423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-23 06:12:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

remove bs condition

Browse Source 
Former-commit-id: 962b9730a7a2940a0d4e5c76d1fe41d0fef76547
This commit is contained in:
Kingsley 2024-10-14 16:55:59 +08:00
parent 8ea1c5c69e
commit 2c5f912e16
1 changed files with 2 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
src/llamafactory/data/mm_plugin.py
View File

@ -477,7 +477,7 @@ class PixtralPlugin(BasePlugin):
if image_input_sizes is None:
raise ValueError("The number of images does not match the number of {} tokens".format(IMAGE_PLACEHOLDER))
image_size = image_input_sizes[0][img_id]
height, width = image_size
num_height_tokens = height // patch_size
@ -500,7 +500,7 @@ class PixtralPlugin(BasePlugin):
raise ValueError("The number of images does not match the number of {} tokens".format(IMAGE_PLACEHOLDER))
return messages
@override
def get_mm_inputs(
self,
@ -516,11 +516,6 @@ class PixtralPlugin(BasePlugin):
if mm_inputs.get("image_sizes"):
mm_inputs.pop("image_sizes")
if isinstance(mm_inputs.get("pixel_values"), list) and len(mm_inputs.get("pixel_values")[0]) >= 2:
raise ValueError("Now it only supports batchsize=1 on per gpu due to `List[tensor]` can not pack into BachEncoding")
mm_inputs["pixel_values"] = mm_inputs.get("pixel_values")[0][0].unsqueeze(0)
return mm_inputs
class Qwen2vlPlugin(BasePlugin):