|
'openai-community/gpt2' |
|
|
|
|
|
training_args = TrainingArguments( |
|
output_dir='ECO_GPT', |
|
evaluation_strategy="epoch", |
|
learning_rate=2e-5, |
|
per_device_train_batch_size=16, |
|
per_device_eval_batch_size=16, |
|
num_train_epochs=32, |
|
weight_decay=0.01, |
|
save_strategy="epoch", |
|
load_best_model_at_end=True, |
|
push_to_hub=True |
|
) |
|
Epoch Training Loss Validation Loss Precision Recall F1 Accuracy |
|
1 No log 0.197675 0.285714 0.212743 0.243887 0.959920 |
|
2 0.368900 0.159557 0.311457 0.302376 0.306849 0.961646 |
|
3 0.368900 0.146986 0.324180 0.344222 0.333901 0.961224 |
|
4 0.129100 0.145931 0.340494 0.372300 0.355687 0.960939 |
|
5 0.129100 0.144521 0.334234 0.367441 0.350051 0.961286 |
|
6 0.098700 0.141342 0.323340 0.393089 0.354819 0.962803 |
|
7 0.098700 0.144871 0.319322 0.381479 0.347644 0.962950 |
|
8 0.079700 0.147587 0.283885 0.377160 0.323942 0.961504 |
|
9 0.079700 0.152369 0.294302 0.355562 0.322044 0.961081 |
|
10 0.066600 0.154832 0.304663 0.373920 0.335758 0.962580 |
|
11 0.066600 0.160076 0.295521 0.381210 0.332940 0.960837 |
|
12 0.056600 0.164543 0.306661 0.370410 0.335534 0.962100 |
|
13 0.056600 0.166050 0.331369 0.379590 0.353844 0.963194 |
|
14 0.048700 0.169381 0.307859 0.383909 0.341704 0.961958 |
|
15 0.048700 0.179932 0.306340 0.380940 0.339591 0.960970 |
|
16 0.042500 0.180197 0.313375 0.368790 0.338832 0.962193 |
|
17 0.042500 0.186068 0.297326 0.384179 0.335218 0.960623 |
|
18 0.037500 0.195882 0.313199 0.377970 0.342550 0.961268 |
|
19 0.037500 0.197563 0.304962 0.384989 0.340334 0.960783 |
|
20 0.034100 0.201612 0.294698 0.370680 0.328351 0.960765 |
|
21 0.034100 0.207106 0.296033 0.382829 0.333883 0.959849 |
|
22 0.030900 0.205857 0.293838 0.383639 0.332787 0.960080 |
|
23 0.030900 0.211530 0.309654 0.383639 0.342699 0.961041 |
|
24 0.028700 0.215475 0.312196 0.381479 0.343378 0.961108 |
|
25 0.028700 0.213058 0.311651 0.377700 0.341511 0.961700 |
|
26 0.026700 0.216595 0.312333 0.378780 0.342362 0.961508 |
|
27 0.026700 0.223365 0.318739 0.379320 0.346400 0.961722 |
|
28 0.025400 0.224544 0.312472 0.373380 0.340221 0.961517 |
|
29 0.025400 0.225070 0.319539 0.381479 0.347773 0.961878 |
|
30 0.024300 0.223071 0.308169 0.375810 0.338645 0.961455 |
|
31 0.024300 0.225455 0.306649 0.372300 0.336300 0.961268 |
|
32 0.023900 0.224405 0.308620 0.373110 0.337815 0.961384 |
|
|
|
|
|
|
|
![image/png](https://cdn-uploads.huggingface.co/production/uploads/6634a6ddbcf56d1302dc1e82/LwJz_e7_EKlh4_c-RwS3f.png) |
|
|