llama.cpp
8abcc70a - model: (qwen3next) correct vectorized key_gdiff calculation (#19324)

Commit

4 days ago

model: (qwen3next) correct vectorized key_gdiff calculation (#19324) * model: (qwen3next) correct vectorized key_gdiff calculation * move transpose to outside of loop

References

#19324 - model: (qwen3next) correct vectorized key_gdiff calculation

Author

ngxson

Parents

eaba92c3

llama.cpp 8abcc70a - model: (qwen3next) correct vectorized key_gdiff calculation (#19324)

llama.cpp
8abcc70a - model: (qwen3next) correct vectorized key_gdiff calculation (#19324)