Skip to content

Conversation

@JackWeiw
Copy link
Contributor

@JackWeiw JackWeiw commented May 7, 2025

Motivation

Dlinfer Ascend 310P device will trigger error when using lmdeploy chat api, since the original default block size is 64.
Set block size to 128 will prevent this error

@jinminxi104 jinminxi104 marked this pull request as draft May 7, 2025 11:22
@jinminxi104 jinminxi104 requested review from grimoire and lvhan028 May 8, 2025 01:23
@JackWeiw JackWeiw marked this pull request as ready for review May 8, 2025 05:18
@jinminxi104 jinminxi104 marked this pull request as draft May 9, 2025 05:09
@jinminxi104
Copy link
Collaborator

waiting for checking on 910B

@JackWeiw JackWeiw marked this pull request as ready for review May 9, 2025 10:44
@jinminxi104 jinminxi104 marked this pull request as draft May 9, 2025 16:06
@jinminxi104
Copy link
Collaborator

better to set block_size=128 on 910b

@jinminxi104 jinminxi104 self-requested a review May 9, 2025 16:07
@jinminxi104 jinminxi104 closed this Jun 9, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants