EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

fix batch error for llava-hd (#98)

Browse Source
This commit is contained in:
shiyi.c_98
2024-01-25 07:56:25 -08:00
committed by GitHub
parent 23950056f0
commit 0147f940dd
1 changed files with 12 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
python/sglang/srt/models/llava.py
View File

@@ -112,24 +112,28 @@ class LlavaLlamaForCausalLM(nn.Module):
need_vision = need_vision & has_pixel need_vision = need_vision & has_pixel
if need_vision.any(): if need_vision.any():
pixel_values = torch.tensor( pixel_values = [pixel_values[i] for i in range(bs) if need_vision[i]]
np.array([pixel_values[i] for i in range(bs) if need_vision[i]]), image_sizes = [image_sizes[i] for i in range(bs) if need_vision[i]]
device=self.vision_tower.device,
)
########## Encode Image ######## ########## Encode Image ########
if pixel_values.ndim == 5: if pixel_values[0].ndim == 4:
# llava-hd: BS, num_patch, C=3, H=336, W=336, num_patch obtained from process_images # llava-hd: BS, num_patch, C=3, H=336, W=336, num_patch obtained from process_images
concat_images = torch.cat( np.concatenate(pixel_values, axis=0)
[image for image in pixel_values], dim=0 # ndim=4
) # ndim=4 concat_images = torch.tensor(
np.concatenate(pixel_values, axis=0),
device=self.vision_tower.device,
)
image_features = self.encode_images(concat_images) image_features = self.encode_images(concat_images)
split_sizes = [image.shape[0] for image in pixel_values] split_sizes = [image.shape[0] for image in pixel_values]
image_features = torch.split(image_features, split_sizes, dim=0) image_features = torch.split(image_features, split_sizes, dim=0)
# hd image_features: BS, num_patch, 576, 4096 # hd image_features: BS, num_patch, 576, 4096
else: else:
# normal pixel: BS, C=3, H=336, W=336 # normal pixel: BS, C=3, H=336, W=336
pixel_values = torch.tensor(
np.array(pixel_values), device=self.vision_tower.device
)
image_features = self.encode_images(pixel_values) image_features = self.encode_images(pixel_values)
# image_features: BS, 576, 4096 # image_features: BS, 576, 4096