Hugging Face's logo

stulcrad
/
CNEC2_0_Supertypes_xlm-roberta-large

Token Classification
Transformers
TensorBoard
Safetensors
xlm-roberta
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
stulcrad commited on
Commit
d72c962
•
1 Parent(s): 1ce4989

Model save

Browse files
Files changed (2) hide show
  1. README.md +18 -26
  2. runs/Mar07_19-15-06_g01/events.out.tfevents.1709835307.g01.769784.6 +2 -2
README.md CHANGED
@@ -25,16 +25,16 @@ model-index:
25
  metrics:
26
  - name: Precision
27
  type: precision
28
- value: 0.8022359290670779
29
  - name: Recall
30
  type: recall
31
- value: 0.8549712407559573
32
  - name: F1
33
  type: f1
34
- value: 0.8277645186953062
35
  - name: Accuracy
36
  type: accuracy
37
- value: 0.9616810519608411
38
  ---
39
 
40
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -44,11 +44,11 @@ should probably proofread and complete it, then remove this comment. -->
44
 
45
  This model is a fine-tuned version of [FacebookAI/xlm-roberta-large](https://huggingface.co/FacebookAI/xlm-roberta-large) on the cnec dataset.
46
  It achieves the following results on the evaluation set:
47
- - Loss: 0.2033
48
- - Precision: 0.8022
49
- - Recall: 0.8550
50
- - F1: 0.8278
51
- - Accuracy: 0.9617
52
 
53
  ## Model description
54
 
@@ -68,33 +68,25 @@ More information needed
68
 
69
  The following hyperparameters were used during training:
70
  - learning_rate: 5e-05
71
- - train_batch_size: 8
72
- - eval_batch_size: 16
73
  - seed: 42
74
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
75
  - lr_scheduler_type: linear
76
  - lr_scheduler_warmup_ratio: 0.1
77
  - lr_scheduler_warmup_steps: 500
78
- - num_epochs: 8
79
 
80
  ### Training results
81
 
82
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
83
  |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
84
- | 0.6981 | 0.56 | 500 | 0.3042 | 0.5141 | 0.6652 | 0.5800 | 0.9121 |
85
- | 0.2782 | 1.11 | 1000 | 0.2128 | 0.7078 | 0.8159 | 0.7580 | 0.9495 |
86
- | 0.2247 | 1.67 | 1500 | 0.2200 | 0.7055 | 0.8081 | 0.7534 | 0.9450 |
87
- | 0.1986 | 2.22 | 2000 | 0.2291 | 0.6569 | 0.8110 | 0.7259 | 0.9460 |
88
- | 0.1697 | 2.78 | 2500 | 0.1819 | 0.7520 | 0.8184 | 0.7838 | 0.9548 |
89
- | 0.1415 | 3.33 | 3000 | 0.1873 | 0.7341 | 0.7975 | 0.7645 | 0.9527 |
90
- | 0.1284 | 3.89 | 3500 | 0.1752 | 0.7618 | 0.8578 | 0.8070 | 0.9590 |
91
- | 0.1073 | 4.44 | 4000 | 0.1903 | 0.7793 | 0.8488 | 0.8126 | 0.9586 |
92
- | 0.1006 | 5.0 | 4500 | 0.1741 | 0.7922 | 0.8661 | 0.8275 | 0.9610 |
93
- | 0.0788 | 5.56 | 5000 | 0.1830 | 0.7995 | 0.8537 | 0.8258 | 0.9623 |
94
- | 0.0838 | 6.11 | 5500 | 0.2096 | 0.8018 | 0.8509 | 0.8256 | 0.9610 |
95
- | 0.0617 | 6.67 | 6000 | 0.1978 | 0.8056 | 0.8632 | 0.8334 | 0.9627 |
96
- | 0.0515 | 7.22 | 6500 | 0.2020 | 0.8061 | 0.8521 | 0.8284 | 0.9616 |
97
- | 0.0455 | 7.78 | 7000 | 0.2033 | 0.8022 | 0.8550 | 0.8278 | 0.9617 |
98
 
99
 
100
  ### Framework versions
 
25
  metrics:
26
  - name: Precision
27
  type: precision
28
+ value: 0.8325581395348837
29
  - name: Recall
30
  type: recall
31
+ value: 0.8824979457682827
32
  - name: F1
33
  type: f1
34
+ value: 0.8568009573195053
35
  - name: Accuracy
36
  type: accuracy
37
+ value: 0.965938712854081
38
  ---
39
 
40
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
44
 
45
  This model is a fine-tuned version of [FacebookAI/xlm-roberta-large](https://huggingface.co/FacebookAI/xlm-roberta-large) on the cnec dataset.
46
  It achieves the following results on the evaluation set:
47
+ - Loss: 0.1992
48
+ - Precision: 0.8326
49
+ - Recall: 0.8825
50
+ - F1: 0.8568
51
+ - Accuracy: 0.9659
52
 
53
  ## Model description
54
 
 
68
 
69
  The following hyperparameters were used during training:
70
  - learning_rate: 5e-05
71
+ - train_batch_size: 32
72
+ - eval_batch_size: 32
73
  - seed: 42
74
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
75
  - lr_scheduler_type: linear
76
  - lr_scheduler_warmup_ratio: 0.1
77
  - lr_scheduler_warmup_steps: 500
78
+ - num_epochs: 15
79
 
80
  ### Training results
81
 
82
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
83
  |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
84
+ | 0.5321 | 2.22 | 500 | 0.1641 | 0.7159 | 0.8065 | 0.7585 | 0.9566 |
85
+ | 0.1512 | 4.44 | 1000 | 0.1831 | 0.7886 | 0.8611 | 0.8233 | 0.9591 |
86
+ | 0.0967 | 6.67 | 1500 | 0.1866 | 0.7628 | 0.8628 | 0.8097 | 0.9596 |
87
+ | 0.0637 | 8.89 | 2000 | 0.1586 | 0.8054 | 0.8841 | 0.8429 | 0.9648 |
88
+ | 0.0422 | 11.11 | 2500 | 0.1777 | 0.8294 | 0.8648 | 0.8467 | 0.9654 |
89
+ | 0.0292 | 13.33 | 3000 | 0.1992 | 0.8326 | 0.8825 | 0.8568 | 0.9659 |
 
 
 
 
 
 
 
 
90
 
91
 
92
  ### Framework versions
runs/Mar07_19-15-06_g01/events.out.tfevents.1709835307.g01.769784.6 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0b1918e326f5ad913815ef2a936e28d0de615e527a8bf102944f8f723e0cbbc8
3
- size 8757
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:995a8a9c3725ca25c80d0640df884eef6f077945004c462bb190a4d8a4f1b560
3
+ size 9111