add word segmentation before tokenization
Browse files- README.md +26 -25
- custom_tokenizer.py +11 -0
- pipeline.py +76 -0
- requirements.txt +1 -0
README.md
CHANGED
|
@@ -1,21 +1,22 @@
|
|
| 1 |
---
|
| 2 |
pipeline_tag: sentence-similarity
|
| 3 |
tags:
|
| 4 |
-
- sentence-transformers
|
| 5 |
-
- feature-extraction
|
| 6 |
-
- sentence-similarity
|
| 7 |
-
- transformers
|
|
|
|
| 8 |
language:
|
| 9 |
-
- vi
|
| 10 |
-
- en
|
| 11 |
widget:
|
| 12 |
-
- source_sentence:
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
---
|
| 20 |
|
| 21 |
# bkai-foundation-models/vietnamese-bi-encoder
|
|
@@ -43,9 +44,8 @@ embeddings = model.encode(sentences)
|
|
| 43 |
print(embeddings)
|
| 44 |
```
|
| 45 |
|
| 46 |
-
|
| 47 |
-
|
| 48 |
## Usage (HuggingFace Transformers)
|
|
|
|
| 49 |
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
|
| 50 |
|
| 51 |
```python
|
|
@@ -81,21 +81,20 @@ print("Sentence embeddings:")
|
|
| 81 |
print(sentence_embeddings)
|
| 82 |
```
|
| 83 |
|
| 84 |
-
|
| 85 |
-
|
| 86 |
## Evaluation Results
|
| 87 |
|
| 88 |
<!--- Describe how your model was evaluated -->
|
| 89 |
|
| 90 |
-
For an automated evaluation of this model, see the
|
| 91 |
-
|
| 92 |
|
| 93 |
## Training
|
|
|
|
| 94 |
The model was trained with the parameters:
|
| 95 |
|
| 96 |
**DataLoader**:
|
| 97 |
|
| 98 |
`torch.utils.data.dataloader.DataLoader` of length 17584 with parameters:
|
|
|
|
| 99 |
```
|
| 100 |
{'batch_size': 32, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
|
| 101 |
```
|
|
@@ -103,11 +102,13 @@ The model was trained with the parameters:
|
|
| 103 |
**Loss**:
|
| 104 |
|
| 105 |
`sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss` with parameters:
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
|
|
|
| 109 |
|
| 110 |
Parameters of the fit()-Method:
|
|
|
|
| 111 |
```
|
| 112 |
{
|
| 113 |
"epochs": 15,
|
|
@@ -125,15 +126,15 @@ Parameters of the fit()-Method:
|
|
| 125 |
}
|
| 126 |
```
|
| 127 |
|
| 128 |
-
|
| 129 |
## Full Model Architecture
|
|
|
|
| 130 |
```
|
| 131 |
SentenceTransformer(
|
| 132 |
-
(0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: RobertaModel
|
| 133 |
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False})
|
| 134 |
)
|
| 135 |
```
|
| 136 |
|
| 137 |
## Citing & Authors
|
| 138 |
|
| 139 |
-
<!--- Describe where people can find more information -->
|
|
|
|
| 1 |
---
|
| 2 |
pipeline_tag: sentence-similarity
|
| 3 |
tags:
|
| 4 |
+
- sentence-transformers
|
| 5 |
+
- feature-extraction
|
| 6 |
+
- sentence-similarity
|
| 7 |
+
- transformers
|
| 8 |
+
library_name: generic
|
| 9 |
language:
|
| 10 |
+
- vi
|
| 11 |
+
- en
|
| 12 |
widget:
|
| 13 |
+
- source_sentence: 'Anh ấy đang là sinh viên năm cuối'
|
| 14 |
+
sentences:
|
| 15 |
+
- 'Anh ấy học tại Đại học Bách khoa Hà Nội, chuyên ngành Khoa học máy tính'
|
| 16 |
+
- 'Anh ấy đang làm việc tại nhà máy sản xuất linh kiện điện tử'
|
| 17 |
+
- 'Anh ấy chuẩn bị đi du học nước ngoài'
|
| 18 |
+
- 'Anh ấy sắp mở cửa hàng bán mỹ phẩm'
|
| 19 |
+
- 'Nhà anh ấy có rất nhiều cây cảnh'
|
| 20 |
---
|
| 21 |
|
| 22 |
# bkai-foundation-models/vietnamese-bi-encoder
|
|
|
|
| 44 |
print(embeddings)
|
| 45 |
```
|
| 46 |
|
|
|
|
|
|
|
| 47 |
## Usage (HuggingFace Transformers)
|
| 48 |
+
|
| 49 |
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
|
| 50 |
|
| 51 |
```python
|
|
|
|
| 81 |
print(sentence_embeddings)
|
| 82 |
```
|
| 83 |
|
|
|
|
|
|
|
| 84 |
## Evaluation Results
|
| 85 |
|
| 86 |
<!--- Describe how your model was evaluated -->
|
| 87 |
|
| 88 |
+
For an automated evaluation of this model, see the _Sentence Embeddings Benchmark_: [https://seb.sbert.net](https://seb.sbert.net?model_name=bkai-foundation-models/vietnamese-bi-encoder)
|
|
|
|
| 89 |
|
| 90 |
## Training
|
| 91 |
+
|
| 92 |
The model was trained with the parameters:
|
| 93 |
|
| 94 |
**DataLoader**:
|
| 95 |
|
| 96 |
`torch.utils.data.dataloader.DataLoader` of length 17584 with parameters:
|
| 97 |
+
|
| 98 |
```
|
| 99 |
{'batch_size': 32, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
|
| 100 |
```
|
|
|
|
| 102 |
**Loss**:
|
| 103 |
|
| 104 |
`sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss` with parameters:
|
| 105 |
+
|
| 106 |
+
```
|
| 107 |
+
{'scale': 20.0, 'similarity_fct': 'cos_sim'}
|
| 108 |
+
```
|
| 109 |
|
| 110 |
Parameters of the fit()-Method:
|
| 111 |
+
|
| 112 |
```
|
| 113 |
{
|
| 114 |
"epochs": 15,
|
|
|
|
| 126 |
}
|
| 127 |
```
|
| 128 |
|
|
|
|
| 129 |
## Full Model Architecture
|
| 130 |
+
|
| 131 |
```
|
| 132 |
SentenceTransformer(
|
| 133 |
+
(0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: RobertaModel
|
| 134 |
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False})
|
| 135 |
)
|
| 136 |
```
|
| 137 |
|
| 138 |
## Citing & Authors
|
| 139 |
|
| 140 |
+
<!--- Describe where people can find more information -->
|
custom_tokenizer.py
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from transformers import PhobertTokenizer
|
| 2 |
+
from pyvi import ViTokenizer
|
| 3 |
+
|
| 4 |
+
|
| 5 |
+
class CustomPhobertTokenizer(PhobertTokenizer):
|
| 6 |
+
def rdr_segment(self, text):
|
| 7 |
+
return ViTokenizer.tokenize(text)
|
| 8 |
+
|
| 9 |
+
def _tokenize(self, text):
|
| 10 |
+
segmented_text = self.rdr_segment(text)
|
| 11 |
+
return super()._tokenize(segmented_text)
|
pipeline.py
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from typing import Dict, List, Union
|
| 2 |
+
import torch
|
| 3 |
+
from transformers import AutoModel
|
| 4 |
+
from custom_tokenizer import CustomPhobertTokenizer
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
def mean_pooling(model_output, attention_mask):
|
| 8 |
+
token_embeddings = model_output[
|
| 9 |
+
0
|
| 10 |
+
] # First element of model_output contains all token embeddings
|
| 11 |
+
input_mask_expanded = (
|
| 12 |
+
attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
|
| 13 |
+
)
|
| 14 |
+
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(
|
| 15 |
+
input_mask_expanded.sum(1), min=1e-9
|
| 16 |
+
)
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
class PreTrainedPipeline:
|
| 20 |
+
def __init__(self, path="."):
|
| 21 |
+
self.model = AutoModel.from_pretrained(path)
|
| 22 |
+
self.tokenizer = CustomPhobertTokenizer.from_pretrained(path)
|
| 23 |
+
|
| 24 |
+
def __call__(self, inputs: Dict[str, Union[str, List[str]]]) -> List[float]:
|
| 25 |
+
"""
|
| 26 |
+
Args:
|
| 27 |
+
inputs (Dict[str, Union[str, List[str]]]):
|
| 28 |
+
a dictionary containing a query sentence and a list of key sentences
|
| 29 |
+
"""
|
| 30 |
+
|
| 31 |
+
# Combine the query sentence and key sentences into one list
|
| 32 |
+
sentences = [inputs["source_sentence"]] + inputs["sentences"]
|
| 33 |
+
|
| 34 |
+
# Tokenize sentences
|
| 35 |
+
encoded_input = self.tokenizer(
|
| 36 |
+
sentences, padding=True, truncation=True, return_tensors="pt"
|
| 37 |
+
)
|
| 38 |
+
|
| 39 |
+
# Compute token embeddings
|
| 40 |
+
with torch.no_grad():
|
| 41 |
+
model_output = self.model(**encoded_input)
|
| 42 |
+
|
| 43 |
+
# Perform pooling to get sentence embeddings
|
| 44 |
+
sentence_embeddings = mean_pooling(
|
| 45 |
+
model_output, encoded_input["attention_mask"]
|
| 46 |
+
)
|
| 47 |
+
|
| 48 |
+
# Separate the query embedding from the key embeddings
|
| 49 |
+
query_embedding = sentence_embeddings[0]
|
| 50 |
+
key_embeddings = sentence_embeddings[1:]
|
| 51 |
+
|
| 52 |
+
# Compute cosine similarities (or any other comparison method you prefer)
|
| 53 |
+
cosine_similarities = torch.nn.functional.cosine_similarity(
|
| 54 |
+
query_embedding.unsqueeze(0), key_embeddings
|
| 55 |
+
)
|
| 56 |
+
|
| 57 |
+
# Convert the tensor of cosine similarities to a list of floats
|
| 58 |
+
scores = cosine_similarities.tolist()
|
| 59 |
+
|
| 60 |
+
return scores
|
| 61 |
+
|
| 62 |
+
|
| 63 |
+
if __name__ == "__main__":
|
| 64 |
+
inputs = {
|
| 65 |
+
"source_sentence": "Anh ấy đang là sinh viên năm cuối",
|
| 66 |
+
"sentences": [
|
| 67 |
+
"Anh ấy học tại Đại học Bách khoa Hà Nội, chuyên ngành Khoa học máy tính",
|
| 68 |
+
"Anh ấy đang làm việc tại nhà máy sản xuất linh kiện điện tử",
|
| 69 |
+
"Anh ấy chuẩn bị đi du học nước ngoài",
|
| 70 |
+
"Anh ấy sắp mở cửa hàng bán mỹ phẩm",
|
| 71 |
+
"Nhà anh ấy có rất nhiều cây cảnh",
|
| 72 |
+
],
|
| 73 |
+
}
|
| 74 |
+
|
| 75 |
+
pipeline = PreTrainedPipeline()
|
| 76 |
+
res = pipeline(inputs)
|
requirements.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
pyvi>=0.1.1
|