KoichiYasuoka commited on
Commit
2c30a68
1 Parent(s): 9e3a804

initial release

Browse files
README.md ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - "japanese"
4
+ - "question-answering"
5
+ - "dependency-parsing"
6
+ datasets:
7
+ - "universal_dependencies"
8
+ license: "cc-by-sa-4.0"
9
+ pipeline_tag: "question-answering"
10
+ widget:
11
+ - text: "国語"
12
+ context: "全学年にわたって小学校の国語の教科書に挿し絵が用いられている"
13
+ - text: "教科書"
14
+ context: "全学年にわたって小学校の国語の教科書に挿し絵が用いられている"
15
+ - text: "の"
16
+ context: "全学年にわたって小学校の国語[MASK]教科書に挿し絵が用いられている"
17
+ ---
18
+
19
+ # deberta-base-japanese-aozora-ud-head
20
+
21
+ ## Model Description
22
+
23
+ This is a DeBERTa(V2) model pretrained on 青空文庫 for dependency-parsing (head-detection on long-unit-words) as question-answering, derived from [deberta-base-japanese-aozora](https://huggingface.co/KoichiYasuoka/deberta-base-japanese-aozora) and [UD_Japanese-GSDLUW](https://huggingface.co/UniversalDependencies/UD_Japanese-GSDLUW). Use [MASK] inside `context` to avoid ambiguity when specifing a multiple-used word as `question`.
24
+
25
+ ## How to Use
26
+
27
+ ```py
28
+ import torch
29
+ from transformers import AutoTokenizer,AutoModelForQuestionAnswering
30
+ tokenizer=AutoTokenizer.from_pretrained("KoichiYasuoka/deberta-base-japanese-aozora-ud-head")
31
+ model=AutoModelForQuestionAnswering.from_pretrained("KoichiYasuoka/deberta-base-japanese-aozora-ud-head")
32
+ question="国語"
33
+ context="全学年にわたって小学校の国語の教科書に挿し絵が用いられている"
34
+ inputs=tokenizer(question,context,return_tensors="pt",return_offsets_mapping=True)
35
+ offsets=inputs.pop("offset_mapping").tolist()[0]
36
+ outputs=model(**inputs)
37
+ start,end=torch.argmax(outputs.start_logits),torch.argmax(outputs.end_logits)
38
+ print(context[offsets[start][0]:offsets[end][-1]])
39
+ ```
40
+
config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForQuestionAnswering"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "eos_token_id": 2,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 3072,
13
+ "layer_norm_eps": 1e-07,
14
+ "max_position_embeddings": 512,
15
+ "max_relative_positions": -1,
16
+ "model_type": "deberta-v2",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "pad_token_id": 1,
20
+ "pooler_dropout": 0,
21
+ "pooler_hidden_act": "gelu",
22
+ "pooler_hidden_size": 768,
23
+ "pos_att_type": null,
24
+ "position_biased_input": true,
25
+ "relative_attention": false,
26
+ "tokenizer_class": "DebertaV2TokenizerFast",
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.19.4",
29
+ "type_vocab_size": 0,
30
+ "vocab_size": 32000
31
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7509e3bc46e732e2b4cada2976207b7f735ace7e6be5262a7424c0cc010e2415
3
+ size 440178035
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "[CLS]", "eos_token": "[SEP]", "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b
3
+ size 1
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": false, "bos_token": "[CLS]", "eos_token": "[SEP]", "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "split_by_punct": true, "keep_accents": true, "model_max_length": 512, "tokenizer_class": "DebertaV2TokenizerFast"}