Hugging Face's logo

nhidinh2
/
xlm-roberta-base-finetuned-ner-thesis-dseb

Token Classification
Transformers
TensorBoard
Safetensors
xlm-roberta
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
nhidinh2 commited on
Commit
9432a6c
•
1 Parent(s): a829bef

End of training

Browse files
Files changed (6) hide show
  1. README.md +20 -18
  2. config.json +4 -24
  3. model.safetensors +2 -2
  4. runs/Jul03_10-14-35_31ca19578870/events.out.tfevents.1720001694.31ca19578870.2037.0 +3 -0
  5. runs/Jul03_10-14-35_31ca19578870/events.out.tfevents.1720001954.31ca19578870.2037.1 +3 -0
  6. training_args.bin +1 -1
README.md CHANGED
@@ -20,11 +20,11 @@ should probably proofread and complete it, then remove this comment. -->
20
 
21
  This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.1471
24
- - Precision: 0.7995
25
- - Recall: 0.9088
26
- - F1: 0.8506
27
- - Accuracy: 0.9605
28
 
29
  ## Model description
30
 
@@ -44,27 +44,29 @@ More information needed
44
 
45
  The following hyperparameters were used during training:
46
  - learning_rate: 2e-05
47
- - train_batch_size: 32
48
  - eval_batch_size: 32
49
  - seed: 42
 
 
50
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
51
  - lr_scheduler_type: linear
52
  - num_epochs: 10
53
 
54
  ### Training results
55
 
56
- | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
57
- |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
58
- | 1.7775 | 1.0 | 31 | 0.3746 | 0.6199 | 0.6839 | 0.6503 | 0.8978 |
59
- | 0.1886 | 2.0 | 62 | 0.0734 | 0.9590 | 0.9301 | 0.9444 | 0.9875 |
60
- | 0.0821 | 3.0 | 93 | 0.0413 | 0.9697 | 0.9651 | 0.9674 | 0.9928 |
61
- | 0.0427 | 4.0 | 124 | 0.0400 | 0.9491 | 0.9635 | 0.9562 | 0.9911 |
62
- | 0.0352 | 5.0 | 155 | 0.0397 | 0.9421 | 0.9571 | 0.9496 | 0.9899 |
63
- | 0.0315 | 6.0 | 186 | 0.0410 | 0.9371 | 0.9579 | 0.9474 | 0.9895 |
64
- | 0.0344 | 7.0 | 217 | 0.0386 | 0.9612 | 0.9643 | 0.9627 | 0.9922 |
65
- | 0.0292 | 8.0 | 248 | 0.0383 | 0.9574 | 0.9651 | 0.9612 | 0.9921 |
66
- | 0.0286 | 9.0 | 279 | 0.0387 | 0.9543 | 0.9619 | 0.9581 | 0.9913 |
67
- | 0.0259 | 10.0 | 310 | 0.0415 | 0.9430 | 0.9595 | 0.9512 | 0.9901 |
68
 
69
 
70
  ### Framework versions
 
20
 
21
  This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.1313
24
+ - Precision: 0.0
25
+ - Recall: 0.0
26
+ - F1: 0.0
27
+ - Accuracy: 0.9315
28
 
29
  ## Model description
30
 
 
44
 
45
  The following hyperparameters were used during training:
46
  - learning_rate: 2e-05
47
+ - train_batch_size: 4
48
  - eval_batch_size: 32
49
  - seed: 42
50
+ - gradient_accumulation_steps: 2
51
+ - total_train_batch_size: 8
52
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
53
  - lr_scheduler_type: linear
54
  - num_epochs: 10
55
 
56
  ### Training results
57
 
58
+ | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
59
+ |:-------------:|:------:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
60
+ | No log | 0.9091 | 5 | 0.7631 | 0.0 | 0.0 | 0.0 | 0.9368 |
61
+ | No log | 2.0 | 11 | 0.2917 | 0.0 | 0.0 | 0.0 | 0.9430 |
62
+ | No log | 2.9091 | 16 | 0.2146 | 0.0 | 0.0 | 0.0 | 0.9430 |
63
+ | 0.5371 | 4.0 | 22 | 0.1465 | 0.0 | 0.0 | 0.0 | 0.9430 |
64
+ | 0.5371 | 4.9091 | 27 | 0.1146 | 0.0 | 0.0 | 0.0 | 0.9430 |
65
+ | 0.5371 | 6.0 | 33 | 0.0938 | 0.0 | 0.0 | 0.0 | 0.9430 |
66
+ | 0.5371 | 6.9091 | 38 | 0.0865 | 0.0 | 0.0 | 0.0 | 0.9430 |
67
+ | 0.1459 | 8.0 | 44 | 0.0824 | 0.0 | 0.0 | 0.0 | 0.9456 |
68
+ | 0.1459 | 8.9091 | 49 | 0.0797 | 0.0833 | 0.0714 | 0.0769 | 0.9514 |
69
+ | 0.1459 | 9.0909 | 50 | 0.0797 | 0.0833 | 0.0714 | 0.0769 | 0.9514 |
70
 
71
 
72
  ### Framework versions
config.json CHANGED
@@ -12,34 +12,14 @@
12
  "hidden_size": 768,
13
  "id2label": {
14
  "0": "O",
15
- "1": "B-NAME",
16
- "2": "I-NAME",
17
- "3": "B-DATE",
18
- "4": "I-DATE",
19
- "5": "B-UNI",
20
- "6": "I-UNI",
21
- "7": "B-MAJ",
22
- "8": "I-MAJ",
23
- "9": "B-MAIL",
24
- "10": "I-MAIL",
25
- "11": "B-UNI_ABRREV",
26
- "12": "I-PHONE"
27
  },
28
  "initializer_range": 0.02,
29
  "intermediate_size": 3072,
30
  "label2id": {
31
- "B-DATE": 3,
32
- "B-MAIL": 9,
33
- "B-MAJ": 7,
34
- "B-NAME": 1,
35
- "B-UNI": 5,
36
- "B-UNI_ABBREV": 11,
37
- "I-DATE": 4,
38
- "I-MAIL": 10,
39
- "I-MAJ": 8,
40
- "I-NAME": 2,
41
- "I-PHONE": 12,
42
- "I-UNI": 6,
43
  "O": 0
44
  },
45
  "layer_norm_eps": 1e-05,
 
12
  "hidden_size": 768,
13
  "id2label": {
14
  "0": "O",
15
+ "1": "B-HL",
16
+ "2": "I-HL"
 
 
 
 
 
 
 
 
 
 
17
  },
18
  "initializer_range": 0.02,
19
  "intermediate_size": 3072,
20
  "label2id": {
21
+ "B-HL": 1,
22
+ "I-HL": 2,
 
 
 
 
 
 
 
 
 
 
23
  "O": 0
24
  },
25
  "layer_norm_eps": 1e-05,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:13cecb2a6666b1ff8f1229bb2448e808e6c397b4542736c221a9a928e4522565
3
- size 1109876260
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d29f391ac21a39b563e2b5c53fe34320274d71ca105ba529db68ad9360ff6d9
3
+ size 1109845500
runs/Jul03_10-14-35_31ca19578870/events.out.tfevents.1720001694.31ca19578870.2037.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d125fd78128d70e32b3dafba1d93c84ecda49814e0c08632352fd34ea5e5466
3
+ size 30433
runs/Jul03_10-14-35_31ca19578870/events.out.tfevents.1720001954.31ca19578870.2037.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e32983657565700edcd28262c519a94f46eb511395d371eb2147a61dc806536
3
+ size 1014
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:527e9c3d048bd7bc6abe2e63ff7b8e37110134364244cee2292407ff334ca3b9
3
  size 5176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d919edb84eb0ae0199530f1de1ae262e407d2db1b5b7b24781cdc88eebeba8ee
3
  size 5176