Spaces:
Running
on
Zero
Running
on
Zero
Men1scus
commited on
Commit
·
66217ed
1
Parent(s):
4b18c4d
Update vision tower name to hf
Browse files
llava/model/multimodal_encoder/clip_encoder.py
CHANGED
|
@@ -2,7 +2,7 @@ import torch
|
|
| 2 |
import torch.nn as nn
|
| 3 |
|
| 4 |
from transformers import CLIPVisionModel, CLIPImageProcessor, CLIPVisionConfig
|
| 5 |
-
from CKPT_PTH import LLAVA_CLIP_PATH
|
| 6 |
|
| 7 |
class CLIPVisionTower(nn.Module):
|
| 8 |
def __init__(self, vision_tower, args, delay_load=False):
|
|
@@ -10,7 +10,7 @@ class CLIPVisionTower(nn.Module):
|
|
| 10 |
|
| 11 |
self.is_loaded = False
|
| 12 |
|
| 13 |
-
self.vision_tower_name =
|
| 14 |
self.select_layer = args.mm_vision_select_layer
|
| 15 |
self.select_feature = getattr(args, 'mm_vision_select_feature', 'patch')
|
| 16 |
|
|
|
|
| 2 |
import torch.nn as nn
|
| 3 |
|
| 4 |
from transformers import CLIPVisionModel, CLIPImageProcessor, CLIPVisionConfig
|
| 5 |
+
# from CKPT_PTH import LLAVA_CLIP_PATH
|
| 6 |
|
| 7 |
class CLIPVisionTower(nn.Module):
|
| 8 |
def __init__(self, vision_tower, args, delay_load=False):
|
|
|
|
| 10 |
|
| 11 |
self.is_loaded = False
|
| 12 |
|
| 13 |
+
self.vision_tower_name = 'openai/clip-vit-large-patch14-336'
|
| 14 |
self.select_layer = args.mm_vision_select_layer
|
| 15 |
self.select_feature = getattr(args, 'mm_vision_select_feature', 'patch')
|
| 16 |
|