Initial commit
Browse files- config.json +42 -0
- pair_classification.py +25 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +14 -0
- vocab.txt +0 -0
config.json
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "sgugger/finetuned-bert-mrpc",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"custom_pipelines": {
|
9 |
+
"pair-classification": {
|
10 |
+
"impl": "pair_classification.PairClassificationPipeline",
|
11 |
+
"pt": "AutoModelForSequenceClassification",
|
12 |
+
"tf": "TFAutoModelForSequenceClassification"
|
13 |
+
}
|
14 |
+
},
|
15 |
+
"gradient_checkpointing": false,
|
16 |
+
"hidden_act": "gelu",
|
17 |
+
"hidden_dropout_prob": 0.1,
|
18 |
+
"hidden_size": 768,
|
19 |
+
"id2label": {
|
20 |
+
"0": "not equivalent",
|
21 |
+
"1": "equivalent"
|
22 |
+
},
|
23 |
+
"initializer_range": 0.02,
|
24 |
+
"intermediate_size": 3072,
|
25 |
+
"label2id": {
|
26 |
+
"equivalent": 1,
|
27 |
+
"not equivalent": 0
|
28 |
+
},
|
29 |
+
"layer_norm_eps": 1e-12,
|
30 |
+
"max_position_embeddings": 512,
|
31 |
+
"model_type": "bert",
|
32 |
+
"num_attention_heads": 12,
|
33 |
+
"num_hidden_layers": 12,
|
34 |
+
"pad_token_id": 0,
|
35 |
+
"position_embedding_type": "absolute",
|
36 |
+
"problem_type": "single_label_classification",
|
37 |
+
"torch_dtype": "float32",
|
38 |
+
"transformers_version": "4.21.0.dev0",
|
39 |
+
"type_vocab_size": 2,
|
40 |
+
"use_cache": true,
|
41 |
+
"vocab_size": 28996
|
42 |
+
}
|
pair_classification.py
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import Pipeline
|
2 |
+
import torch
|
3 |
+
|
4 |
+
class PairClassificationPipeline(Pipeline):
|
5 |
+
def _sanitize_parameters(self, **kwargs):
|
6 |
+
preprocess_kwargs = {}
|
7 |
+
if "text_pair" in kwargs:
|
8 |
+
preprocess_kwargs["text_pair"] = kwargs["text_pair"]
|
9 |
+
return preprocess_kwargs, {}, {}
|
10 |
+
|
11 |
+
def preprocess(self, text, text_pair=None):
|
12 |
+
return self.tokenizer(text, text_pair=text_pair, return_tensors="pt")
|
13 |
+
|
14 |
+
def _forward(self, model_inputs):
|
15 |
+
return self.model(**model_inputs)
|
16 |
+
|
17 |
+
def postprocess(self, model_outputs):
|
18 |
+
logits = model_outputs.logits
|
19 |
+
probabilities = torch.nn.functional.softmax(logits, dim=-1)
|
20 |
+
|
21 |
+
best_class = probabilities.argmax().item()
|
22 |
+
label = self.model.config.id2label[best_class]
|
23 |
+
score = probabilities.squeeze()[best_class].item()
|
24 |
+
logits = logits.squeeze().tolist()
|
25 |
+
return {"label": label, "score": score, "logits": logits}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fb4ff71c41b993ad8531e00ca325ef3c7938156427dccae355771c026a75d4ad
|
3 |
+
size 433315437
|
special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
tokenizer_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"do_lower_case": false,
|
4 |
+
"mask_token": "[MASK]",
|
5 |
+
"model_max_length": 512,
|
6 |
+
"name_or_path": "sgugger/finetuned-bert-mrpc",
|
7 |
+
"pad_token": "[PAD]",
|
8 |
+
"sep_token": "[SEP]",
|
9 |
+
"special_tokens_map_file": null,
|
10 |
+
"strip_accents": null,
|
11 |
+
"tokenize_chinese_chars": true,
|
12 |
+
"tokenizer_class": "BertTokenizer",
|
13 |
+
"unk_token": "[UNK]"
|
14 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|