Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

metal : utilize max shared memory for mul_mat_id #7935

Merged
merged 1 commit into from
Jun 14, 2024

Conversation

ggerganov
Copy link
Member

fix #7652

Allows larger batch sizes for MoE models (e.g. DeepSeek v2), though using -ub 256 remains more efficient in terms of speed

@ggerganov ggerganov merged commit 66ef1ce into master Jun 14, 2024
68 checks passed
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.

Bug: DeepSeek-V2-Lite GGML_ASSERT: ggml-metal.m:1857: dst_rows <= 2048 and aborts
2 participants