visheratin
commited on
Commit
•
89d694e
1
Parent(s):
2f07d8b
Update nllb_mrl.py
Browse files- nllb_mrl.py +31 -19
nllb_mrl.py
CHANGED
@@ -1,21 +1,26 @@
|
|
1 |
-
from dataclasses import dataclass
|
2 |
from typing import List, Union
|
3 |
|
4 |
import torch
|
5 |
import torch.nn as nn
|
6 |
import torch.nn.functional as F
|
7 |
-
from
|
8 |
-
from open_clip import
|
9 |
from PIL import Image
|
10 |
-
from transformers import PretrainedConfig
|
11 |
|
12 |
|
13 |
-
@dataclass
|
14 |
class MatryoshkaNllbClipConfig(PretrainedConfig):
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
|
21 |
class MatryoshkaLayer(nn.Module):
|
@@ -37,16 +42,23 @@ class MatryoshkaLayer(nn.Module):
|
|
37 |
return outputs
|
38 |
|
39 |
|
40 |
-
class MatryoshkaNllbClip(
|
|
|
|
|
41 |
def __init__(self, config: MatryoshkaNllbClipConfig, device):
|
42 |
-
super().__init__()
|
43 |
if isinstance(device, str):
|
44 |
device = torch.device(device)
|
45 |
self.config = config
|
46 |
-
self.model
|
47 |
-
config.clip_model_name,
|
|
|
|
|
|
|
|
|
|
|
48 |
)
|
49 |
-
self.
|
50 |
self.model.to(device)
|
51 |
self.matryoshka_layer = MatryoshkaLayer(
|
52 |
config.mrl_resolutions, config.target_resolution
|
@@ -55,8 +67,8 @@ class MatryoshkaNllbClip(nn.Module, PyTorchModelHubMixin):
|
|
55 |
self.tokenizer = get_tokenizer(config.clip_model_name)
|
56 |
|
57 |
def forward(self, image_inputs, input_ids, resolution: Union[int, None] = None):
|
58 |
-
image_inputs = image_inputs.to(self.
|
59 |
-
input_ids = input_ids.to(self.
|
60 |
outputs = self.model(
|
61 |
image=image_inputs,
|
62 |
text=input_ids,
|
@@ -118,7 +130,7 @@ class MatryoshkaNllbClip(nn.Module, PyTorchModelHubMixin):
|
|
118 |
resolution: Union[int, None] = None,
|
119 |
):
|
120 |
image_inputs = [self.transform(image) for image in images]
|
121 |
-
image_inputs = torch.stack(image_inputs, dim=0).to(self.
|
122 |
with torch.inference_mode():
|
123 |
features = self.model.visual(image_inputs)
|
124 |
if resolution is not None:
|
@@ -138,10 +150,10 @@ class MatryoshkaNllbClip(nn.Module, PyTorchModelHubMixin):
|
|
138 |
):
|
139 |
if langs is None:
|
140 |
langs = ["eng_Latn"] * len(texts)
|
141 |
-
texts = [f"{lang}
|
142 |
input_ids = self.tokenizer.tokenizer.batch_encode_plus(
|
143 |
texts, return_tensors="pt", padding="longest", add_special_tokens=False
|
144 |
-
)["input_ids"].to(self.
|
145 |
with torch.inference_mode():
|
146 |
features = self.model.text(input_ids)
|
147 |
if resolution is not None:
|
|
|
|
|
1 |
from typing import List, Union
|
2 |
|
3 |
import torch
|
4 |
import torch.nn as nn
|
5 |
import torch.nn.functional as F
|
6 |
+
from open_clip import create_model, get_tokenizer
|
7 |
+
from open_clip.transform import PreprocessCfg, image_transform_v2
|
8 |
from PIL import Image
|
9 |
+
from transformers import PretrainedConfig, PreTrainedModel
|
10 |
|
11 |
|
|
|
12 |
class MatryoshkaNllbClipConfig(PretrainedConfig):
|
13 |
+
def __init__(
|
14 |
+
self,
|
15 |
+
clip_model_name: str = "",
|
16 |
+
target_resolution: int = -1,
|
17 |
+
mrl_resolutions: List[int] = [],
|
18 |
+
**kwargs,
|
19 |
+
):
|
20 |
+
super().__init__(**kwargs)
|
21 |
+
self.clip_model_name = clip_model_name
|
22 |
+
self.target_resolution = target_resolution
|
23 |
+
self.mrl_resolutions = mrl_resolutions
|
24 |
|
25 |
|
26 |
class MatryoshkaLayer(nn.Module):
|
|
|
42 |
return outputs
|
43 |
|
44 |
|
45 |
+
class MatryoshkaNllbClip(PreTrainedModel):
|
46 |
+
config_class = MatryoshkaNllbClipConfig
|
47 |
+
|
48 |
def __init__(self, config: MatryoshkaNllbClipConfig, device):
|
49 |
+
super().__init__(config)
|
50 |
if isinstance(device, str):
|
51 |
device = torch.device(device)
|
52 |
self.config = config
|
53 |
+
self.model = create_model(
|
54 |
+
config.clip_model_name, output_dict=True
|
55 |
+
)
|
56 |
+
pp_cfg = PreprocessCfg(**self.model.visual.preprocess_cfg)
|
57 |
+
self.transform = image_transform_v2(
|
58 |
+
pp_cfg,
|
59 |
+
is_train=False,
|
60 |
)
|
61 |
+
self._device = device
|
62 |
self.model.to(device)
|
63 |
self.matryoshka_layer = MatryoshkaLayer(
|
64 |
config.mrl_resolutions, config.target_resolution
|
|
|
67 |
self.tokenizer = get_tokenizer(config.clip_model_name)
|
68 |
|
69 |
def forward(self, image_inputs, input_ids, resolution: Union[int, None] = None):
|
70 |
+
image_inputs = image_inputs.to(self._device)
|
71 |
+
input_ids = input_ids.to(self._device)
|
72 |
outputs = self.model(
|
73 |
image=image_inputs,
|
74 |
text=input_ids,
|
|
|
130 |
resolution: Union[int, None] = None,
|
131 |
):
|
132 |
image_inputs = [self.transform(image) for image in images]
|
133 |
+
image_inputs = torch.stack(image_inputs, dim=0).to(self._device)
|
134 |
with torch.inference_mode():
|
135 |
features = self.model.visual(image_inputs)
|
136 |
if resolution is not None:
|
|
|
150 |
):
|
151 |
if langs is None:
|
152 |
langs = ["eng_Latn"] * len(texts)
|
153 |
+
texts = [f"{lang}{text}" for lang, text in zip(langs, texts)]
|
154 |
input_ids = self.tokenizer.tokenizer.batch_encode_plus(
|
155 |
texts, return_tensors="pt", padding="longest", add_special_tokens=False
|
156 |
+
)["input_ids"].to(self._device)
|
157 |
with torch.inference_mode():
|
158 |
features = self.model.text(input_ids)
|
159 |
if resolution is not None:
|