vedantjumle commited on
Commit
a5762a1
1 Parent(s): 946b1e0

Training in progress epoch 0

Browse files
Files changed (3) hide show
  1. README.md +5 -34
  2. config.json +1 -1
  3. tf_model.h5 +1 -1
README.md CHANGED
@@ -15,10 +15,10 @@ probably proofread and complete it, then remove this comment. -->
15
 
16
  This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
- - Train Loss: 0.0645
19
- - Validation Loss: 0.5256
20
- - Train Accuracy: 0.89
21
- - Epoch: 29
22
 
23
  ## Model description
24
 
@@ -44,36 +44,7 @@ The following hyperparameters were used during training:
44
 
45
  | Train Loss | Validation Loss | Train Accuracy | Epoch |
46
  |:----------:|:---------------:|:--------------:|:-----:|
47
- | 5.0926 | 4.9378 | 0.0267 | 0 |
48
- | 4.8091 | 4.5518 | 0.19 | 1 |
49
- | 4.2565 | 3.9529 | 0.4833 | 2 |
50
- | 3.5496 | 3.3870 | 0.6733 | 3 |
51
- | 2.9320 | 2.8781 | 0.76 | 4 |
52
- | 2.3851 | 2.4109 | 0.8233 | 5 |
53
- | 1.8879 | 2.0233 | 0.8633 | 6 |
54
- | 1.4473 | 1.6912 | 0.8667 | 7 |
55
- | 1.1197 | 1.4349 | 0.8867 | 8 |
56
- | 0.8654 | 1.2351 | 0.8767 | 9 |
57
- | 0.6786 | 1.1051 | 0.8933 | 10 |
58
- | 0.5352 | 0.9739 | 0.8967 | 11 |
59
- | 0.4209 | 0.8814 | 0.8967 | 12 |
60
- | 0.3455 | 0.8079 | 0.8933 | 13 |
61
- | 0.2878 | 0.7635 | 0.8967 | 14 |
62
- | 0.2364 | 0.7142 | 0.9 | 15 |
63
- | 0.2115 | 0.6899 | 0.9 | 16 |
64
- | 0.1827 | 0.6554 | 0.9067 | 17 |
65
- | 0.1625 | 0.6388 | 0.8967 | 18 |
66
- | 0.1422 | 0.6158 | 0.9033 | 19 |
67
- | 0.1321 | 0.6181 | 0.9 | 20 |
68
- | 0.1187 | 0.5910 | 0.8933 | 21 |
69
- | 0.1072 | 0.5873 | 0.8967 | 22 |
70
- | 0.0988 | 0.5725 | 0.8967 | 23 |
71
- | 0.0902 | 0.5619 | 0.8967 | 24 |
72
- | 0.0844 | 0.5506 | 0.8967 | 25 |
73
- | 0.0776 | 0.5489 | 0.89 | 26 |
74
- | 0.0724 | 0.5411 | 0.8933 | 27 |
75
- | 0.0687 | 0.5280 | 0.8933 | 28 |
76
- | 0.0645 | 0.5256 | 0.89 | 29 |
77
 
78
 
79
  ### Framework versions
 
15
 
16
  This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Train Loss: 5.1229
19
+ - Validation Loss: 5.0343
20
+ - Train Accuracy: 0.0067
21
+ - Epoch: 0
22
 
23
  ## Model description
24
 
 
44
 
45
  | Train Loss | Validation Loss | Train Accuracy | Epoch |
46
  |:----------:|:---------------:|:--------------:|:-----:|
47
+ | 5.1229 | 5.0343 | 0.0067 | 0 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
48
 
49
 
50
  ### Framework versions
config.json CHANGED
@@ -7,7 +7,7 @@
7
  "classifier_dropout": 0.5,
8
  "gradient_checkpointing": false,
9
  "hidden_act": "gelu",
10
- "hidden_dropout_prob": 0.0,
11
  "hidden_size": 1024,
12
  "id2label": {
13
  "0": "LABEL_0",
 
7
  "classifier_dropout": 0.5,
8
  "gradient_checkpointing": false,
9
  "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.4,
11
  "hidden_size": 1024,
12
  "id2label": {
13
  "0": "LABEL_0",
tf_model.h5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4355bca940c806ec4db38fce331c7a04def5a7f20c48b8e88b34d1f4b1e93246
3
  size 1341734528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ceb1cb66d5c12908f681272d87fa414f0da25747c3d3938af4c0d693866738fe
3
  size 1341734528