| .. |
|
CMakeLists.txt
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-adapter.cpp
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-adapter.h
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-arch.cpp
|
fbdfefe74e
llama : gemma3 : use output tensor if it exists in model weight (#12506)
|
пре 10 месеци |
|
llama-arch.h
|
7dfad387e3
llama: Add support for RWKV v7 architecture (#12412)
|
пре 10 месеци |
|
llama-batch.cpp
|
f66f582927
llama : refactor `src/llama.cpp` (#10902)
|
пре 1 година |
|
llama-batch.h
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-chat.cpp
|
5bbe6a9fe9
ggml : portability fixes for VS 2017 (#12150)
|
пре 11 месеци |
|
llama-chat.h
|
0cec062a63
llama : add support for GLM-Edge and GLM-Edge-V series models (#10573)
|
пре 1 година |
|
llama-context.cpp
|
568013d0cd
context : clear sets containing encoder output sequence ids before storing new values (#12470)
|
пре 10 месеци |
|
llama-context.h
|
8fcb563613
Load all MoE experts during warmup (#11571)
|
пре 10 месеци |
|
llama-cparams.cpp
|
f66f582927
llama : refactor `src/llama.cpp` (#10902)
|
пре 1 година |
|
llama-cparams.h
|
8fcb563613
Load all MoE experts during warmup (#11571)
|
пре 10 месеци |
|
llama-grammar.cpp
|
669912d9a5
`tool-call`: fix Qwen 2.5 Coder support, add micro benchmarks, support trigger patterns for lazy grammars (#12034)
|
пре 11 месеци |
|
llama-grammar.h
|
669912d9a5
`tool-call`: fix Qwen 2.5 Coder support, add micro benchmarks, support trigger patterns for lazy grammars (#12034)
|
пре 11 месеци |
|
llama-graph.cpp
|
75422e8bc4
graph : normalize Q, K, V shapes + sync cross attention (#12449)
|
пре 10 месеци |
|
llama-graph.h
|
75422e8bc4
graph : normalize Q, K, V shapes + sync cross attention (#12449)
|
пре 10 месеци |
|
llama-hparams.cpp
|
081bee8c64
hparams : add SWA rope parameters (#12374)
|
пре 10 месеци |
|
llama-hparams.h
|
7dfad387e3
llama: Add support for RWKV v7 architecture (#12412)
|
пре 10 месеци |
|
llama-impl.cpp
|
53ff6b9b9f
GGUF: C++ refactor, backend support, misc fixes (#11030)
|
пре 1 година |
|
llama-impl.h
|
fef0cbeadf
cleanup: fix compile warnings associated with gnu_printf (#11811)
|
пре 11 месеци |
|
llama-io.cpp
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-io.h
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-kv-cache.cpp
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-kv-cache.h
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-memory.cpp
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-memory.h
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
llama-mmap.cpp
|
c54f6b7988
mmap : skip resource limit checks on AIX (#12541)
|
пре 10 месеци |
|
llama-mmap.h
|
19b392d58d
llama-mmap: fix missing include (#11796)
|
пре 11 месеци |
|
llama-model-loader.cpp
|
a5203b4465
llama : minor fixes for up llama load model speed (#11448)
|
пре 1 година |
|
llama-model-loader.h
|
681149ced2
llama : add `llama_model_load_from_splits` (#11255)
|
пре 1 година |
|
llama-model.cpp
|
fbdfefe74e
llama : gemma3 : use output tensor if it exists in model weight (#12506)
|
пре 10 месеци |
|
llama-model.h
|
7dfad387e3
llama: Add support for RWKV v7 architecture (#12412)
|
пре 10 месеци |
|
llama-quant.cpp
|
7dfad387e3
llama: Add support for RWKV v7 architecture (#12412)
|
пре 10 месеци |
|
llama-quant.h
|
f66f582927
llama : refactor `src/llama.cpp` (#10902)
|
пре 1 година |
|
llama-sampling.cpp
|
669912d9a5
`tool-call`: fix Qwen 2.5 Coder support, add micro benchmarks, support trigger patterns for lazy grammars (#12034)
|
пре 11 месеци |
|
llama-sampling.h
|
afa8a9ec9b
llama : add `llama_vocab`, functions -> methods, naming (#11110)
|
пре 1 година |
|
llama-vocab.cpp
|
00d53800e0
llama-vocab : add SuperBPE pre-tokenizer (#12532)
|
пре 10 месеци |
|
llama-vocab.h
|
08f10f69c3
llama : remove notion of CLS token (#11064)
|
пре 1 година |
|
llama.cpp
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
пре 10 месеци |
|
unicode-data.cpp
|
458367a906
server : better security control for public deployments (#9776)
|
пре 1 година |
|
unicode-data.h
|
a39ab216aa
llama : reduce compile time and binary size (#9712)
|
пре 1 година |
|
unicode.cpp
|
68ff663a04
repo : update links to new url (#11886)
|
пре 11 месеци |
|
unicode.h
|
08ea539df2
unicode : improve naming style (#10838)
|
пре 1 година |
compilade
00d53800e0
llama-vocab : add SuperBPE pre-tokenizer (#12532)