tingtone commited on
Commit
ed974ef
1 Parent(s): 959fe36

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -11
README.md CHANGED
@@ -16,12 +16,12 @@ model-index:
16
  name: emotion
17
  type: emotion
18
  config: split
19
- split: validation[:1000]
20
  args: split
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
- value: 0.575
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 1.2677
35
- - Accuracy: 0.575
36
 
37
  ## Model description
38
 
@@ -51,25 +51,34 @@ More information needed
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
54
- - learning_rate: 2e-05
55
  - train_batch_size: 16
56
  - eval_batch_size: 16
57
  - seed: 42
58
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
  - lr_scheduler_type: linear
60
- - num_epochs: 2
 
61
 
62
  ### Training results
63
 
64
- | Training Loss | Epoch | Step | Validation Loss | Accuracy |
65
- |:-------------:|:-----:|:----:|:---------------:|:--------:|
66
- | No log | 1.0 | 63 | 1.4519 | 0.54 |
67
- | No log | 2.0 | 126 | 1.2677 | 0.575 |
 
 
 
 
 
 
 
 
68
 
69
 
70
  ### Framework versions
71
 
72
  - Transformers 4.26.1
73
- - Pytorch 2.1.0.dev20230322
74
  - Datasets 2.12.0
75
  - Tokenizers 0.13.3
 
16
  name: emotion
17
  type: emotion
18
  config: split
19
+ split: validation
20
  args: split
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
+ value: 0.939
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
31
 
32
  This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 0.2100
35
+ - Accuracy: 0.939
36
 
37
  ## Model description
38
 
 
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
54
+ - learning_rate: 1e-05
55
  - train_batch_size: 16
56
  - eval_batch_size: 16
57
  - seed: 42
58
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
  - lr_scheduler_type: linear
60
+ - lr_scheduler_warmup_steps: 16000
61
+ - num_epochs: 10
62
 
63
  ### Training results
64
 
65
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
66
+ |:-------------:|:-----:|:-----:|:---------------:|:--------:|
67
+ | 1.6731 | 1.0 | 1000 | 1.5737 | 0.4805 |
68
+ | 1.224 | 2.0 | 2000 | 1.1049 | 0.5915 |
69
+ | 0.8266 | 3.0 | 3000 | 0.7033 | 0.761 |
70
+ | 0.4635 | 4.0 | 4000 | 0.3884 | 0.8845 |
71
+ | 0.2832 | 5.0 | 5000 | 0.2466 | 0.9145 |
72
+ | 0.1855 | 6.0 | 6000 | 0.2112 | 0.926 |
73
+ | 0.1552 | 7.0 | 7000 | 0.1978 | 0.9285 |
74
+ | 0.1426 | 8.0 | 8000 | 0.1937 | 0.9345 |
75
+ | 0.1308 | 9.0 | 9000 | 0.2093 | 0.932 |
76
+ | 0.1127 | 10.0 | 10000 | 0.2100 | 0.939 |
77
 
78
 
79
  ### Framework versions
80
 
81
  - Transformers 4.26.1
82
+ - Pytorch 1.13.1+cu117
83
  - Datasets 2.12.0
84
  - Tokenizers 0.13.3