Joe Davison commited on
Commit
a6dc0fd
1 Parent(s): db61cbd
README.md ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: en
3
+ tags:
4
+ - text-classification
5
+ - pytorch
6
+ - tensorflow
7
+ datasets:
8
+ - ag_news
9
+ license: mit
10
+ widget
11
+ - text: "Armed conflict has been a near-constant policial and economic burden."
12
+ - text: "Tom Brady won his seventh Super Bowl last night."
13
+ - text: "Dow falls more than 100 points after disappointing jobs data":
14
+ - text: "A new moon has been discovered in Jupter's orbit."
15
+ ---
16
+
17
+ # distilbert-base-uncased-agnews-student
18
+
19
+ ## Model Description
20
+
21
+ This model is distilled from the zero-shot classification pipeline on the unlabeled AG's News dataset using [this
22
+ script](https://github.com/huggingface/transformers/tree/master/examples/research_projects/zero-shot-distillation).
23
+ It is the result of the demo notebook
24
+ [here](https://colab.research.google.com/drive/1mjBjd0cR8G57ZpsnFCS3ngGyo5nCa9ya?usp=sharing), where more details
25
+ about the model can be found.
26
+
27
+ ## Intended Usage
28
+
29
+ The model can be used like any other model trained on AG's News, but will likely not perform as well as a model
30
+ trained with full supervision. It is primarily intended as a demo of how an expensive NLI-based zero-shot model
31
+ can be distilled to a more efficient student.
config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./distilbert-base-uncased-agnews-student/",
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": "the world",
13
+ "1": "sports",
14
+ "2": "business",
15
+ "3": "science/tech"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "label2id": {
19
+ "business": 2,
20
+ "science/tech": 3,
21
+ "sports": 1,
22
+ "the world": 0
23
+ },
24
+ "max_position_embeddings": 512,
25
+ "model_type": "distilbert",
26
+ "n_heads": 12,
27
+ "n_layers": 6,
28
+ "pad_token_id": 0,
29
+ "qa_dropout": 0.1,
30
+ "seq_classif_dropout": 0.2,
31
+ "sinusoidal_pos_embds": false,
32
+ "tie_weights_": true,
33
+ "transformers_version": "4.4.0.dev0",
34
+ "vocab_size": 30522
35
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ae45738d2e01161bb5de6af060e043ef7e0592fc9b8ffd54bc7aeb9848c3cd30
3
+ size 267866825
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tf_model.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd6486c59cf836b7b62d40f010d619bbc45ff2ee316b3ca353e6751e5056b47a
3
+ size 267958656
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "name_or_path": "distilbert-base-uncased"}
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6004c90490755211a561ac7fd2c86ce8840a4fa12734cd4d493df54e07cc4b19
3
+ size 2095
vocab.txt ADDED
The diff for this file is too large to render. See raw diff