[GH-ISSUE #109] dispatch_model cannot put some parameter to cuda #90

Open
opened 2026-10-26 21:11:46 +00:00 by bsakaguchi · 0 comments
Owner

Originally created by @erikchwang on GitHub (Mar 4, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-VL2/issues/109

accelerate.dispatch_model() cannot put the following parameters to cuda:

self.image_newline
self.view_seperator
self.tile_indicators

they are still on cpu after dispatch_model()

Originally created by @erikchwang on GitHub (Mar 4, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-VL2/issues/109 accelerate.dispatch_model() cannot put the following parameters to cuda: > self.image_newline > self.view_seperator > self.tile_indicators they are still on cpu after dispatch_model()
Sign in to join this conversation.
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
deepseek-ai/DeepSeek-VL2#90
No description provided.