llama.cpp
1466621e - llama : Support llama 4 text-only (#12791)

Commit

112 days ago

llama : Support llama 4 text-only (#12791) * llama4 conversion * initial support, no chat template * clean up a bit * fix tokenizer conversion * correct hparams * try this * fix shexp * ffn_inp_normed * chat template * clean up model conversion * add_bos * add scale_before_ffn * fix order * weight_before_ffn * llm_graph_input_attn_temp * add chunk attn mask * build_inp_attn_scale() * add comment about ggml_repeat * clarify comments * fix build

References

#12791 - llama : Support llama 4 text-only

Author

ngxson

Parents

82974011

Files17

convert_hf_to_gguf.py
convert_hf_to_gguf_update.py
gguf-py/gguf
- constants.py
- gguf_writer.py
include
- llama.h
models
- ggml-vocab-llama4.gguf.inp
- ggml-vocab-llama4.gguf.out
src
- llama-arch.cpp
- llama-arch.h
- llama-chat.cpp
- llama-chat.h
- llama-graph.cpp
- llama-graph.h
- llama-hparams.h
- llama-model.cpp
- llama-model.h
- llama-vocab.cpp

llama.cpp 1466621e - llama : Support llama 4 text-only (#12791)

llama.cpp
1466621e - llama : Support llama 4 text-only (#12791)