End of training
Browse files- README.md +16 -4
- all_results.json +17 -17
- eval_results.json +13 -13
- train_results.json +5 -5
- trainer_state.json +950 -12
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_nq_train600_eval300_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
|
|
@@ -15,10 +27,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# lmind_nq_train600_eval300_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.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 3 |
base_model: gpt2-xl
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- tyzhu/lmind_nq_train600_eval300_v1_recite_qa
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
| 10 |
model-index:
|
| 11 |
- name: lmind_nq_train600_eval300_v1_recite_qa_gpt2-xl
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Causal Language Modeling
|
| 15 |
+
type: text-generation
|
| 16 |
+
dataset:
|
| 17 |
+
name: tyzhu/lmind_nq_train600_eval300_v1_recite_qa
|
| 18 |
+
type: tyzhu/lmind_nq_train600_eval300_v1_recite_qa
|
| 19 |
+
metrics:
|
| 20 |
+
- name: Accuracy
|
| 21 |
+
type: accuracy
|
| 22 |
+
value: 0.8407058823529412
|
| 23 |
---
|
| 24 |
|
| 25 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 27 |
|
| 28 |
# lmind_nq_train600_eval300_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_nq_train600_eval300_v1_recite_qa dataset.
|
| 31 |
It achieves the following results on the evaluation set:
|
| 32 |
+
- Loss: 0.3660
|
| 33 |
+
- Accuracy: 0.8407
|
| 34 |
|
| 35 |
## Model description
|
| 36 |
|
all_results.json
CHANGED
|
@@ -1,21 +1,21 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_exact_match":
|
| 5 |
-
"eval_f1":
|
| 6 |
-
"eval_loss": 0.
|
| 7 |
-
"eval_qa_bleu":
|
| 8 |
-
"eval_qa_exact_match": 0.
|
| 9 |
-
"eval_recite_bleu":
|
| 10 |
-
"eval_recite_exact_match": 0.
|
| 11 |
-
"eval_runtime": 8.
|
| 12 |
"eval_samples": 300,
|
| 13 |
-
"eval_samples_per_second": 36.
|
| 14 |
-
"eval_steps_per_second": 2.
|
| 15 |
-
"perplexity": 1.
|
| 16 |
-
"train_loss": 0.
|
| 17 |
-
"train_runtime":
|
| 18 |
"train_samples": 1483,
|
| 19 |
-
"train_samples_per_second":
|
| 20 |
-
"train_steps_per_second": 0.
|
| 21 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.8407058823529412,
|
| 4 |
+
"eval_exact_match": 12.0,
|
| 5 |
+
"eval_f1": 15.791798941798934,
|
| 6 |
+
"eval_loss": 0.365990549325943,
|
| 7 |
+
"eval_qa_bleu": 9.37996870880597,
|
| 8 |
+
"eval_qa_exact_match": 0.08666666666666667,
|
| 9 |
+
"eval_recite_bleu": 26.43773861474592,
|
| 10 |
+
"eval_recite_exact_match": 0.1,
|
| 11 |
+
"eval_runtime": 8.2014,
|
| 12 |
"eval_samples": 300,
|
| 13 |
+
"eval_samples_per_second": 36.579,
|
| 14 |
+
"eval_steps_per_second": 2.317,
|
| 15 |
+
"perplexity": 1.441941615269907,
|
| 16 |
+
"train_loss": 0.03236631401764449,
|
| 17 |
+
"train_runtime": 2828.6852,
|
| 18 |
"train_samples": 1483,
|
| 19 |
+
"train_samples_per_second": 10.485,
|
| 20 |
+
"train_steps_per_second": 0.658
|
| 21 |
}
|
eval_results.json
CHANGED
|
@@ -1,16 +1,16 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_exact_match":
|
| 5 |
-
"eval_f1":
|
| 6 |
-
"eval_loss": 0.
|
| 7 |
-
"eval_qa_bleu":
|
| 8 |
-
"eval_qa_exact_match": 0.
|
| 9 |
-
"eval_recite_bleu":
|
| 10 |
-
"eval_recite_exact_match": 0.
|
| 11 |
-
"eval_runtime": 8.
|
| 12 |
"eval_samples": 300,
|
| 13 |
-
"eval_samples_per_second": 36.
|
| 14 |
-
"eval_steps_per_second": 2.
|
| 15 |
-
"perplexity": 1.
|
| 16 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.8407058823529412,
|
| 4 |
+
"eval_exact_match": 12.0,
|
| 5 |
+
"eval_f1": 15.791798941798934,
|
| 6 |
+
"eval_loss": 0.365990549325943,
|
| 7 |
+
"eval_qa_bleu": 9.37996870880597,
|
| 8 |
+
"eval_qa_exact_match": 0.08666666666666667,
|
| 9 |
+
"eval_recite_bleu": 26.43773861474592,
|
| 10 |
+
"eval_recite_exact_match": 0.1,
|
| 11 |
+
"eval_runtime": 8.2014,
|
| 12 |
"eval_samples": 300,
|
| 13 |
+
"eval_samples_per_second": 36.579,
|
| 14 |
+
"eval_steps_per_second": 2.317,
|
| 15 |
+
"perplexity": 1.441941615269907
|
| 16 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 1483,
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"train_loss": 0.03236631401764449,
|
| 4 |
+
"train_runtime": 2828.6852,
|
| 5 |
"train_samples": 1483,
|
| 6 |
+
"train_samples_per_second": 10.485,
|
| 7 |
+
"train_steps_per_second": 0.658
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 47,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -928,20 +928,958 @@
|
|
| 928 |
"step": 930
|
| 929 |
},
|
| 930 |
{
|
| 931 |
-
"epoch": 10.
|
| 932 |
-
"
|
| 933 |
-
"
|
| 934 |
-
"
|
| 935 |
-
|
| 936 |
-
|
| 937 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 938 |
}
|
| 939 |
],
|
| 940 |
"logging_steps": 10,
|
| 941 |
-
"max_steps":
|
| 942 |
-
"num_train_epochs":
|
| 943 |
"save_steps": 500,
|
| 944 |
-
"total_flos":
|
| 945 |
"trial_name": null,
|
| 946 |
"trial_params": null
|
| 947 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
"eval_steps": 47,
|
| 6 |
+
"global_step": 1860,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 928 |
"step": 930
|
| 929 |
},
|
| 930 |
{
|
| 931 |
+
"epoch": 10.11,
|
| 932 |
+
"learning_rate": 3e-05,
|
| 933 |
+
"loss": 0.0637,
|
| 934 |
+
"step": 940
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 10.11,
|
| 938 |
+
"eval_accuracy": 0.8411764705882353,
|
| 939 |
+
"eval_loss": 0.3618181347846985,
|
| 940 |
+
"eval_runtime": 11.3655,
|
| 941 |
+
"eval_samples_per_second": 26.396,
|
| 942 |
+
"eval_steps_per_second": 1.672,
|
| 943 |
+
"step": 940
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 10.11,
|
| 947 |
+
"eval_exact_match": 11.666666666666666,
|
| 948 |
+
"eval_f1": 16.085185185185182,
|
| 949 |
+
"eval_qa_bleu": 8.74902056299004,
|
| 950 |
+
"eval_qa_exact_match": 0.09,
|
| 951 |
+
"eval_recite_bleu": 24.197336184324623,
|
| 952 |
+
"eval_recite_exact_match": 0.056666666666666664,
|
| 953 |
+
"step": 940
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 10.22,
|
| 957 |
+
"learning_rate": 3e-05,
|
| 958 |
+
"loss": 0.0643,
|
| 959 |
+
"step": 950
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 10.32,
|
| 963 |
+
"learning_rate": 3e-05,
|
| 964 |
+
"loss": 0.0694,
|
| 965 |
+
"step": 960
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 10.43,
|
| 969 |
+
"learning_rate": 3e-05,
|
| 970 |
+
"loss": 0.0691,
|
| 971 |
+
"step": 970
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 10.54,
|
| 975 |
+
"learning_rate": 3e-05,
|
| 976 |
+
"loss": 0.069,
|
| 977 |
+
"step": 980
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 10.61,
|
| 981 |
+
"eval_accuracy": 0.8414509803921568,
|
| 982 |
+
"eval_loss": 0.3563094437122345,
|
| 983 |
+
"eval_runtime": 8.2328,
|
| 984 |
+
"eval_samples_per_second": 36.439,
|
| 985 |
+
"eval_steps_per_second": 2.308,
|
| 986 |
+
"step": 987
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 10.61,
|
| 990 |
+
"eval_exact_match": 11.333333333333334,
|
| 991 |
+
"eval_f1": 15.137566137566136,
|
| 992 |
+
"eval_qa_bleu": 9.554429137249072,
|
| 993 |
+
"eval_qa_exact_match": 0.07666666666666666,
|
| 994 |
+
"eval_recite_bleu": 23.910757721350535,
|
| 995 |
+
"eval_recite_exact_match": 0.07333333333333333,
|
| 996 |
+
"step": 987
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 10.65,
|
| 1000 |
+
"learning_rate": 3e-05,
|
| 1001 |
+
"loss": 0.0698,
|
| 1002 |
+
"step": 990
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 10.75,
|
| 1006 |
+
"learning_rate": 3e-05,
|
| 1007 |
+
"loss": 0.0696,
|
| 1008 |
+
"step": 1000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 10.86,
|
| 1012 |
+
"learning_rate": 3e-05,
|
| 1013 |
+
"loss": 0.0682,
|
| 1014 |
+
"step": 1010
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 10.97,
|
| 1018 |
+
"learning_rate": 3e-05,
|
| 1019 |
+
"loss": 0.071,
|
| 1020 |
+
"step": 1020
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 11.08,
|
| 1024 |
+
"learning_rate": 3e-05,
|
| 1025 |
+
"loss": 0.0636,
|
| 1026 |
+
"step": 1030
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 11.12,
|
| 1030 |
+
"eval_accuracy": 0.840875816993464,
|
| 1031 |
+
"eval_loss": 0.36504313349723816,
|
| 1032 |
+
"eval_runtime": 8.5962,
|
| 1033 |
+
"eval_samples_per_second": 34.899,
|
| 1034 |
+
"eval_steps_per_second": 2.21,
|
| 1035 |
+
"step": 1034
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 11.12,
|
| 1039 |
+
"eval_exact_match": 9.0,
|
| 1040 |
+
"eval_f1": 13.016137566137564,
|
| 1041 |
+
"eval_qa_bleu": 6.847070322558309,
|
| 1042 |
+
"eval_qa_exact_match": 0.06,
|
| 1043 |
+
"eval_recite_bleu": 20.79272983231636,
|
| 1044 |
+
"eval_recite_exact_match": 0.043333333333333335,
|
| 1045 |
+
"step": 1034
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 11.18,
|
| 1049 |
+
"learning_rate": 3e-05,
|
| 1050 |
+
"loss": 0.0633,
|
| 1051 |
+
"step": 1040
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 11.29,
|
| 1055 |
+
"learning_rate": 3e-05,
|
| 1056 |
+
"loss": 0.0626,
|
| 1057 |
+
"step": 1050
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 11.4,
|
| 1061 |
+
"learning_rate": 3e-05,
|
| 1062 |
+
"loss": 0.067,
|
| 1063 |
+
"step": 1060
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 11.51,
|
| 1067 |
+
"learning_rate": 3e-05,
|
| 1068 |
+
"loss": 0.0658,
|
| 1069 |
+
"step": 1070
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 11.61,
|
| 1073 |
+
"learning_rate": 3e-05,
|
| 1074 |
+
"loss": 0.0663,
|
| 1075 |
+
"step": 1080
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 11.62,
|
| 1079 |
+
"eval_accuracy": 0.8418039215686275,
|
| 1080 |
+
"eval_loss": 0.3554021716117859,
|
| 1081 |
+
"eval_runtime": 8.476,
|
| 1082 |
+
"eval_samples_per_second": 35.394,
|
| 1083 |
+
"eval_steps_per_second": 2.242,
|
| 1084 |
+
"step": 1081
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 11.62,
|
| 1088 |
+
"eval_exact_match": 10.666666666666666,
|
| 1089 |
+
"eval_f1": 14.759547859547856,
|
| 1090 |
+
"eval_qa_bleu": 7.764051246001461,
|
| 1091 |
+
"eval_qa_exact_match": 0.07666666666666666,
|
| 1092 |
+
"eval_recite_bleu": 24.921719211518564,
|
| 1093 |
+
"eval_recite_exact_match": 0.08333333333333333,
|
| 1094 |
+
"step": 1081
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 11.72,
|
| 1098 |
+
"learning_rate": 3e-05,
|
| 1099 |
+
"loss": 0.0674,
|
| 1100 |
+
"step": 1090
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 11.83,
|
| 1104 |
+
"learning_rate": 3e-05,
|
| 1105 |
+
"loss": 0.0716,
|
| 1106 |
+
"step": 1100
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 11.94,
|
| 1110 |
+
"learning_rate": 3e-05,
|
| 1111 |
+
"loss": 0.0689,
|
| 1112 |
+
"step": 1110
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 12.04,
|
| 1116 |
+
"learning_rate": 3e-05,
|
| 1117 |
+
"loss": 0.0636,
|
| 1118 |
+
"step": 1120
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 12.13,
|
| 1122 |
+
"eval_accuracy": 0.840875816993464,
|
| 1123 |
+
"eval_loss": 0.3674863576889038,
|
| 1124 |
+
"eval_runtime": 8.463,
|
| 1125 |
+
"eval_samples_per_second": 35.449,
|
| 1126 |
+
"eval_steps_per_second": 2.245,
|
| 1127 |
+
"step": 1128
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 12.13,
|
| 1131 |
+
"eval_exact_match": 10.666666666666666,
|
| 1132 |
+
"eval_f1": 14.650553150553147,
|
| 1133 |
+
"eval_qa_bleu": 9.093912749957752,
|
| 1134 |
+
"eval_qa_exact_match": 0.06666666666666667,
|
| 1135 |
+
"eval_recite_bleu": 24.62110487495627,
|
| 1136 |
+
"eval_recite_exact_match": 0.06666666666666667,
|
| 1137 |
+
"step": 1128
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 12.15,
|
| 1141 |
+
"learning_rate": 3e-05,
|
| 1142 |
+
"loss": 0.0617,
|
| 1143 |
+
"step": 1130
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 12.26,
|
| 1147 |
+
"learning_rate": 3e-05,
|
| 1148 |
+
"loss": 0.0629,
|
| 1149 |
+
"step": 1140
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 12.37,
|
| 1153 |
+
"learning_rate": 3e-05,
|
| 1154 |
+
"loss": 0.061,
|
| 1155 |
+
"step": 1150
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 12.47,
|
| 1159 |
+
"learning_rate": 3e-05,
|
| 1160 |
+
"loss": 0.0621,
|
| 1161 |
+
"step": 1160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 12.58,
|
| 1165 |
+
"learning_rate": 3e-05,
|
| 1166 |
+
"loss": 0.0632,
|
| 1167 |
+
"step": 1170
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 12.63,
|
| 1171 |
+
"eval_accuracy": 0.8418039215686275,
|
| 1172 |
+
"eval_loss": 0.3590278625488281,
|
| 1173 |
+
"eval_runtime": 8.472,
|
| 1174 |
+
"eval_samples_per_second": 35.411,
|
| 1175 |
+
"eval_steps_per_second": 2.243,
|
| 1176 |
+
"step": 1175
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 12.63,
|
| 1180 |
+
"eval_exact_match": 12.666666666666666,
|
| 1181 |
+
"eval_f1": 16.126984126984127,
|
| 1182 |
+
"eval_qa_bleu": 9.250714961910592,
|
| 1183 |
+
"eval_qa_exact_match": 0.08666666666666667,
|
| 1184 |
+
"eval_recite_bleu": 26.441847770240706,
|
| 1185 |
+
"eval_recite_exact_match": 0.08333333333333333,
|
| 1186 |
+
"step": 1175
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 12.69,
|
| 1190 |
+
"learning_rate": 3e-05,
|
| 1191 |
+
"loss": 0.0665,
|
| 1192 |
+
"step": 1180
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 12.8,
|
| 1196 |
+
"learning_rate": 3e-05,
|
| 1197 |
+
"loss": 0.0656,
|
| 1198 |
+
"step": 1190
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 12.9,
|
| 1202 |
+
"learning_rate": 3e-05,
|
| 1203 |
+
"loss": 0.0647,
|
| 1204 |
+
"step": 1200
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 13.01,
|
| 1208 |
+
"learning_rate": 3e-05,
|
| 1209 |
+
"loss": 0.0656,
|
| 1210 |
+
"step": 1210
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 13.12,
|
| 1214 |
+
"learning_rate": 3e-05,
|
| 1215 |
+
"loss": 0.0605,
|
| 1216 |
+
"step": 1220
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 13.14,
|
| 1220 |
+
"eval_accuracy": 0.8412679738562091,
|
| 1221 |
+
"eval_loss": 0.36349713802337646,
|
| 1222 |
+
"eval_runtime": 8.5252,
|
| 1223 |
+
"eval_samples_per_second": 35.19,
|
| 1224 |
+
"eval_steps_per_second": 2.229,
|
| 1225 |
+
"step": 1222
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 13.14,
|
| 1229 |
+
"eval_exact_match": 11.0,
|
| 1230 |
+
"eval_f1": 14.664021164021163,
|
| 1231 |
+
"eval_qa_bleu": 8.790900925095583,
|
| 1232 |
+
"eval_qa_exact_match": 0.08,
|
| 1233 |
+
"eval_recite_bleu": 25.058252467710584,
|
| 1234 |
+
"eval_recite_exact_match": 0.07,
|
| 1235 |
+
"step": 1222
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 13.23,
|
| 1239 |
+
"learning_rate": 3e-05,
|
| 1240 |
+
"loss": 0.0594,
|
| 1241 |
+
"step": 1230
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 13.33,
|
| 1245 |
+
"learning_rate": 3e-05,
|
| 1246 |
+
"loss": 0.0612,
|
| 1247 |
+
"step": 1240
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 13.44,
|
| 1251 |
+
"learning_rate": 3e-05,
|
| 1252 |
+
"loss": 0.0611,
|
| 1253 |
+
"step": 1250
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 13.55,
|
| 1257 |
+
"learning_rate": 3e-05,
|
| 1258 |
+
"loss": 0.0621,
|
| 1259 |
+
"step": 1260
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 13.65,
|
| 1263 |
+
"eval_accuracy": 0.8413071895424836,
|
| 1264 |
+
"eval_loss": 0.35581451654434204,
|
| 1265 |
+
"eval_runtime": 8.6726,
|
| 1266 |
+
"eval_samples_per_second": 34.592,
|
| 1267 |
+
"eval_steps_per_second": 2.191,
|
| 1268 |
+
"step": 1269
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 13.65,
|
| 1272 |
+
"eval_exact_match": 12.666666666666666,
|
| 1273 |
+
"eval_f1": 16.87830687830688,
|
| 1274 |
+
"eval_qa_bleu": 9.046507615706695,
|
| 1275 |
+
"eval_qa_exact_match": 0.09333333333333334,
|
| 1276 |
+
"eval_recite_bleu": 26.358751880820996,
|
| 1277 |
+
"eval_recite_exact_match": 0.08333333333333333,
|
| 1278 |
+
"step": 1269
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 13.66,
|
| 1282 |
+
"learning_rate": 3e-05,
|
| 1283 |
+
"loss": 0.0662,
|
| 1284 |
+
"step": 1270
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 13.76,
|
| 1288 |
+
"learning_rate": 3e-05,
|
| 1289 |
+
"loss": 0.0636,
|
| 1290 |
+
"step": 1280
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 13.87,
|
| 1294 |
+
"learning_rate": 3e-05,
|
| 1295 |
+
"loss": 0.0651,
|
| 1296 |
+
"step": 1290
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 13.98,
|
| 1300 |
+
"learning_rate": 3e-05,
|
| 1301 |
+
"loss": 0.0638,
|
| 1302 |
+
"step": 1300
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 14.09,
|
| 1306 |
+
"learning_rate": 3e-05,
|
| 1307 |
+
"loss": 0.0579,
|
| 1308 |
+
"step": 1310
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 14.15,
|
| 1312 |
+
"eval_accuracy": 0.840967320261438,
|
| 1313 |
+
"eval_loss": 0.36824384331703186,
|
| 1314 |
+
"eval_runtime": 8.5554,
|
| 1315 |
+
"eval_samples_per_second": 35.065,
|
| 1316 |
+
"eval_steps_per_second": 2.221,
|
| 1317 |
+
"step": 1316
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 14.15,
|
| 1321 |
+
"eval_exact_match": 8.666666666666666,
|
| 1322 |
+
"eval_f1": 12.518518518518519,
|
| 1323 |
+
"eval_qa_bleu": 5.946814014717943,
|
| 1324 |
+
"eval_qa_exact_match": 0.06,
|
| 1325 |
+
"eval_recite_bleu": 23.295443887810283,
|
| 1326 |
+
"eval_recite_exact_match": 0.06,
|
| 1327 |
+
"step": 1316
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 14.19,
|
| 1331 |
+
"learning_rate": 3e-05,
|
| 1332 |
+
"loss": 0.0577,
|
| 1333 |
+
"step": 1320
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 14.3,
|
| 1337 |
+
"learning_rate": 3e-05,
|
| 1338 |
+
"loss": 0.0603,
|
| 1339 |
+
"step": 1330
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 14.41,
|
| 1343 |
+
"learning_rate": 3e-05,
|
| 1344 |
+
"loss": 0.0602,
|
| 1345 |
+
"step": 1340
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 14.52,
|
| 1349 |
+
"learning_rate": 3e-05,
|
| 1350 |
+
"loss": 0.0613,
|
| 1351 |
+
"step": 1350
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 14.62,
|
| 1355 |
+
"learning_rate": 3e-05,
|
| 1356 |
+
"loss": 0.0618,
|
| 1357 |
+
"step": 1360
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 14.66,
|
| 1361 |
+
"eval_accuracy": 0.8405490196078431,
|
| 1362 |
+
"eval_loss": 0.36532720923423767,
|
| 1363 |
+
"eval_runtime": 8.6431,
|
| 1364 |
+
"eval_samples_per_second": 34.71,
|
| 1365 |
+
"eval_steps_per_second": 2.198,
|
| 1366 |
+
"step": 1363
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 14.66,
|
| 1370 |
+
"eval_exact_match": 11.0,
|
| 1371 |
+
"eval_f1": 15.13042328042328,
|
| 1372 |
+
"eval_qa_bleu": 8.759905510533487,
|
| 1373 |
+
"eval_qa_exact_match": 0.07666666666666666,
|
| 1374 |
+
"eval_recite_bleu": 26.14569012302347,
|
| 1375 |
+
"eval_recite_exact_match": 0.07,
|
| 1376 |
+
"step": 1363
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 14.73,
|
| 1380 |
+
"learning_rate": 3e-05,
|
| 1381 |
+
"loss": 0.064,
|
| 1382 |
+
"step": 1370
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 14.84,
|
| 1386 |
+
"learning_rate": 3e-05,
|
| 1387 |
+
"loss": 0.0631,
|
| 1388 |
+
"step": 1380
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 14.95,
|
| 1392 |
+
"learning_rate": 3e-05,
|
| 1393 |
+
"loss": 0.0611,
|
| 1394 |
+
"step": 1390
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 15.05,
|
| 1398 |
+
"learning_rate": 3e-05,
|
| 1399 |
+
"loss": 0.0582,
|
| 1400 |
+
"step": 1400
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 15.16,
|
| 1404 |
+
"learning_rate": 3e-05,
|
| 1405 |
+
"loss": 0.0552,
|
| 1406 |
+
"step": 1410
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 15.16,
|
| 1410 |
+
"eval_accuracy": 0.8413071895424836,
|
| 1411 |
+
"eval_loss": 0.3660917282104492,
|
| 1412 |
+
"eval_runtime": 8.4709,
|
| 1413 |
+
"eval_samples_per_second": 35.416,
|
| 1414 |
+
"eval_steps_per_second": 2.243,
|
| 1415 |
+
"step": 1410
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 15.16,
|
| 1419 |
+
"eval_exact_match": 13.0,
|
| 1420 |
+
"eval_f1": 16.49047619047619,
|
| 1421 |
+
"eval_qa_bleu": 13.081354815231128,
|
| 1422 |
+
"eval_qa_exact_match": 0.09333333333333334,
|
| 1423 |
+
"eval_recite_bleu": 28.050116321012986,
|
| 1424 |
+
"eval_recite_exact_match": 0.09333333333333334,
|
| 1425 |
+
"step": 1410
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 15.27,
|
| 1429 |
+
"learning_rate": 3e-05,
|
| 1430 |
+
"loss": 0.0566,
|
| 1431 |
+
"step": 1420
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 15.38,
|
| 1435 |
+
"learning_rate": 3e-05,
|
| 1436 |
+
"loss": 0.059,
|
| 1437 |
+
"step": 1430
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 15.48,
|
| 1441 |
+
"learning_rate": 3e-05,
|
| 1442 |
+
"loss": 0.0611,
|
| 1443 |
+
"step": 1440
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 15.59,
|
| 1447 |
+
"learning_rate": 3e-05,
|
| 1448 |
+
"loss": 0.0619,
|
| 1449 |
+
"step": 1450
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 15.67,
|
| 1453 |
+
"eval_accuracy": 0.8416209150326798,
|
| 1454 |
+
"eval_loss": 0.3595679700374603,
|
| 1455 |
+
"eval_runtime": 8.4125,
|
| 1456 |
+
"eval_samples_per_second": 35.661,
|
| 1457 |
+
"eval_steps_per_second": 2.259,
|
| 1458 |
+
"step": 1457
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 15.67,
|
| 1462 |
+
"eval_exact_match": 10.0,
|
| 1463 |
+
"eval_f1": 14.556084656084652,
|
| 1464 |
+
"eval_qa_bleu": 8.766722232594686,
|
| 1465 |
+
"eval_qa_exact_match": 0.07,
|
| 1466 |
+
"eval_recite_bleu": 25.696114831592638,
|
| 1467 |
+
"eval_recite_exact_match": 0.08666666666666667,
|
| 1468 |
+
"step": 1457
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 15.7,
|
| 1472 |
+
"learning_rate": 3e-05,
|
| 1473 |
+
"loss": 0.0619,
|
| 1474 |
+
"step": 1460
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 15.81,
|
| 1478 |
+
"learning_rate": 3e-05,
|
| 1479 |
+
"loss": 0.063,
|
| 1480 |
+
"step": 1470
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 15.91,
|
| 1484 |
+
"learning_rate": 3e-05,
|
| 1485 |
+
"loss": 0.0625,
|
| 1486 |
+
"step": 1480
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 16.02,
|
| 1490 |
+
"learning_rate": 3e-05,
|
| 1491 |
+
"loss": 0.0614,
|
| 1492 |
+
"step": 1490
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 16.13,
|
| 1496 |
+
"learning_rate": 3e-05,
|
| 1497 |
+
"loss": 0.0536,
|
| 1498 |
+
"step": 1500
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 16.17,
|
| 1502 |
+
"eval_accuracy": 0.8413856209150327,
|
| 1503 |
+
"eval_loss": 0.3710184693336487,
|
| 1504 |
+
"eval_runtime": 8.2941,
|
| 1505 |
+
"eval_samples_per_second": 36.17,
|
| 1506 |
+
"eval_steps_per_second": 2.291,
|
| 1507 |
+
"step": 1504
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 16.17,
|
| 1511 |
+
"eval_exact_match": 12.0,
|
| 1512 |
+
"eval_f1": 16.22356902356902,
|
| 1513 |
+
"eval_qa_bleu": 10.03449398398276,
|
| 1514 |
+
"eval_qa_exact_match": 0.09666666666666666,
|
| 1515 |
+
"eval_recite_bleu": 27.21503335381207,
|
| 1516 |
+
"eval_recite_exact_match": 0.09,
|
| 1517 |
+
"step": 1504
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 16.24,
|
| 1521 |
+
"learning_rate": 3e-05,
|
| 1522 |
+
"loss": 0.0594,
|
| 1523 |
+
"step": 1510
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 16.34,
|
| 1527 |
+
"learning_rate": 3e-05,
|
| 1528 |
+
"loss": 0.0583,
|
| 1529 |
+
"step": 1520
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 16.45,
|
| 1533 |
+
"learning_rate": 3e-05,
|
| 1534 |
+
"loss": 0.0596,
|
| 1535 |
+
"step": 1530
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 16.56,
|
| 1539 |
+
"learning_rate": 3e-05,
|
| 1540 |
+
"loss": 0.0581,
|
| 1541 |
+
"step": 1540
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 16.67,
|
| 1545 |
+
"learning_rate": 3e-05,
|
| 1546 |
+
"loss": 0.0602,
|
| 1547 |
+
"step": 1550
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 16.68,
|
| 1551 |
+
"eval_accuracy": 0.8417777777777777,
|
| 1552 |
+
"eval_loss": 0.3608551621437073,
|
| 1553 |
+
"eval_runtime": 8.5041,
|
| 1554 |
+
"eval_samples_per_second": 35.277,
|
| 1555 |
+
"eval_steps_per_second": 2.234,
|
| 1556 |
+
"step": 1551
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 16.68,
|
| 1560 |
+
"eval_exact_match": 12.666666666666666,
|
| 1561 |
+
"eval_f1": 16.286243386243385,
|
| 1562 |
+
"eval_qa_bleu": 10.983250277461813,
|
| 1563 |
+
"eval_qa_exact_match": 0.08666666666666667,
|
| 1564 |
+
"eval_recite_bleu": 27.970963658310353,
|
| 1565 |
+
"eval_recite_exact_match": 0.1,
|
| 1566 |
+
"step": 1551
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 16.77,
|
| 1570 |
+
"learning_rate": 3e-05,
|
| 1571 |
+
"loss": 0.0595,
|
| 1572 |
+
"step": 1560
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 16.88,
|
| 1576 |
+
"learning_rate": 3e-05,
|
| 1577 |
+
"loss": 0.0607,
|
| 1578 |
+
"step": 1570
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 16.99,
|
| 1582 |
+
"learning_rate": 3e-05,
|
| 1583 |
+
"loss": 0.0609,
|
| 1584 |
+
"step": 1580
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 17.1,
|
| 1588 |
+
"learning_rate": 3e-05,
|
| 1589 |
+
"loss": 0.054,
|
| 1590 |
+
"step": 1590
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 17.18,
|
| 1594 |
+
"eval_accuracy": 0.8409934640522876,
|
| 1595 |
+
"eval_loss": 0.37587499618530273,
|
| 1596 |
+
"eval_runtime": 8.4497,
|
| 1597 |
+
"eval_samples_per_second": 35.504,
|
| 1598 |
+
"eval_steps_per_second": 2.249,
|
| 1599 |
+
"step": 1598
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 17.18,
|
| 1603 |
+
"eval_exact_match": 12.666666666666666,
|
| 1604 |
+
"eval_f1": 16.83068783068783,
|
| 1605 |
+
"eval_qa_bleu": 8.529481408255021,
|
| 1606 |
+
"eval_qa_exact_match": 0.1,
|
| 1607 |
+
"eval_recite_bleu": 27.164265779035382,
|
| 1608 |
+
"eval_recite_exact_match": 0.08333333333333333,
|
| 1609 |
+
"step": 1598
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 17.2,
|
| 1613 |
+
"learning_rate": 3e-05,
|
| 1614 |
+
"loss": 0.0564,
|
| 1615 |
+
"step": 1600
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 17.31,
|
| 1619 |
+
"learning_rate": 3e-05,
|
| 1620 |
+
"loss": 0.061,
|
| 1621 |
+
"step": 1610
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 17.42,
|
| 1625 |
+
"learning_rate": 3e-05,
|
| 1626 |
+
"loss": 0.058,
|
| 1627 |
+
"step": 1620
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 17.53,
|
| 1631 |
+
"learning_rate": 3e-05,
|
| 1632 |
+
"loss": 0.0597,
|
| 1633 |
+
"step": 1630
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 17.63,
|
| 1637 |
+
"learning_rate": 3e-05,
|
| 1638 |
+
"loss": 0.0635,
|
| 1639 |
+
"step": 1640
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 17.69,
|
| 1643 |
+
"eval_accuracy": 0.8414248366013072,
|
| 1644 |
+
"eval_loss": 0.3596791625022888,
|
| 1645 |
+
"eval_runtime": 8.3259,
|
| 1646 |
+
"eval_samples_per_second": 36.032,
|
| 1647 |
+
"eval_steps_per_second": 2.282,
|
| 1648 |
+
"step": 1645
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 17.69,
|
| 1652 |
+
"eval_exact_match": 10.666666666666666,
|
| 1653 |
+
"eval_f1": 14.222751322751321,
|
| 1654 |
+
"eval_qa_bleu": 7.525051258494909,
|
| 1655 |
+
"eval_qa_exact_match": 0.07333333333333333,
|
| 1656 |
+
"eval_recite_bleu": 27.177796043976347,
|
| 1657 |
+
"eval_recite_exact_match": 0.09333333333333334,
|
| 1658 |
+
"step": 1645
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 17.74,
|
| 1662 |
+
"learning_rate": 3e-05,
|
| 1663 |
+
"loss": 0.0596,
|
| 1664 |
+
"step": 1650
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 17.85,
|
| 1668 |
+
"learning_rate": 3e-05,
|
| 1669 |
+
"loss": 0.0595,
|
| 1670 |
+
"step": 1660
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 17.96,
|
| 1674 |
+
"learning_rate": 3e-05,
|
| 1675 |
+
"loss": 0.0617,
|
| 1676 |
+
"step": 1670
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 18.06,
|
| 1680 |
+
"learning_rate": 3e-05,
|
| 1681 |
+
"loss": 0.0558,
|
| 1682 |
+
"step": 1680
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 18.17,
|
| 1686 |
+
"learning_rate": 3e-05,
|
| 1687 |
+
"loss": 0.0536,
|
| 1688 |
+
"step": 1690
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 18.19,
|
| 1692 |
+
"eval_accuracy": 0.840967320261438,
|
| 1693 |
+
"eval_loss": 0.3750253915786743,
|
| 1694 |
+
"eval_runtime": 8.5676,
|
| 1695 |
+
"eval_samples_per_second": 35.015,
|
| 1696 |
+
"eval_steps_per_second": 2.218,
|
| 1697 |
+
"step": 1692
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 18.19,
|
| 1701 |
+
"eval_exact_match": 12.333333333333334,
|
| 1702 |
+
"eval_f1": 15.669312169312168,
|
| 1703 |
+
"eval_qa_bleu": 8.37631748679836,
|
| 1704 |
+
"eval_qa_exact_match": 0.07,
|
| 1705 |
+
"eval_recite_bleu": 28.231867454797584,
|
| 1706 |
+
"eval_recite_exact_match": 0.1,
|
| 1707 |
+
"step": 1692
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 18.28,
|
| 1711 |
+
"learning_rate": 3e-05,
|
| 1712 |
+
"loss": 0.0572,
|
| 1713 |
+
"step": 1700
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 18.39,
|
| 1717 |
+
"learning_rate": 3e-05,
|
| 1718 |
+
"loss": 0.0578,
|
| 1719 |
+
"step": 1710
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 18.49,
|
| 1723 |
+
"learning_rate": 3e-05,
|
| 1724 |
+
"loss": 0.0577,
|
| 1725 |
+
"step": 1720
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 18.6,
|
| 1729 |
+
"learning_rate": 3e-05,
|
| 1730 |
+
"loss": 0.0588,
|
| 1731 |
+
"step": 1730
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 18.7,
|
| 1735 |
+
"eval_accuracy": 0.841359477124183,
|
| 1736 |
+
"eval_loss": 0.36842381954193115,
|
| 1737 |
+
"eval_runtime": 8.7203,
|
| 1738 |
+
"eval_samples_per_second": 34.402,
|
| 1739 |
+
"eval_steps_per_second": 2.179,
|
| 1740 |
+
"step": 1739
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 18.7,
|
| 1744 |
+
"eval_exact_match": 13.666666666666666,
|
| 1745 |
+
"eval_f1": 16.964021164021162,
|
| 1746 |
+
"eval_qa_bleu": 10.412884310411028,
|
| 1747 |
+
"eval_qa_exact_match": 0.08666666666666667,
|
| 1748 |
+
"eval_recite_bleu": 27.91245954086591,
|
| 1749 |
+
"eval_recite_exact_match": 0.09333333333333334,
|
| 1750 |
+
"step": 1739
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 18.71,
|
| 1754 |
+
"learning_rate": 3e-05,
|
| 1755 |
+
"loss": 0.0579,
|
| 1756 |
+
"step": 1740
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 18.82,
|
| 1760 |
+
"learning_rate": 3e-05,
|
| 1761 |
+
"loss": 0.0623,
|
| 1762 |
+
"step": 1750
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 18.92,
|
| 1766 |
+
"learning_rate": 3e-05,
|
| 1767 |
+
"loss": 0.1972,
|
| 1768 |
+
"step": 1760
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 19.03,
|
| 1772 |
+
"learning_rate": 3e-05,
|
| 1773 |
+
"loss": 0.0903,
|
| 1774 |
+
"step": 1770
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 19.14,
|
| 1778 |
+
"learning_rate": 3e-05,
|
| 1779 |
+
"loss": 0.0713,
|
| 1780 |
+
"step": 1780
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 19.2,
|
| 1784 |
+
"eval_accuracy": 0.8410849673202614,
|
| 1785 |
+
"eval_loss": 0.36910614371299744,
|
| 1786 |
+
"eval_runtime": 8.785,
|
| 1787 |
+
"eval_samples_per_second": 34.149,
|
| 1788 |
+
"eval_steps_per_second": 2.163,
|
| 1789 |
+
"step": 1786
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 19.2,
|
| 1793 |
+
"eval_exact_match": 9.666666666666666,
|
| 1794 |
+
"eval_f1": 13.245502645502643,
|
| 1795 |
+
"eval_qa_bleu": 9.06198561813076,
|
| 1796 |
+
"eval_qa_exact_match": 0.06333333333333334,
|
| 1797 |
+
"eval_recite_bleu": 26.9668850069419,
|
| 1798 |
+
"eval_recite_exact_match": 0.1,
|
| 1799 |
+
"step": 1786
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 19.25,
|
| 1803 |
+
"learning_rate": 3e-05,
|
| 1804 |
+
"loss": 0.0723,
|
| 1805 |
+
"step": 1790
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 19.35,
|
| 1809 |
+
"learning_rate": 3e-05,
|
| 1810 |
+
"loss": 0.074,
|
| 1811 |
+
"step": 1800
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 19.46,
|
| 1815 |
+
"learning_rate": 3e-05,
|
| 1816 |
+
"loss": 0.0748,
|
| 1817 |
+
"step": 1810
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 19.57,
|
| 1821 |
+
"learning_rate": 3e-05,
|
| 1822 |
+
"loss": 0.0718,
|
| 1823 |
+
"step": 1820
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 19.68,
|
| 1827 |
+
"learning_rate": 3e-05,
|
| 1828 |
+
"loss": 0.0704,
|
| 1829 |
+
"step": 1830
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 19.71,
|
| 1833 |
+
"eval_accuracy": 0.8412679738562091,
|
| 1834 |
+
"eval_loss": 0.36380141973495483,
|
| 1835 |
+
"eval_runtime": 8.4756,
|
| 1836 |
+
"eval_samples_per_second": 35.396,
|
| 1837 |
+
"eval_steps_per_second": 2.242,
|
| 1838 |
+
"step": 1833
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 19.71,
|
| 1842 |
+
"eval_exact_match": 10.333333333333334,
|
| 1843 |
+
"eval_f1": 14.167724867724864,
|
| 1844 |
+
"eval_qa_bleu": 9.330356281157737,
|
| 1845 |
+
"eval_qa_exact_match": 0.07666666666666666,
|
| 1846 |
+
"eval_recite_bleu": 26.044375085111305,
|
| 1847 |
+
"eval_recite_exact_match": 0.09666666666666666,
|
| 1848 |
+
"step": 1833
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 19.78,
|
| 1852 |
+
"learning_rate": 3e-05,
|
| 1853 |
+
"loss": 0.0699,
|
| 1854 |
+
"step": 1840
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 19.89,
|
| 1858 |
+
"learning_rate": 3e-05,
|
| 1859 |
+
"loss": 0.0742,
|
| 1860 |
+
"step": 1850
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 20.0,
|
| 1864 |
+
"learning_rate": 3e-05,
|
| 1865 |
+
"loss": 0.0708,
|
| 1866 |
+
"step": 1860
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 20.0,
|
| 1870 |
+
"step": 1860,
|
| 1871 |
+
"total_flos": 4.95036515318784e+16,
|
| 1872 |
+
"train_loss": 0.03236631401764449,
|
| 1873 |
+
"train_runtime": 2828.6852,
|
| 1874 |
+
"train_samples_per_second": 10.485,
|
| 1875 |
+
"train_steps_per_second": 0.658
|
| 1876 |
}
|
| 1877 |
],
|
| 1878 |
"logging_steps": 10,
|
| 1879 |
+
"max_steps": 1860,
|
| 1880 |
+
"num_train_epochs": 20,
|
| 1881 |
"save_steps": 500,
|
| 1882 |
+
"total_flos": 4.95036515318784e+16,
|
| 1883 |
"trial_name": null,
|
| 1884 |
"trial_params": null
|
| 1885 |
}
|