Hugging Face's logo

Mingyi
/
classify_title_subject

Text Classification
Transformers
TensorFlow
bert
generated_from_keras_callback
Model card Files
xet
Community
1
Mingyi commited on
Commit
2bef3db
·
1 Parent(s): 51b2246

add model

Browse files
Files changed (3) hide show
  1. README.md +97 -11
  2. config.json +23 -16
  3. tf_model.h5 +2 -2
README.md CHANGED
@@ -1,15 +1,101 @@
1
- This model classifies the title of a content (e.g., YouTube video, article, or podcast episode) into 1 of 8 subjects
 
 
 
 
 
 
 
2
 
3
- 0. art
4
- 1. personal development
5
- 2. world
6
- 3. health
7
- 4. science
8
- 5. business
9
- 6. humanities
10
- 7. technology.
11
 
12
- This model is fine-tuned from **distilbert-base-uncased** using ~1k labeled data.
13
 
14
- This model is used to support [Sanderling](https://sanderling.app)
 
 
 
 
15
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_keras_callback
5
+ model-index:
6
+ - name: tmp6tsjsfbf
7
+ results: []
8
+ ---
9
 
10
+ <!-- This model card has been generated automatically according to the information Keras had access to. You should
11
+ probably proofread and complete it, then remove this comment. -->
 
 
 
 
 
 
12
 
13
+ # tmp6tsjsfbf
14
 
15
+ This model is a fine-tuned version of [bert-base-multilingual-cased](https://huggingface.co/bert-base-multilingual-cased) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Train Loss: 0.0178
18
+ - Train Sparse Categorical Accuracy: 0.9962
19
+ - Epoch: 49
20
 
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - optimizer: {'name': 'Adam', 'learning_rate': 5e-06, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
39
+ - training_precision: float32
40
+
41
+ ### Training results
42
+
43
+ | Train Loss | Train Sparse Categorical Accuracy | Epoch |
44
+ |:----------:|:---------------------------------:|:-----:|
45
+ | 1.8005 | 0.3956 | 0 |
46
+ | 1.3302 | 0.5916 | 1 |
47
+ | 0.8998 | 0.7575 | 2 |
48
+ | 0.6268 | 0.8468 | 3 |
49
+ | 0.4239 | 0.9062 | 4 |
50
+ | 0.2982 | 0.9414 | 5 |
51
+ | 0.2245 | 0.9625 | 6 |
52
+ | 0.1678 | 0.9730 | 7 |
53
+ | 0.1399 | 0.9745 | 8 |
54
+ | 0.1059 | 0.9827 | 9 |
55
+ | 0.0822 | 0.9850 | 10 |
56
+ | 0.0601 | 0.9902 | 11 |
57
+ | 0.0481 | 0.9932 | 12 |
58
+ | 0.0386 | 0.9955 | 13 |
59
+ | 0.0292 | 0.9977 | 14 |
60
+ | 0.0353 | 0.9940 | 15 |
61
+ | 0.0336 | 0.9932 | 16 |
62
+ | 0.0345 | 0.9910 | 17 |
63
+ | 0.0179 | 0.9985 | 18 |
64
+ | 0.0150 | 0.9985 | 19 |
65
+ | 0.0365 | 0.9895 | 20 |
66
+ | 0.0431 | 0.9895 | 21 |
67
+ | 0.0243 | 0.9955 | 22 |
68
+ | 0.0317 | 0.9925 | 23 |
69
+ | 0.0375 | 0.9902 | 24 |
70
+ | 0.0138 | 0.9970 | 25 |
71
+ | 0.0159 | 0.9977 | 26 |
72
+ | 0.0160 | 0.9962 | 27 |
73
+ | 0.0151 | 0.9977 | 28 |
74
+ | 0.0337 | 0.9902 | 29 |
75
+ | 0.0119 | 0.9977 | 30 |
76
+ | 0.0165 | 0.9955 | 31 |
77
+ | 0.0133 | 0.9977 | 32 |
78
+ | 0.0047 | 1.0 | 33 |
79
+ | 0.0037 | 1.0 | 34 |
80
+ | 0.0033 | 1.0 | 35 |
81
+ | 0.0031 | 1.0 | 36 |
82
+ | 0.0036 | 1.0 | 37 |
83
+ | 0.0343 | 0.9887 | 38 |
84
+ | 0.0234 | 0.9962 | 39 |
85
+ | 0.0034 | 1.0 | 40 |
86
+ | 0.0036 | 1.0 | 41 |
87
+ | 0.0261 | 0.9917 | 42 |
88
+ | 0.0111 | 0.9970 | 43 |
89
+ | 0.0039 | 1.0 | 44 |
90
+ | 0.0214 | 0.9932 | 45 |
91
+ | 0.0044 | 0.9985 | 46 |
92
+ | 0.0122 | 0.9985 | 47 |
93
+ | 0.0119 | 0.9962 | 48 |
94
+ | 0.0178 | 0.9962 | 49 |
95
+
96
+
97
+ ### Framework versions
98
+
99
+ - Transformers 4.15.0
100
+ - TensorFlow 2.7.0
101
+ - Tokenizers 0.10.3
config.json CHANGED
@@ -1,13 +1,14 @@
1
  {
2
- "_name_or_path": "distilbert-base-uncased",
3
- "activation": "gelu",
4
  "architectures": [
5
- "DistilBertForSequenceClassification"
6
  ],
7
- "attention_dropout": 0.1,
8
- "dim": 768,
9
- "dropout": 0.1,
10
- "hidden_dim": 3072,
 
 
11
  "id2label": {
12
  "0": "LABEL_0",
13
  "1": "LABEL_1",
@@ -19,6 +20,7 @@
19
  "7": "LABEL_7"
20
  },
21
  "initializer_range": 0.02,
 
22
  "label2id": {
23
  "LABEL_0": 0,
24
  "LABEL_1": 1,
@@ -29,15 +31,20 @@
29
  "LABEL_6": 6,
30
  "LABEL_7": 7
31
  },
 
32
  "max_position_embeddings": 512,
33
- "model_type": "distilbert",
34
- "n_heads": 12,
35
- "n_layers": 6,
36
  "pad_token_id": 0,
37
- "qa_dropout": 0.1,
38
- "seq_classif_dropout": 0.2,
39
- "sinusoidal_pos_embds": false,
40
- "tie_weights_": true,
41
- "transformers_version": "4.12.3",
42
- "vocab_size": 30522
 
 
 
 
43
  }
 
1
  {
2
+ "_name_or_path": "bert-base-multilingual-cased",
 
3
  "architectures": [
4
+ "BertForSequenceClassification"
5
  ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
  "id2label": {
13
  "0": "LABEL_0",
14
  "1": "LABEL_1",
 
20
  "7": "LABEL_7"
21
  },
22
  "initializer_range": 0.02,
23
+ "intermediate_size": 3072,
24
  "label2id": {
25
  "LABEL_0": 0,
26
  "LABEL_1": 1,
 
31
  "LABEL_6": 6,
32
  "LABEL_7": 7
33
  },
34
+ "layer_norm_eps": 1e-12,
35
  "max_position_embeddings": 512,
36
+ "model_type": "bert",
37
+ "num_attention_heads": 12,
38
+ "num_hidden_layers": 12,
39
  "pad_token_id": 0,
40
+ "pooler_fc_size": 768,
41
+ "pooler_num_attention_heads": 12,
42
+ "pooler_num_fc_layers": 3,
43
+ "pooler_size_per_head": 128,
44
+ "pooler_type": "first_token_transform",
45
+ "position_embedding_type": "absolute",
46
+ "transformers_version": "4.15.0",
47
+ "type_vocab_size": 2,
48
+ "use_cache": true,
49
+ "vocab_size": 119547
50
  }
tf_model.h5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bcb40642f6ddb7b5580aaa67d1db1e691b698c3fc783b36482a3ecb2d5514e02
3
- size 267970328
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2907ce89ca81a850bb2c8e29b108aebbbb83d6bd55afa85bde8359d4673c576
3
+ size 711726448