llama.cpp
9b4d63ae - convert : add "tokenizer.ggml.pre" GGUF KV (wip)

Commit

1 year ago

convert : add "tokenizer.ggml.pre" GGUF KV (wip)

References

#6920 - llama : improve BPE pre-processing + LLaMA 3 and Deepseek support

Author

ggerganov

ggerganov

Parents

Loading