Hugging Face's logo

Conrad747
/
luganda-ner-v2

Token Classification
Transformers
PyTorch
roberta
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
1
Conrad747 commited on
Commit
b8cf27e
•
1 Parent(s): 83d8bf6

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +19 -19
README.md CHANGED
@@ -24,16 +24,16 @@ model-index:
24
  metrics:
25
  - name: Precision
26
  type: precision
27
- value: 0.9360137288551116
28
  - name: Recall
29
  type: recall
30
- value: 0.936702649656526
31
  - name: F1
32
  type: f1
33
- value: 0.9363580625383201
34
  - name: Accuracy
35
  type: accuracy
36
- value: 0.982504001163975
37
  ---
38
 
39
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -43,11 +43,11 @@ should probably proofread and complete it, then remove this comment. -->
43
 
44
  This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the lg-ner dataset.
45
  It achieves the following results on the evaluation set:
46
- - Loss: 0.0948
47
- - Precision: 0.9360
48
- - Recall: 0.9367
49
- - F1: 0.9364
50
- - Accuracy: 0.9825
51
 
52
  ## Model description
53
 
@@ -78,16 +78,16 @@ The following hyperparameters were used during training:
78
 
79
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
80
  |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
81
- | 0.5678 | 1.0 | 609 | 0.2591 | 0.6885 | 0.7473 | 0.7167 | 0.9346 |
82
- | 0.235 | 2.0 | 1218 | 0.1651 | 0.7946 | 0.8675 | 0.8295 | 0.9572 |
83
- | 0.1756 | 3.0 | 1827 | 0.1307 | 0.8643 | 0.8810 | 0.8726 | 0.9660 |
84
- | 0.1312 | 4.0 | 2436 | 0.1074 | 0.8711 | 0.9134 | 0.8917 | 0.9718 |
85
- | 0.084 | 5.0 | 3045 | 0.0922 | 0.9141 | 0.9220 | 0.9180 | 0.9780 |
86
- | 0.0686 | 6.0 | 3654 | 0.0937 | 0.9212 | 0.9269 | 0.9241 | 0.9798 |
87
- | 0.0571 | 7.0 | 4263 | 0.0851 | 0.9258 | 0.9335 | 0.9296 | 0.9806 |
88
- | 0.0541 | 8.0 | 4872 | 0.0926 | 0.9338 | 0.9345 | 0.9342 | 0.9817 |
89
- | 0.0477 | 9.0 | 5481 | 0.0939 | 0.9364 | 0.9355 | 0.9359 | 0.9820 |
90
- | 0.0367 | 10.0 | 6090 | 0.0948 | 0.9360 | 0.9367 | 0.9364 | 0.9825 |
91
 
92
 
93
  ### Framework versions
 
24
  metrics:
25
  - name: Precision
26
  type: precision
27
+ value: 0.79182156133829
28
  - name: Recall
29
  type: recall
30
+ value: 0.7842415316642121
31
  - name: F1
32
  type: f1
33
+ value: 0.788013318534961
34
  - name: Accuracy
35
  type: accuracy
36
+ value: 0.9559346774929295
37
  ---
38
 
39
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
43
 
44
  This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the lg-ner dataset.
45
  It achieves the following results on the evaluation set:
46
+ - Loss: 0.3199
47
+ - Precision: 0.7918
48
+ - Recall: 0.7842
49
+ - F1: 0.7880
50
+ - Accuracy: 0.9559
51
 
52
  ## Model description
53
 
 
78
 
79
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
80
  |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
81
+ | No log | 1.0 | 261 | 0.2380 | 0.7942 | 0.7106 | 0.7501 | 0.9526 |
82
+ | 0.0954 | 2.0 | 522 | 0.2345 | 0.7954 | 0.7872 | 0.7913 | 0.9558 |
83
+ | 0.0954 | 3.0 | 783 | 0.2560 | 0.8168 | 0.7518 | 0.7830 | 0.9555 |
84
+ | 0.0562 | 4.0 | 1044 | 0.2815 | 0.7261 | 0.7791 | 0.7517 | 0.9477 |
85
+ | 0.0562 | 5.0 | 1305 | 0.2738 | 0.7744 | 0.8012 | 0.7875 | 0.9566 |
86
+ | 0.0345 | 6.0 | 1566 | 0.2951 | 0.8083 | 0.7732 | 0.7904 | 0.9556 |
87
+ | 0.0345 | 7.0 | 1827 | 0.3026 | 0.7741 | 0.7872 | 0.7806 | 0.9547 |
88
+ | 0.0215 | 8.0 | 2088 | 0.3062 | 0.8159 | 0.7636 | 0.7889 | 0.9563 |
89
+ | 0.0215 | 9.0 | 2349 | 0.3157 | 0.7959 | 0.7813 | 0.7886 | 0.9563 |
90
+ | 0.017 | 10.0 | 2610 | 0.3199 | 0.7918 | 0.7842 | 0.7880 | 0.9559 |
91
 
92
 
93
  ### Framework versions