Hugging Face's logo

Salvatore
/
bert-finetuned-ner

Token Classification
Transformers
PyTorch
TensorBoard
bert
Generated from Trainer
Model card Files
xet
 Metrics Community
1
Salvatore commited on
Commit
c675621
·
1 Parent(s): ce66961

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +21 -21
README.md CHANGED
@@ -17,16 +17,16 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  # bert-finetuned-ner
19
 
20
- This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on an unknown dataset.
21
  It achieves the following results on the evaluation set:
22
- - Loss: 0.0359
23
- - Proteinmutation F1: 0.9212
24
- - Dnamutation F1: 0.7655
25
- - Snp F1: 1.0
26
- - Precision: 0.8351
27
- - Recall: 0.8728
28
- - F1: 0.8535
29
- - Accuracy: 0.9951
30
 
31
  ## Model description
32
 
@@ -55,18 +55,18 @@ The following hyperparameters were used during training:
55
 
56
  ### Training results
57
 
58
- | Training Loss | Epoch | Step | Validation Loss | Proteinmutation F1 | Dnamutation F1 | Snp F1 | Precision | Recall | F1 | Accuracy |
59
- |:-------------:|:-----:|:----:|:---------------:|:------------------:|:--------------:|:------:|:---------:|:------:|:------:|:--------:|
60
- | No log | 1.0 | 227 | 0.0265 | 0.8478 | 0.5493 | 1.0 | 0.6776 | 0.7565 | 0.7149 | 0.9931 |
61
- | No log | 2.0 | 454 | 0.0275 | 0.8738 | 0.6460 | 0.9744 | 0.7257 | 0.8211 | 0.7705 | 0.9934 |
62
- | 0.0683 | 3.0 | 681 | 0.0291 | 0.9257 | 0.7121 | 1.0 | 0.8107 | 0.8491 | 0.8295 | 0.9947 |
63
- | 0.0683 | 4.0 | 908 | 0.0325 | 0.9279 | 0.7392 | 1.0 | 0.8386 | 0.8513 | 0.8449 | 0.9947 |
64
- | 0.0044 | 5.0 | 1135 | 0.0346 | 0.9065 | 0.7572 | 1.0 | 0.8292 | 0.8578 | 0.8432 | 0.9946 |
65
- | 0.0044 | 6.0 | 1362 | 0.0358 | 0.8981 | 0.7566 | 1.0 | 0.8264 | 0.8513 | 0.8386 | 0.9949 |
66
- | 0.0009 | 7.0 | 1589 | 0.0413 | 0.8819 | 0.7439 | 1.0 | 0.8138 | 0.8384 | 0.8259 | 0.9942 |
67
- | 0.0009 | 8.0 | 1816 | 0.0359 | 0.9038 | 0.7506 | 1.0 | 0.8219 | 0.8556 | 0.8384 | 0.9948 |
68
- | 0.0005 | 9.0 | 2043 | 0.0355 | 0.9151 | 0.7522 | 1.0 | 0.8153 | 0.875 | 0.8441 | 0.9948 |
69
- | 0.0005 | 10.0 | 2270 | 0.0359 | 0.9212 | 0.7655 | 1.0 | 0.8351 | 0.8728 | 0.8535 | 0.9951 |
70
 
71
 
72
  ### Framework versions
 
17
 
18
  # bert-finetuned-ner
19
 
20
+ This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on the None dataset.
21
  It achieves the following results on the evaluation set:
22
+ - Loss: 0.0997
23
+ - Proteinmutation F1: 0.1309
24
+ - Snp F1: 0.1953
25
+ - Dnamutation F1: 0.3778
26
+ - Precision: 0.2380
27
+ - Recall: 0.2416
28
+ - F1: 0.2398
29
+ - Accuracy: 0.9703
30
 
31
  ## Model description
32
 
 
55
 
56
  ### Training results
57
 
58
+ | Training Loss | Epoch | Step | Validation Loss | Proteinmutation F1 | Snp F1 | Dnamutation F1 | Precision | Recall | F1 | Accuracy |
59
+ |:-------------:|:-----:|:----:|:---------------:|:------------------:|:------:|:--------------:|:---------:|:------:|:------:|:--------:|
60
+ | No log | 1.0 | 324 | 0.0533 | 0.0396 | 0.2830 | 0.4667 | 0.2334 | 0.3221 | 0.2707 | 0.9788 |
61
+ | 0.1072 | 2.0 | 648 | 0.0437 | 0.6065 | 0.4906 | 0.5009 | 0.4802 | 0.6348 | 0.5468 | 0.9868 |
62
+ | 0.1072 | 3.0 | 972 | 0.0592 | 0.1379 | 0.2485 | 0.2005 | 0.1639 | 0.2228 | 0.1889 | 0.9731 |
63
+ | 0.0573 | 4.0 | 1296 | 0.0722 | 0.0749 | 0.2530 | 0.4692 | 0.2705 | 0.2959 | 0.2826 | 0.9749 |
64
+ | 0.0431 | 5.0 | 1620 | 0.0766 | 0.1574 | 0.1847 | 0.2540 | 0.1766 | 0.2285 | 0.1992 | 0.9723 |
65
+ | 0.0431 | 6.0 | 1944 | 0.0805 | 0.1099 | 0.2202 | 0.2383 | 0.1657 | 0.2097 | 0.1851 | 0.9715 |
66
+ | 0.0396 | 7.0 | 2268 | 0.0886 | 0.1337 | 0.2138 | 0.4318 | 0.2683 | 0.2678 | 0.2680 | 0.9724 |
67
+ | 0.0354 | 8.0 | 2592 | 0.0927 | 0.1535 | 0.2113 | 0.3769 | 0.2505 | 0.2528 | 0.2516 | 0.9714 |
68
+ | 0.0354 | 9.0 | 2916 | 0.0978 | 0.1011 | 0.2540 | 0.3812 | 0.2495 | 0.2528 | 0.2512 | 0.9705 |
69
+ | 0.0312 | 10.0 | 3240 | 0.0997 | 0.1309 | 0.1953 | 0.3778 | 0.2380 | 0.2416 | 0.2398 | 0.9703 |
70
 
71
 
72
  ### Framework versions