llama.cpp
9fd1e83f
- Use Q4_K for attn_v for Q2_K_S when n_gqa >= 4
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
2 years ago
Use Q4_K for attn_v for Q2_K_S when n_gqa >= 4
References
ik/better_q2_k_s
#4996 - Use Q4_K for attn_v for Q2_K_S when n_gqa >= 4
Author
Iwan Kawrakow
Committer
Iwan Kawrakow
Parents
75632936
Loading