05,
|
| 485 |
+
"loss": 0.9918,
|
| 486 |
+
"step": 6300
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 12.0,
|
| 490 |
+
"eval_accuracy": 0.6007179487179487,
|
| 491 |
+
"eval_loss": 1.7632156610488892,
|
| 492 |
+
"eval_runtime": 5.2634,
|
| 493 |
+
"eval_samples_per_second": 94.995,
|
| 494 |
+
"eval_steps_per_second": 11.969,
|
| 495 |
+
"step": 6348
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 12.1,
|
| 499 |
+
"learning_rate": 3e-05,
|
| 500 |
+
"loss": 0.967,
|
| 501 |
+
"step": 6400
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 12.29,
|
| 505 |
+
"learning_rate": 3e-05,
|
| 506 |
+
"loss": 0.9341,
|
| 507 |
+
"step": 6500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 12.48,
|
| 511 |
+
"learning_rate": 3e-05,
|
| 512 |
+
"loss": 0.9656,
|
| 513 |
+
"step": 6600
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 12.67,
|
| 517 |
+
"learning_rate": 3e-05,
|
| 518 |
+
"loss": 0.9519,
|
| 519 |
+
"step": 6700
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 12.85,
|
| 523 |
+
"learning_rate": 3e-05,
|
| 524 |
+
"loss": 0.947,
|
| 525 |
+
"step": 6800
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 13.0,
|
| 529 |
+
"eval_accuracy": 0.6010769230769231,
|
| 530 |
+
"eval_loss": 1.8302537202835083,
|
| 531 |
+
"eval_runtime": 4.8387,
|
| 532 |
+
"eval_samples_per_second": 103.333,
|
| 533 |
+
"eval_steps_per_second": 13.02,
|
| 534 |
+
"step": 6877
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 13.04,
|
| 538 |
+
"learning_rate": 3e-05,
|
| 539 |
+
"loss": 0.9461,
|
| 540 |
+
"step": 6900
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 13.23,
|
| 544 |
+
"learning_rate": 3e-05,
|
| 545 |
+
"loss": 0.9153,
|
| 546 |
+
"step": 7000
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 13.42,
|
| 550 |
+
"learning_rate": 3e-05,
|
| 551 |
+
"loss": 0.9166,
|
| 552 |
+
"step": 7100
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 13.61,
|
| 556 |
+
"learning_rate": 3e-05,
|
| 557 |
+
"loss": 0.9266,
|
| 558 |
+
"step": 7200
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 13.8,
|
| 562 |
+
"learning_rate": 3e-05,
|
| 563 |
+
"loss": 0.9089,
|
| 564 |
+
"step": 7300
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 13.99,
|
| 568 |
+
"learning_rate": 3e-05,
|
| 569 |
+
"loss": 0.9376,
|
| 570 |
+
"step": 7400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 14.0,
|
| 574 |
+
"eval_accuracy": 0.5990769230769231,
|
| 575 |
+
"eval_loss": 1.8872731924057007,
|
| 576 |
+
"eval_runtime": 4.8378,
|
| 577 |
+
"eval_samples_per_second": 103.354,
|
| 578 |
+
"eval_steps_per_second": 13.023,
|
| 579 |
+
"step": 7406
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 14.18,
|
| 583 |
+
"learning_rate": 3e-05,
|
| 584 |
+
"loss": 0.8792,
|
| 585 |
+
"step": 7500
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 14.37,
|
| 589 |
+
"learning_rate": 3e-05,
|
| 590 |
+
"loss": 0.8787,
|
| 591 |
+
"step": 7600
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 14.56,
|
| 595 |
+
"learning_rate": 3e-05,
|
| 596 |
+
"loss": 0.874,
|
| 597 |
+
"step": 7700
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 14.74,
|
| 601 |
+
"learning_rate": 3e-05,
|
| 602 |
+
"loss": 0.9123,
|
| 603 |
+
"step": 7800
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 14.93,
|
| 607 |
+
"learning_rate": 3e-05,
|
| 608 |
+
"loss": 0.898,
|
| 609 |
+
"step": 7900
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 15.0,
|
| 613 |
+
"eval_accuracy": 0.5975897435897436,
|
| 614 |
+
"eval_loss": 1.96884024143219,
|
| 615 |
+
"eval_runtime": 4.4382,
|
| 616 |
+
"eval_samples_per_second": 112.657,
|
| 617 |
+
"eval_steps_per_second": 14.195,
|
| 618 |
+
"step": 7935
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 15.12,
|
| 622 |
+
"learning_rate": 3e-05,
|
| 623 |
+
"loss": 0.8631,
|
| 624 |
+
"step": 8000
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 15.31,
|
| 628 |
+
"learning_rate": 3e-05,
|
| 629 |
+
"loss": 0.8353,
|
| 630 |
+
"step": 8100
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 15.5,
|
| 634 |
+
"learning_rate": 3e-05,
|
| 635 |
+
"loss": 0.8673,
|
| 636 |
+
"step": 8200
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 15.69,
|
| 640 |
+
"learning_rate": 3e-05,
|
| 641 |
+
"loss": 0.8515,
|
| 642 |
+
"step": 8300
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 15.88,
|
| 646 |
+
"learning_rate": 3e-05,
|
| 647 |
+
"loss": 0.8559,
|
| 648 |
+
"step": 8400
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 16.0,
|
| 652 |
+
"eval_accuracy": 0.5987692307692307,
|
| 653 |
+
"eval_loss": 1.9724328517913818,
|
| 654 |
+
"eval_runtime": 4.4368,
|
| 655 |
+
"eval_samples_per_second": 112.695,
|
| 656 |
+
"eval_steps_per_second": 14.2,
|
| 657 |
+
"step": 8464
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 16.07,
|
| 661 |
+
"learning_rate": 3e-05,
|
| 662 |
+
"loss": 0.8508,
|
| 663 |
+
"step": 8500
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 16.26,
|
| 667 |
+
"learning_rate": 3e-05,
|
| 668 |
+
"loss": 0.8122,
|
| 669 |
+
"step": 8600
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 16.45,
|
| 673 |
+
"learning_rate": 3e-05,
|
| 674 |
+
"loss": 0.8223,
|
| 675 |
+
"step": 8700
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 16.64,
|
| 679 |
+
"learning_rate": 3e-05,
|
| 680 |
+
"loss": 0.8275,
|
| 681 |
+
"step": 8800
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 16.82,
|
| 685 |
+
"learning_rate": 3e-05,
|
| 686 |
+
"loss": 0.8348,
|
| 687 |
+
"step": 8900
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 17.0,
|
| 691 |
+
"eval_accuracy": 0.5713846153846154,
|
| 692 |
+
"eval_loss": 1.9815243482589722,
|
| 693 |
+
"eval_runtime": 4.8518,
|
| 694 |
+
"eval_samples_per_second": 103.054,
|
| 695 |
+
"eval_steps_per_second": 12.985,
|
| 696 |
+
"step": 8993
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 17.01,
|
| 700 |
+
"learning_rate": 3e-05,
|
| 701 |
+
"loss": 0.837,
|
| 702 |
+
"step": 9000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 17.2,
|
| 706 |
+
"learning_rate": 3e-05,
|
| 707 |
+
"loss": 0.7764,
|
| 708 |
+
"step": 9100
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 17.39,
|
| 712 |
+
"learning_rate": 3e-05,
|
| 713 |
+
"loss": 0.7817,
|
| 714 |
+
"step": 9200
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 17.58,
|
| 718 |
+
"learning_rate": 3e-05,
|
| 719 |
+
"loss": 0.8068,
|
| 720 |
+
"step": 9300
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 17.77,
|
| 724 |
+
"learning_rate": 3e-05,
|
| 725 |
+
"loss": 0.7968,
|
| 726 |
+
"step": 9400
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 17.96,
|
| 730 |
+
"learning_rate": 3e-05,
|
| 731 |
+
"loss": 0.8106,
|
| 732 |
+
"step": 9500
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 18.0,
|
| 736 |
+
"eval_accuracy": 0.598,
|
| 737 |
+
"eval_loss": 2.038572072982788,
|
| 738 |
+
"eval_runtime": 4.8974,
|
| 739 |
+
"eval_samples_per_second": 102.096,
|
| 740 |
+
"eval_steps_per_second": 12.864,
|
| 741 |
+
"step": 9522
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 18.15,
|
| 745 |
+
"learning_rate": 3e-05,
|
| 746 |
+
"loss": 0.7697,
|
| 747 |
+
"step": 9600
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 18.34,
|
| 751 |
+
"learning_rate": 3e-05,
|
| 752 |
+
"loss": 0.7429,
|
| 753 |
+
"step": 9700
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 18.53,
|
| 757 |
+
"learning_rate": 3e-05,
|
| 758 |
+
"loss": 0.7468,
|
| 759 |
+
"step": 9800
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 18.71,
|
| 763 |
+
"learning_rate": 3e-05,
|
| 764 |
+
"loss": 0.7667,
|
| 765 |
+
"step": 9900
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 18.9,
|
| 769 |
+
"learning_rate": 3e-05,
|
| 770 |
+
"loss": 0.7848,
|
| 771 |
+
"step": 10000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 19.0,
|
| 775 |
+
"eval_accuracy": 0.5963589743589743,
|
| 776 |
+
"eval_loss": 2.0627450942993164,
|
| 777 |
+
"eval_runtime": 5.1596,
|
| 778 |
+
"eval_samples_per_second": 96.906,
|
| 779 |
+
"eval_steps_per_second": 12.21,
|
| 780 |
+
"step": 10051
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 19.09,
|
| 784 |
+
"learning_rate": 3e-05,
|
| 785 |
+
"loss": 0.7537,
|
| 786 |
+
"step": 10100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 19.28,
|
| 790 |
+
"learning_rate": 3e-05,
|
| 791 |
+
"loss": 0.718,
|
| 792 |
+
"step": 10200
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 19.47,
|
| 796 |
+
"learning_rate": 3e-05,
|
| 797 |
+
"loss": 0.7241,
|
| 798 |
+
"step": 10300
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 19.66,
|
| 802 |
+
"learning_rate": 3e-05,
|
| 803 |
+
"loss": 0.7421,
|
| 804 |
+
"step": 10400
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 19.85,
|
| 808 |
+
"learning_rate": 3e-05,
|
| 809 |
+
"loss": 0.745,
|
| 810 |
+
"step": 10500
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 20.0,
|
| 814 |
+
"eval_accuracy": 0.5965641025641025,
|
| 815 |
+
"eval_loss": 2.0825321674346924,
|
| 816 |
+
"eval_runtime": 4.1414,
|
| 817 |
+
"eval_samples_per_second": 120.733,
|
| 818 |
+
"eval_steps_per_second": 15.212,
|
| 819 |
+
"step": 10580
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 20.0,
|
| 823 |
+
"step": 10580,
|
| 824 |
+
"total_flos": 1.7870734670660895e+18,
|
| 825 |
+
"train_loss": 1.0529077490245913,
|
| 826 |
+
"train_runtime": 7366.0235,
|
| 827 |
+
"train_samples_per_second": 45.954,
|
| 828 |
+
"train_steps_per_second": 1.436
|
| 829 |
+
}
|
| 830 |
+
],
|
| 831 |
+
"logging_steps": 100,
|
| 832 |
+
"max_steps": 10580,
|
| 833 |
+
"num_train_epochs": 20,
|
| 834 |
+
"save_steps": 500,
|
| 835 |
+
"total_flos": 1.7870734670660895e+18,
|
| 836 |
+
"trial_name": null,
|
| 837 |
+
"trial_params": null
|
| 838 |
+
}
|