File size: 1,659 Bytes
cb9cb86
 
5760b3d
cb9cb86
5760b3d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cb9cb86
5760b3d
cb9cb86
 
5760b3d
cb9cb86
 
 
 
 
5760b3d
 
cb9cb86
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
import torch
import torch.nn as nn
from transformers import AutoModel, AutoTokenizer, XLMRobertaModel

# カスタムレイヤーの定義
class SparseLinear(nn.Module):
    def __init__(self, input_dim, output_dim):
        super(SparseLinear, self).__init__()
        self.linear = nn.Linear(input_dim, output_dim)

    def forward(self, x):
        return self.linear(x)

# カスタムモデルの定義
class CustomXLMRobertaModel(XLMRobertaModel):
    def __init__(self, config):
        super(CustomXLMRobertaModel, self).__init__(config)
        self.sparse_linear = SparseLinear(config.hidden_size, 1)  # 適切な出力次元を設定

    def forward(self, *args, **kwargs):
        outputs = super(CustomXLMRobertaModel, self).forward(*args, **kwargs)
        dense_embeddings = outputs.last_hidden_state
        sparse_embeddings = self.sparse_linear(dense_embeddings)
        return outputs, sparse_embeddings

# モデルとトークナイザーのロード
model_name = "."  # ローカルディレクトリを指定
tokenizer = AutoTokenizer.from_pretrained(model_name)
config = AutoModel.from_pretrained(model_name).config

# マージされたモデルのロード
merged_model = CustomXLMRobertaModel.from_pretrained(model_name, config=config)
merged_model.load_state_dict(torch.load("merged_pytorch_model.bin"))

# テキストのエンコード
def encode_text(text):
    inputs = tokenizer(text, return_tensors="pt")
    outputs, sparse_embeddings = merged_model(**inputs)
    return outputs, sparse_embeddings

# テキストのエンコード例
text = "こんにちは"
sparse_embeddings = encode_text(text)
print(sparse_embeddings)