Hugging Face's logo

TTian
/
deberta-classifier-feedback-1024-pseudo-final

Token Classification
Transformers
PyTorch
TensorBoard
deberta-v2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
TTian commited on
Commit
8106c11
·
1 Parent(s): e819850

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +105 -0
README.md ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ tags:
4
+ - generated_from_trainer
5
+ model-index:
6
+ - name: deberta-classifier-feedback-1024-pseudo-final
7
+ results: []
8
+ ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+ # deberta-classifier-feedback-1024-pseudo-final
14
+
15
+ This model is a fine-tuned version of [TTian/deberta-classifier-feedback-1024-pseudo](https://huggingface.co/TTian/deberta-classifier-feedback-1024-pseudo) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 0.5263
18
+
19
+ ## Model description
20
+
21
+ More information needed
22
+
23
+ ## Intended uses & limitations
24
+
25
+ More information needed
26
+
27
+ ## Training and evaluation data
28
+
29
+ More information needed
30
+
31
+ ## Training procedure
32
+
33
+ ### Training hyperparameters
34
+
35
+ The following hyperparameters were used during training:
36
+ - learning_rate: 2e-05
37
+ - train_batch_size: 8
38
+ - eval_batch_size: 8
39
+ - seed: 42
40
+ - gradient_accumulation_steps: 2
41
+ - total_train_batch_size: 16
42
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
+ - lr_scheduler_type: linear
44
+ - num_epochs: 2
45
+ - mixed_precision_training: Native AMP
46
+
47
+ ### Training results
48
+
49
+ | Training Loss | Epoch | Step | Validation Loss |
50
+ |:-------------:|:-----:|:----:|:---------------:|
51
+ | 0.5814 | 0.04 | 10 | 0.5888 |
52
+ | 0.5521 | 0.08 | 20 | 0.5736 |
53
+ | 0.5685 | 0.13 | 30 | 0.5809 |
54
+ | 0.6052 | 0.17 | 40 | 0.5702 |
55
+ | 0.5532 | 0.21 | 50 | 0.5571 |
56
+ | 0.6177 | 0.25 | 60 | 0.5848 |
57
+ | 0.6196 | 0.3 | 70 | 0.5464 |
58
+ | 0.5772 | 0.34 | 80 | 0.5307 |
59
+ | 0.5805 | 0.38 | 90 | 0.5550 |
60
+ | 0.6453 | 0.42 | 100 | 0.5467 |
61
+ | 0.5756 | 0.47 | 110 | 0.5587 |
62
+ | 0.5901 | 0.51 | 120 | 0.5482 |
63
+ | 0.568 | 0.55 | 130 | 0.5263 |
64
+ | 0.5452 | 0.59 | 140 | 0.5698 |
65
+ | 0.5949 | 0.64 | 150 | 0.5484 |
66
+ | 0.5537 | 0.68 | 160 | 0.5783 |
67
+ | 0.5327 | 0.72 | 170 | 0.5202 |
68
+ | 0.5449 | 0.76 | 180 | 0.5272 |
69
+ | 0.5345 | 0.81 | 190 | 0.5621 |
70
+ | 0.5837 | 0.85 | 200 | 0.5501 |
71
+ | 0.5969 | 0.89 | 210 | 0.5470 |
72
+ | 0.5905 | 0.93 | 220 | 0.5924 |
73
+ | 0.5481 | 0.97 | 230 | 0.5415 |
74
+ | 0.5035 | 1.02 | 240 | 0.5321 |
75
+ | 0.4508 | 1.06 | 250 | 0.5371 |
76
+ | 0.4227 | 1.1 | 260 | 0.5276 |
77
+ | 0.4423 | 1.14 | 270 | 0.5324 |
78
+ | 0.432 | 1.19 | 280 | 0.5378 |
79
+ | 0.4317 | 1.23 | 290 | 0.5302 |
80
+ | 0.46 | 1.27 | 300 | 0.5302 |
81
+ | 0.435 | 1.31 | 310 | 0.5326 |
82
+ | 0.3813 | 1.36 | 320 | 0.5431 |
83
+ | 0.4422 | 1.4 | 330 | 0.5323 |
84
+ | 0.4298 | 1.44 | 340 | 0.5575 |
85
+ | 0.5068 | 1.48 | 350 | 0.5529 |
86
+ | 0.4619 | 1.53 | 360 | 0.5589 |
87
+ | 0.4852 | 1.57 | 370 | 0.5256 |
88
+ | 0.3888 | 1.61 | 380 | 0.5731 |
89
+ | 0.4319 | 1.65 | 390 | 0.5335 |
90
+ | 0.4422 | 1.69 | 400 | 0.5419 |
91
+ | 0.4522 | 1.74 | 410 | 0.5547 |
92
+ | 0.4276 | 1.78 | 420 | 0.5263 |
93
+ | 0.3988 | 1.82 | 430 | 0.5481 |
94
+ | 0.4063 | 1.86 | 440 | 0.5404 |
95
+ | 0.4141 | 1.91 | 450 | 0.5292 |
96
+ | 0.4149 | 1.95 | 460 | 0.5241 |
97
+ | 0.4104 | 1.99 | 470 | 0.5263 |
98
+
99
+
100
+ ### Framework versions
101
+
102
+ - Transformers 4.24.0
103
+ - Pytorch 1.12.1+cu113
104
+ - Datasets 2.7.1
105
+ - Tokenizers 0.13.2