File size: 1,374 Bytes
8c45550
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
from transformers import PretrainedConfig



class M3DCLIPConfig(PretrainedConfig):
    model_type = "m3d_clip"

    def __init__(

        self,

        language_model_name_or_path: str = 'bert-base-uncased',

        local_loss: bool = False,

        gather_loss: bool = True,

        in_channels: int = 1,

        img_size: tuple = (32, 256, 256),

        patch_size: tuple = (4, 16, 16),

        hidden_size: int = 768,

        mlp_dim: int = 3072,

        num_layers: int = 12,

        num_heads: int = 12,

        pos_embed: str = "perceptron",

        dropout_rate: float = 0,

        spatial_dims: int = 3,

        max_text_len: int = 128,

        vocab_size: int = 30522,

        **kwargs,

    ):
        self.language_model_name_or_path = language_model_name_or_path
        self.in_channels = in_channels
        self.img_size = img_size
        self.patch_size = patch_size
        self.hidden_size = hidden_size
        self.mlp_dim = mlp_dim
        self.num_layers = num_layers
        self.num_heads = num_heads
        self.pos_embed = pos_embed
        self.dropout_rate = dropout_rate
        self.spatial_dims = spatial_dims
        self.local_loss = local_loss
        self.gather_loss = gather_loss
        self.max_text_len = max_text_len
        self.vocab_size = vocab_size
        super().__init__(**kwargs)