End of training
Browse files- README.md +14 -2
- all_results.json +21 -0
- eval_results.json +16 -0
- train_results.json +8 -0
- trainer_state.json +788 -0
README.md
CHANGED
|
@@ -3,11 +3,23 @@ license: mit
|
|
| 3 |
base_model: gpt2-xl
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
- name: lmind_hotpot_train1000_eval200_v1_recite_qa_gpt2-xl
|
| 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_hotpot_train1000_eval200_v1_recite_qa_gpt2-xl
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [gpt2-xl](https://huggingface.co/gpt2-xl) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.4436
|
| 21 |
- Accuracy: 0.6989
|
|
|
|
| 3 |
base_model: gpt2-xl
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- tyzhu/lmind_hotpot_train1000_eval200_v1_recite_qa
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
| 10 |
model-index:
|
| 11 |
- name: lmind_hotpot_train1000_eval200_v1_recite_qa_gpt2-xl
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Causal Language Modeling
|
| 15 |
+
type: text-generation
|
| 16 |
+
dataset:
|
| 17 |
+
name: tyzhu/lmind_hotpot_train1000_eval200_v1_recite_qa
|
| 18 |
+
type: tyzhu/lmind_hotpot_train1000_eval200_v1_recite_qa
|
| 19 |
+
metrics:
|
| 20 |
+
- name: Accuracy
|
| 21 |
+
type: accuracy
|
| 22 |
+
value: 0.6988971684053651
|
| 23 |
---
|
| 24 |
|
| 25 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 27 |
|
| 28 |
# lmind_hotpot_train1000_eval200_v1_recite_qa_gpt2-xl
|
| 29 |
|
| 30 |
+
This model is a fine-tuned version of [gpt2-xl](https://huggingface.co/gpt2-xl) on the tyzhu/lmind_hotpot_train1000_eval200_v1_recite_qa dataset.
|
| 31 |
It achieves the following results on the evaluation set:
|
| 32 |
- Loss: 0.4436
|
| 33 |
- Accuracy: 0.6989
|
all_results.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.6988971684053651,
|
| 4 |
+
"eval_exact_match": 20.5,
|
| 5 |
+
"eval_f1": 25.95833333333333,
|
| 6 |
+
"eval_loss": 0.4435840845108032,
|
| 7 |
+
"eval_qa_bleu": 12.082719988904218,
|
| 8 |
+
"eval_qa_exact_match": 0.18,
|
| 9 |
+
"eval_recite_bleu": 44.74289478974747,
|
| 10 |
+
"eval_recite_exact_match": 0.075,
|
| 11 |
+
"eval_runtime": 6.1906,
|
| 12 |
+
"eval_samples": 200,
|
| 13 |
+
"eval_samples_per_second": 32.307,
|
| 14 |
+
"eval_steps_per_second": 2.1,
|
| 15 |
+
"perplexity": 1.5582822371040086,
|
| 16 |
+
"train_loss": 0.29445283661521443,
|
| 17 |
+
"train_runtime": 3197.9615,
|
| 18 |
+
"train_samples": 3373,
|
| 19 |
+
"train_samples_per_second": 10.547,
|
| 20 |
+
"train_steps_per_second": 0.66
|
| 21 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.6988971684053651,
|
| 4 |
+
"eval_exact_match": 20.5,
|
| 5 |
+
"eval_f1": 25.95833333333333,
|
| 6 |
+
"eval_loss": 0.4435840845108032,
|
| 7 |
+
"eval_qa_bleu": 12.082719988904218,
|
| 8 |
+
"eval_qa_exact_match": 0.18,
|
| 9 |
+
"eval_recite_bleu": 44.74289478974747,
|
| 10 |
+
"eval_recite_exact_match": 0.075,
|
| 11 |
+
"eval_runtime": 6.1906,
|
| 12 |
+
"eval_samples": 200,
|
| 13 |
+
"eval_samples_per_second": 32.307,
|
| 14 |
+
"eval_steps_per_second": 2.1,
|
| 15 |
+
"perplexity": 1.5582822371040086
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"train_loss": 0.29445283661521443,
|
| 4 |
+
"train_runtime": 3197.9615,
|
| 5 |
+
"train_samples": 3373,
|
| 6 |
+
"train_samples_per_second": 10.547,
|
| 7 |
+
"train_steps_per_second": 0.66
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,788 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2110,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"learning_rate": 6.226415094339623e-06,
|
| 14 |
+
"loss": 1.1807,
|
| 15 |
+
"step": 22
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.21,
|
| 19 |
+
"learning_rate": 1.2452830188679246e-05,
|
| 20 |
+
"loss": 1.1325,
|
| 21 |
+
"step": 44
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.31,
|
| 25 |
+
"learning_rate": 1.8679245283018867e-05,
|
| 26 |
+
"loss": 0.9876,
|
| 27 |
+
"step": 66
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.42,
|
| 31 |
+
"learning_rate": 2.4905660377358492e-05,
|
| 32 |
+
"loss": 0.873,
|
| 33 |
+
"step": 88
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.52,
|
| 37 |
+
"learning_rate": 2.994011976047904e-05,
|
| 38 |
+
"loss": 0.7931,
|
| 39 |
+
"step": 110
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.63,
|
| 43 |
+
"learning_rate": 2.9610778443113774e-05,
|
| 44 |
+
"loss": 0.7764,
|
| 45 |
+
"step": 132
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.73,
|
| 49 |
+
"learning_rate": 2.9281437125748504e-05,
|
| 50 |
+
"loss": 0.7701,
|
| 51 |
+
"step": 154
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.83,
|
| 55 |
+
"learning_rate": 2.8952095808383233e-05,
|
| 56 |
+
"loss": 0.739,
|
| 57 |
+
"step": 176
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.94,
|
| 61 |
+
"learning_rate": 2.8622754491017966e-05,
|
| 62 |
+
"loss": 0.7644,
|
| 63 |
+
"step": 198
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0,
|
| 67 |
+
"eval_accuracy": 0.6429210134128167,
|
| 68 |
+
"eval_loss": 1.0036486387252808,
|
| 69 |
+
"eval_runtime": 6.1897,
|
| 70 |
+
"eval_samples_per_second": 32.312,
|
| 71 |
+
"eval_steps_per_second": 2.1,
|
| 72 |
+
"step": 211
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_exact_match": 14.5,
|
| 77 |
+
"eval_f1": 19.727380952380955,
|
| 78 |
+
"eval_qa_bleu": 9.309381343075087,
|
| 79 |
+
"eval_qa_exact_match": 0.14,
|
| 80 |
+
"eval_recite_bleu": 15.818078125717689,
|
| 81 |
+
"eval_recite_exact_match": 0.0,
|
| 82 |
+
"step": 211
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 1.04,
|
| 86 |
+
"learning_rate": 2.8293413173652696e-05,
|
| 87 |
+
"loss": 0.6488,
|
| 88 |
+
"step": 220
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.15,
|
| 92 |
+
"learning_rate": 2.7964071856287425e-05,
|
| 93 |
+
"loss": 0.5286,
|
| 94 |
+
"step": 242
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.25,
|
| 98 |
+
"learning_rate": 2.763473053892216e-05,
|
| 99 |
+
"loss": 0.5157,
|
| 100 |
+
"step": 264
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.36,
|
| 104 |
+
"learning_rate": 2.7305389221556884e-05,
|
| 105 |
+
"loss": 0.4838,
|
| 106 |
+
"step": 286
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 1.46,
|
| 110 |
+
"learning_rate": 2.6976047904191617e-05,
|
| 111 |
+
"loss": 0.5013,
|
| 112 |
+
"step": 308
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.56,
|
| 116 |
+
"learning_rate": 2.664670658682635e-05,
|
| 117 |
+
"loss": 0.4649,
|
| 118 |
+
"step": 330
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 1.67,
|
| 122 |
+
"learning_rate": 2.6317365269461076e-05,
|
| 123 |
+
"loss": 0.4604,
|
| 124 |
+
"step": 352
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 1.77,
|
| 128 |
+
"learning_rate": 2.598802395209581e-05,
|
| 129 |
+
"loss": 0.5064,
|
| 130 |
+
"step": 374
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.88,
|
| 134 |
+
"learning_rate": 2.565868263473054e-05,
|
| 135 |
+
"loss": 0.4732,
|
| 136 |
+
"step": 396
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.98,
|
| 140 |
+
"learning_rate": 2.5329341317365268e-05,
|
| 141 |
+
"loss": 0.526,
|
| 142 |
+
"step": 418
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.0,
|
| 146 |
+
"eval_accuracy": 0.6657600596125186,
|
| 147 |
+
"eval_loss": 0.7350580096244812,
|
| 148 |
+
"eval_runtime": 6.5425,
|
| 149 |
+
"eval_samples_per_second": 30.569,
|
| 150 |
+
"eval_steps_per_second": 1.987,
|
| 151 |
+
"step": 422
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_exact_match": 12.5,
|
| 156 |
+
"eval_f1": 19.69947691197691,
|
| 157 |
+
"eval_qa_bleu": 9.452490656702446,
|
| 158 |
+
"eval_qa_exact_match": 0.12,
|
| 159 |
+
"eval_recite_bleu": 17.493375608153375,
|
| 160 |
+
"eval_recite_exact_match": 0.0,
|
| 161 |
+
"step": 422
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 2.09,
|
| 165 |
+
"learning_rate": 2.5e-05,
|
| 166 |
+
"loss": 0.4741,
|
| 167 |
+
"step": 440
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.19,
|
| 171 |
+
"learning_rate": 2.467065868263473e-05,
|
| 172 |
+
"loss": 0.4338,
|
| 173 |
+
"step": 462
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 2.29,
|
| 177 |
+
"learning_rate": 2.4341317365269464e-05,
|
| 178 |
+
"loss": 0.4451,
|
| 179 |
+
"step": 484
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.4,
|
| 183 |
+
"learning_rate": 2.4011976047904193e-05,
|
| 184 |
+
"loss": 0.4312,
|
| 185 |
+
"step": 506
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 2.5,
|
| 189 |
+
"learning_rate": 2.3682634730538923e-05,
|
| 190 |
+
"loss": 0.4363,
|
| 191 |
+
"step": 528
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 2.61,
|
| 195 |
+
"learning_rate": 2.3353293413173656e-05,
|
| 196 |
+
"loss": 0.4302,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 2.71,
|
| 201 |
+
"learning_rate": 2.302395209580838e-05,
|
| 202 |
+
"loss": 0.4528,
|
| 203 |
+
"step": 572
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 2.82,
|
| 207 |
+
"learning_rate": 2.2694610778443115e-05,
|
| 208 |
+
"loss": 0.4086,
|
| 209 |
+
"step": 594
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.92,
|
| 213 |
+
"learning_rate": 2.2365269461077847e-05,
|
| 214 |
+
"loss": 0.4163,
|
| 215 |
+
"step": 616
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 3.0,
|
| 219 |
+
"eval_accuracy": 0.6815499254843517,
|
| 220 |
+
"eval_loss": 0.5743635296821594,
|
| 221 |
+
"eval_runtime": 6.2699,
|
| 222 |
+
"eval_samples_per_second": 31.898,
|
| 223 |
+
"eval_steps_per_second": 2.073,
|
| 224 |
+
"step": 633
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 3.0,
|
| 228 |
+
"eval_exact_match": 13.0,
|
| 229 |
+
"eval_f1": 20.81829004329004,
|
| 230 |
+
"eval_qa_bleu": 7.245823931129993,
|
| 231 |
+
"eval_qa_exact_match": 0.12,
|
| 232 |
+
"eval_recite_bleu": 22.248138618982026,
|
| 233 |
+
"eval_recite_exact_match": 0.0,
|
| 234 |
+
"step": 633
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 3.02,
|
| 238 |
+
"learning_rate": 2.2035928143712574e-05,
|
| 239 |
+
"loss": 0.3866,
|
| 240 |
+
"step": 638
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.13,
|
| 244 |
+
"learning_rate": 2.1706586826347306e-05,
|
| 245 |
+
"loss": 0.2789,
|
| 246 |
+
"step": 660
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.23,
|
| 250 |
+
"learning_rate": 2.1377245508982036e-05,
|
| 251 |
+
"loss": 0.2932,
|
| 252 |
+
"step": 682
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 3.34,
|
| 256 |
+
"learning_rate": 2.1047904191616766e-05,
|
| 257 |
+
"loss": 0.2883,
|
| 258 |
+
"step": 704
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 3.44,
|
| 262 |
+
"learning_rate": 2.07185628742515e-05,
|
| 263 |
+
"loss": 0.3048,
|
| 264 |
+
"step": 726
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 3.55,
|
| 268 |
+
"learning_rate": 2.0389221556886228e-05,
|
| 269 |
+
"loss": 0.28,
|
| 270 |
+
"step": 748
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 3.65,
|
| 274 |
+
"learning_rate": 2.0059880239520957e-05,
|
| 275 |
+
"loss": 0.2847,
|
| 276 |
+
"step": 770
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 3.75,
|
| 280 |
+
"learning_rate": 1.9730538922155687e-05,
|
| 281 |
+
"loss": 0.2949,
|
| 282 |
+
"step": 792
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 3.86,
|
| 286 |
+
"learning_rate": 1.940119760479042e-05,
|
| 287 |
+
"loss": 0.3002,
|
| 288 |
+
"step": 814
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.96,
|
| 292 |
+
"learning_rate": 1.9071856287425153e-05,
|
| 293 |
+
"loss": 0.2864,
|
| 294 |
+
"step": 836
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 4.0,
|
| 298 |
+
"eval_accuracy": 0.689903129657228,
|
| 299 |
+
"eval_loss": 0.4952711760997772,
|
| 300 |
+
"eval_runtime": 6.337,
|
| 301 |
+
"eval_samples_per_second": 31.561,
|
| 302 |
+
"eval_steps_per_second": 2.051,
|
| 303 |
+
"step": 844
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 4.0,
|
| 307 |
+
"eval_exact_match": 17.0,
|
| 308 |
+
"eval_f1": 22.154761904761905,
|
| 309 |
+
"eval_qa_bleu": 10.09488722739415,
|
| 310 |
+
"eval_qa_exact_match": 0.155,
|
| 311 |
+
"eval_recite_bleu": 27.91257178143564,
|
| 312 |
+
"eval_recite_exact_match": 0.005,
|
| 313 |
+
"step": 844
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 4.07,
|
| 317 |
+
"learning_rate": 1.874251497005988e-05,
|
| 318 |
+
"loss": 0.2368,
|
| 319 |
+
"step": 858
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 4.17,
|
| 323 |
+
"learning_rate": 1.8413173652694612e-05,
|
| 324 |
+
"loss": 0.203,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.27,
|
| 329 |
+
"learning_rate": 1.8083832335329345e-05,
|
| 330 |
+
"loss": 0.2092,
|
| 331 |
+
"step": 902
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 4.38,
|
| 335 |
+
"learning_rate": 1.775449101796407e-05,
|
| 336 |
+
"loss": 0.2188,
|
| 337 |
+
"step": 924
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 4.48,
|
| 341 |
+
"learning_rate": 1.7425149700598804e-05,
|
| 342 |
+
"loss": 0.2092,
|
| 343 |
+
"step": 946
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 4.59,
|
| 347 |
+
"learning_rate": 1.7095808383233533e-05,
|
| 348 |
+
"loss": 0.2144,
|
| 349 |
+
"step": 968
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 4.69,
|
| 353 |
+
"learning_rate": 1.6766467065868263e-05,
|
| 354 |
+
"loss": 0.2061,
|
| 355 |
+
"step": 990
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 4.8,
|
| 359 |
+
"learning_rate": 1.6437125748502996e-05,
|
| 360 |
+
"loss": 0.2161,
|
| 361 |
+
"step": 1012
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 4.9,
|
| 365 |
+
"learning_rate": 1.6107784431137725e-05,
|
| 366 |
+
"loss": 0.2118,
|
| 367 |
+
"step": 1034
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 5.0,
|
| 371 |
+
"eval_accuracy": 0.6943889716840537,
|
| 372 |
+
"eval_loss": 0.45938587188720703,
|
| 373 |
+
"eval_runtime": 6.2541,
|
| 374 |
+
"eval_samples_per_second": 31.979,
|
| 375 |
+
"eval_steps_per_second": 2.079,
|
| 376 |
+
"step": 1055
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.0,
|
| 380 |
+
"eval_exact_match": 16.0,
|
| 381 |
+
"eval_f1": 22.22857142857142,
|
| 382 |
+
"eval_qa_bleu": 10.69246525438524,
|
| 383 |
+
"eval_qa_exact_match": 0.14,
|
| 384 |
+
"eval_recite_bleu": 33.64683237936858,
|
| 385 |
+
"eval_recite_exact_match": 0.025,
|
| 386 |
+
"step": 1055
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 5.0,
|
| 390 |
+
"learning_rate": 1.5778443113772455e-05,
|
| 391 |
+
"loss": 0.2129,
|
| 392 |
+
"step": 1056
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 5.11,
|
| 396 |
+
"learning_rate": 1.5449101796407184e-05,
|
| 397 |
+
"loss": 0.1559,
|
| 398 |
+
"step": 1078
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 5.21,
|
| 402 |
+
"learning_rate": 1.5119760479041917e-05,
|
| 403 |
+
"loss": 0.1631,
|
| 404 |
+
"step": 1100
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.32,
|
| 408 |
+
"learning_rate": 1.4790419161676647e-05,
|
| 409 |
+
"loss": 0.1614,
|
| 410 |
+
"step": 1122
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 5.42,
|
| 414 |
+
"learning_rate": 1.4461077844311378e-05,
|
| 415 |
+
"loss": 0.1636,
|
| 416 |
+
"step": 1144
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 5.53,
|
| 420 |
+
"learning_rate": 1.4131736526946109e-05,
|
| 421 |
+
"loss": 0.1645,
|
| 422 |
+
"step": 1166
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 5.63,
|
| 426 |
+
"learning_rate": 1.3802395209580839e-05,
|
| 427 |
+
"loss": 0.1613,
|
| 428 |
+
"step": 1188
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 5.73,
|
| 432 |
+
"learning_rate": 1.3473053892215568e-05,
|
| 433 |
+
"loss": 0.1724,
|
| 434 |
+
"step": 1210
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 5.84,
|
| 438 |
+
"learning_rate": 1.31437125748503e-05,
|
| 439 |
+
"loss": 0.1643,
|
| 440 |
+
"step": 1232
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 5.94,
|
| 444 |
+
"learning_rate": 1.281437125748503e-05,
|
| 445 |
+
"loss": 0.17,
|
| 446 |
+
"step": 1254
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 6.0,
|
| 450 |
+
"eval_accuracy": 0.6964456035767511,
|
| 451 |
+
"eval_loss": 0.44896385073661804,
|
| 452 |
+
"eval_runtime": 6.3718,
|
| 453 |
+
"eval_samples_per_second": 31.389,
|
| 454 |
+
"eval_steps_per_second": 2.04,
|
| 455 |
+
"step": 1266
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.0,
|
| 459 |
+
"eval_exact_match": 15.0,
|
| 460 |
+
"eval_f1": 21.070238095238086,
|
| 461 |
+
"eval_qa_bleu": 11.294234950233255,
|
| 462 |
+
"eval_qa_exact_match": 0.13,
|
| 463 |
+
"eval_recite_bleu": 35.4728355349377,
|
| 464 |
+
"eval_recite_exact_match": 0.025,
|
| 465 |
+
"step": 1266
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 6.05,
|
| 469 |
+
"learning_rate": 1.2485029940119762e-05,
|
| 470 |
+
"loss": 0.1441,
|
| 471 |
+
"step": 1276
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 6.15,
|
| 475 |
+
"learning_rate": 1.2155688622754491e-05,
|
| 476 |
+
"loss": 0.1356,
|
| 477 |
+
"step": 1298
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 6.26,
|
| 481 |
+
"learning_rate": 1.182634730538922e-05,
|
| 482 |
+
"loss": 0.1363,
|
| 483 |
+
"step": 1320
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 6.36,
|
| 487 |
+
"learning_rate": 1.1497005988023954e-05,
|
| 488 |
+
"loss": 0.137,
|
| 489 |
+
"step": 1342
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 6.46,
|
| 493 |
+
"learning_rate": 1.1167664670658683e-05,
|
| 494 |
+
"loss": 0.1382,
|
| 495 |
+
"step": 1364
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 6.57,
|
| 499 |
+
"learning_rate": 1.0838323353293413e-05,
|
| 500 |
+
"loss": 0.1416,
|
| 501 |
+
"step": 1386
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 6.67,
|
| 505 |
+
"learning_rate": 1.0508982035928144e-05,
|
| 506 |
+
"loss": 0.1389,
|
| 507 |
+
"step": 1408
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 6.78,
|
| 511 |
+
"learning_rate": 1.0179640718562873e-05,
|
| 512 |
+
"loss": 0.1392,
|
| 513 |
+
"step": 1430
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 6.88,
|
| 517 |
+
"learning_rate": 9.850299401197606e-06,
|
| 518 |
+
"loss": 0.1398,
|
| 519 |
+
"step": 1452
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 6.99,
|
| 523 |
+
"learning_rate": 9.520958083832336e-06,
|
| 524 |
+
"loss": 0.134,
|
| 525 |
+
"step": 1474
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 7.0,
|
| 529 |
+
"eval_accuracy": 0.697876304023845,
|
| 530 |
+
"eval_loss": 0.43685418367385864,
|
| 531 |
+
"eval_runtime": 6.2539,
|
| 532 |
+
"eval_samples_per_second": 31.98,
|
| 533 |
+
"eval_steps_per_second": 2.079,
|
| 534 |
+
"step": 1477
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.0,
|
| 538 |
+
"eval_exact_match": 16.5,
|
| 539 |
+
"eval_f1": 22.119444444444444,
|
| 540 |
+
"eval_qa_bleu": 13.989667861778496,
|
| 541 |
+
"eval_qa_exact_match": 0.14,
|
| 542 |
+
"eval_recite_bleu": 42.188868266934165,
|
| 543 |
+
"eval_recite_exact_match": 0.04,
|
| 544 |
+
"step": 1477
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 7.09,
|
| 548 |
+
"learning_rate": 9.191616766467065e-06,
|
| 549 |
+
"loss": 0.1188,
|
| 550 |
+
"step": 1496
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 7.19,
|
| 554 |
+
"learning_rate": 8.862275449101796e-06,
|
| 555 |
+
"loss": 0.1207,
|
| 556 |
+
"step": 1518
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 7.3,
|
| 560 |
+
"learning_rate": 8.532934131736528e-06,
|
| 561 |
+
"loss": 0.1223,
|
| 562 |
+
"step": 1540
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.4,
|
| 566 |
+
"learning_rate": 8.203592814371257e-06,
|
| 567 |
+
"loss": 0.115,
|
| 568 |
+
"step": 1562
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 7.51,
|
| 572 |
+
"learning_rate": 7.874251497005988e-06,
|
| 573 |
+
"loss": 0.1139,
|
| 574 |
+
"step": 1584
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 7.61,
|
| 578 |
+
"learning_rate": 7.544910179640718e-06,
|
| 579 |
+
"loss": 0.117,
|
| 580 |
+
"step": 1606
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 7.72,
|
| 584 |
+
"learning_rate": 7.215568862275449e-06,
|
| 585 |
+
"loss": 0.1203,
|
| 586 |
+
"step": 1628
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 7.82,
|
| 590 |
+
"learning_rate": 6.8862275449101795e-06,
|
| 591 |
+
"loss": 0.1229,
|
| 592 |
+
"step": 1650
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 7.92,
|
| 596 |
+
"learning_rate": 6.556886227544911e-06,
|
| 597 |
+
"loss": 0.1206,
|
| 598 |
+
"step": 1672
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 8.0,
|
| 602 |
+
"eval_accuracy": 0.6986959761549926,
|
| 603 |
+
"eval_loss": 0.43722403049468994,
|
| 604 |
+
"eval_runtime": 6.191,
|
| 605 |
+
"eval_samples_per_second": 32.305,
|
| 606 |
+
"eval_steps_per_second": 2.1,
|
| 607 |
+
"step": 1688
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 8.0,
|
| 611 |
+
"eval_exact_match": 20.0,
|
| 612 |
+
"eval_f1": 25.83571428571428,
|
| 613 |
+
"eval_qa_bleu": 8.333490045944334,
|
| 614 |
+
"eval_qa_exact_match": 0.18,
|
| 615 |
+
"eval_recite_bleu": 43.68492978075338,
|
| 616 |
+
"eval_recite_exact_match": 0.065,
|
| 617 |
+
"step": 1688
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 8.03,
|
| 621 |
+
"learning_rate": 6.22754491017964e-06,
|
| 622 |
+
"loss": 0.1145,
|
| 623 |
+
"step": 1694
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 8.13,
|
| 627 |
+
"learning_rate": 5.898203592814371e-06,
|
| 628 |
+
"loss": 0.1066,
|
| 629 |
+
"step": 1716
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 8.24,
|
| 633 |
+
"learning_rate": 5.568862275449102e-06,
|
| 634 |
+
"loss": 0.1064,
|
| 635 |
+
"step": 1738
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 8.34,
|
| 639 |
+
"learning_rate": 5.239520958083833e-06,
|
| 640 |
+
"loss": 0.1077,
|
| 641 |
+
"step": 1760
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 8.45,
|
| 645 |
+
"learning_rate": 4.9101796407185625e-06,
|
| 646 |
+
"loss": 0.1108,
|
| 647 |
+
"step": 1782
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 8.55,
|
| 651 |
+
"learning_rate": 4.580838323353294e-06,
|
| 652 |
+
"loss": 0.1078,
|
| 653 |
+
"step": 1804
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 8.65,
|
| 657 |
+
"learning_rate": 4.251497005988024e-06,
|
| 658 |
+
"loss": 0.1072,
|
| 659 |
+
"step": 1826
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 8.76,
|
| 663 |
+
"learning_rate": 3.922155688622755e-06,
|
| 664 |
+
"loss": 0.1098,
|
| 665 |
+
"step": 1848
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 8.86,
|
| 669 |
+
"learning_rate": 3.592814371257485e-06,
|
| 670 |
+
"loss": 0.1093,
|
| 671 |
+
"step": 1870
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 8.97,
|
| 675 |
+
"learning_rate": 3.2634730538922155e-06,
|
| 676 |
+
"loss": 0.1081,
|
| 677 |
+
"step": 1892
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 9.0,
|
| 681 |
+
"eval_accuracy": 0.698725782414307,
|
| 682 |
+
"eval_loss": 0.44227516651153564,
|
| 683 |
+
"eval_runtime": 6.2856,
|
| 684 |
+
"eval_samples_per_second": 31.819,
|
| 685 |
+
"eval_steps_per_second": 2.068,
|
| 686 |
+
"step": 1899
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 9.0,
|
| 690 |
+
"eval_exact_match": 17.5,
|
| 691 |
+
"eval_f1": 22.30357142857143,
|
| 692 |
+
"eval_qa_bleu": 10.817736383091892,
|
| 693 |
+
"eval_qa_exact_match": 0.15,
|
| 694 |
+
"eval_recite_bleu": 43.37666725316897,
|
| 695 |
+
"eval_recite_exact_match": 0.06,
|
| 696 |
+
"step": 1899
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 9.07,
|
| 700 |
+
"learning_rate": 2.9341317365269463e-06,
|
| 701 |
+
"loss": 0.1079,
|
| 702 |
+
"step": 1914
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 9.18,
|
| 706 |
+
"learning_rate": 2.6047904191616767e-06,
|
| 707 |
+
"loss": 0.102,
|
| 708 |
+
"step": 1936
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 9.28,
|
| 712 |
+
"learning_rate": 2.2754491017964075e-06,
|
| 713 |
+
"loss": 0.1003,
|
| 714 |
+
"step": 1958
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 9.38,
|
| 718 |
+
"learning_rate": 1.9461077844311374e-06,
|
| 719 |
+
"loss": 0.1023,
|
| 720 |
+
"step": 1980
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 9.49,
|
| 724 |
+
"learning_rate": 1.6167664670658684e-06,
|
| 725 |
+
"loss": 0.1003,
|
| 726 |
+
"step": 2002
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 9.59,
|
| 730 |
+
"learning_rate": 1.287425149700599e-06,
|
| 731 |
+
"loss": 0.1058,
|
| 732 |
+
"step": 2024
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 9.7,
|
| 736 |
+
"learning_rate": 9.580838323353293e-07,
|
| 737 |
+
"loss": 0.1067,
|
| 738 |
+
"step": 2046
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 9.8,
|
| 742 |
+
"learning_rate": 6.287425149700599e-07,
|
| 743 |
+
"loss": 0.1047,
|
| 744 |
+
"step": 2068
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 9.91,
|
| 748 |
+
"learning_rate": 2.994011976047904e-07,
|
| 749 |
+
"loss": 0.1053,
|
| 750 |
+
"step": 2090
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 10.0,
|
| 754 |
+
"eval_accuracy": 0.6988971684053651,
|
| 755 |
+
"eval_loss": 0.4435840845108032,
|
| 756 |
+
"eval_runtime": 6.3653,
|
| 757 |
+
"eval_samples_per_second": 31.42,
|
| 758 |
+
"eval_steps_per_second": 2.042,
|
| 759 |
+
"step": 2110
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 10.0,
|
| 763 |
+
"eval_exact_match": 20.5,
|
| 764 |
+
"eval_f1": 25.95833333333333,
|
| 765 |
+
"eval_qa_bleu": 12.082719988904218,
|
| 766 |
+
"eval_qa_exact_match": 0.18,
|
| 767 |
+
"eval_recite_bleu": 44.74289478974747,
|
| 768 |
+
"eval_recite_exact_match": 0.075,
|
| 769 |
+
"step": 2110
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 10.0,
|
| 773 |
+
"step": 2110,
|
| 774 |
+
"total_flos": 9.64020042848256e+16,
|
| 775 |
+
"train_loss": 0.29445283661521443,
|
| 776 |
+
"train_runtime": 3197.9615,
|
| 777 |
+
"train_samples_per_second": 10.547,
|
| 778 |
+
"train_steps_per_second": 0.66
|
| 779 |
+
}
|
| 780 |
+
],
|
| 781 |
+
"logging_steps": 22,
|
| 782 |
+
"max_steps": 2110,
|
| 783 |
+
"num_train_epochs": 10,
|
| 784 |
+
"save_steps": 500,
|
| 785 |
+
"total_flos": 9.64020042848256e+16,
|
| 786 |
+
"trial_name": null,
|
| 787 |
+
"trial_params": null
|
| 788 |
+
}
|