Hugging Face's logo

stulcrad
/
CNEC1_1_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
159afc5
·
verified ·
1 Parent(s): 0cdbedc

Model save

Browse files
Files changed (1) hide show
  1. README.md +22 -29
README.md CHANGED
@@ -25,16 +25,16 @@ model-index:
25
  metrics:
26
  - name: Precision
27
  type: precision
28
- value: 0.8608470181503889
29
  - name: Recall
30
  type: recall
31
- value: 0.8794701986754967
32
  - name: F1
33
  type: f1
34
- value: 0.8700589648394845
35
  - name: Accuracy
36
  type: accuracy
37
- value: 0.9588976931183887
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: 0.3712
48
- - Precision: 0.8608
49
- - Recall: 0.8795
50
- - F1: 0.8701
51
- - Accuracy: 0.9589
52
 
53
  ## Model description
54
 
@@ -68,34 +68,27 @@ More information needed
68
 
69
  The following hyperparameters were used during training:
70
  - learning_rate: 2e-05
71
- - train_batch_size: 4
72
- - eval_batch_size: 4
73
  - seed: 42
74
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
75
  - lr_scheduler_type: linear
76
- - num_epochs: 15
77
 
78
  ### Training results
79
 
80
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
81
  |:-------------:|:-----:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
82
- | 0.3118 | 0.85 | 1000 | 0.2415 | 0.8255 | 0.8247 | 0.8251 | 0.9485 |
83
- | 0.219 | 1.7 | 2000 | 0.2222 | 0.8144 | 0.8406 | 0.8273 | 0.9505 |
84
- | 0.1421 | 2.56 | 3000 | 0.2214 | 0.8595 | 0.8450 | 0.8522 | 0.9509 |
85
- | 0.0972 | 3.41 | 4000 | 0.2378 | 0.8240 | 0.8534 | 0.8384 | 0.9529 |
86
- | 0.0861 | 4.26 | 5000 | 0.2462 | 0.8550 | 0.8640 | 0.8595 | 0.9574 |
87
- | 0.0735 | 5.11 | 6000 | 0.2688 | 0.8359 | 0.8636 | 0.8495 | 0.9563 |
88
- | 0.0556 | 5.96 | 7000 | 0.2874 | 0.8613 | 0.8720 | 0.8666 | 0.9573 |
89
- | 0.0337 | 6.81 | 8000 | 0.3092 | 0.8596 | 0.8675 | 0.8635 | 0.9572 |
90
- | 0.0316 | 7.67 | 9000 | 0.3166 | 0.8592 | 0.8786 | 0.8688 | 0.9578 |
91
- | 0.0207 | 8.52 | 10000 | 0.3391 | 0.8631 | 0.8799 | 0.8714 | 0.9590 |
92
- | 0.0154 | 9.37 | 11000 | 0.3660 | 0.8621 | 0.8693 | 0.8657 | 0.9568 |
93
- | 0.0096 | 10.22 | 12000 | 0.3851 | 0.8565 | 0.8746 | 0.8654 | 0.9570 |
94
- | 0.0077 | 11.07 | 13000 | 0.3553 | 0.8567 | 0.8737 | 0.8651 | 0.9576 |
95
- | 0.0094 | 11.93 | 14000 | 0.3742 | 0.8560 | 0.8684 | 0.8622 | 0.9573 |
96
- | 0.0064 | 12.78 | 15000 | 0.3656 | 0.8570 | 0.8755 | 0.8661 | 0.9582 |
97
- | 0.0032 | 13.63 | 16000 | 0.3607 | 0.8607 | 0.8812 | 0.8709 | 0.9594 |
98
- | 0.0026 | 14.48 | 17000 | 0.3712 | 0.8608 | 0.8795 | 0.8701 | 0.9589 |
99
 
100
 
101
  ### Framework versions
 
25
  metrics:
26
  - name: Precision
27
  type: precision
28
+ value: 0.8643410852713178
29
  - name: Recall
30
  type: recall
31
+ value: 0.8860927152317881
32
  - name: F1
33
  type: f1
34
+ value: 0.8750817527795944
35
  - name: Accuracy
36
  type: accuracy
37
+ value: 0.9606151684296811
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: 0.3309
48
+ - Precision: 0.8643
49
+ - Recall: 0.8861
50
+ - F1: 0.8751
51
+ - Accuracy: 0.9606
52
 
53
  ## Model description
54
 
 
68
 
69
  The following hyperparameters were used during training:
70
  - learning_rate: 2e-05
71
+ - train_batch_size: 2
72
+ - eval_batch_size: 2
73
  - seed: 42
74
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
75
  - lr_scheduler_type: linear
76
+ - num_epochs: 10
77
 
78
  ### Training results
79
 
80
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
81
  |:-------------:|:-----:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
82
+ | 0.3987 | 1.0 | 2348 | 0.2694 | 0.7643 | 0.8102 | 0.7865 | 0.9411 |
83
+ | 0.2937 | 2.0 | 4696 | 0.2530 | 0.8060 | 0.8252 | 0.8154 | 0.9491 |
84
+ | 0.2096 | 3.0 | 7044 | 0.2699 | 0.8285 | 0.8552 | 0.8416 | 0.9516 |
85
+ | 0.2114 | 4.0 | 9392 | 0.2632 | 0.8361 | 0.8693 | 0.8524 | 0.9572 |
86
+ | 0.1678 | 5.0 | 11740 | 0.2695 | 0.8344 | 0.8565 | 0.8453 | 0.9543 |
87
+ | 0.1352 | 6.0 | 14088 | 0.2680 | 0.8557 | 0.8821 | 0.8687 | 0.9578 |
88
+ | 0.0875 | 7.0 | 16436 | 0.2894 | 0.8532 | 0.8826 | 0.8676 | 0.9599 |
89
+ | 0.0735 | 8.0 | 18784 | 0.2816 | 0.8537 | 0.8834 | 0.8683 | 0.9606 |
90
+ | 0.0641 | 9.0 | 21132 | 0.3170 | 0.8627 | 0.8852 | 0.8738 | 0.9594 |
91
+ | 0.0629 | 10.0 | 23480 | 0.3309 | 0.8643 | 0.8861 | 0.8751 | 0.9606 |
 
 
 
 
 
 
 
92
 
93
 
94
  ### Framework versions