| Safe Haskell | None |
|---|---|
| Language | Haskell2010 |
Llama.Internal.Foreign.Model
Synopsis
- p_llama_free :: FinalizerPtr CLlamaContext
- p_llama_model_free :: FinalizerPtr CLlamaModel
- c_llama_free :: CLlamaContext -> IO ()
- c_llama_init_from_model_wrap :: CLlamaModel -> CLlamaContextParams -> IO CLlamaContext
- c_llama_model_default_params :: CLlamaModelParams -> IO ()
- c_llama_model_load_from_file_wrap :: CString -> CLlamaModelParams -> IO CLlamaModel
- c_llama_model_free :: CLlamaModel -> IO ()
- c_llama_model_get_vocab :: CLlamaModel -> IO CLlamaVocab
- c_llama_model_load_from_splits :: Ptr CString -> CSize -> CLlamaModelParams -> IO CLlamaModel
- c_llama_model_rope_type_into :: CLlamaModel -> Ptr CInt -> IO ()
- c_llama_model_n_ctx_train :: CLlamaModel -> IO Int32
- c_llama_model_n_embd :: CLlamaModel -> IO Int32
- c_llama_model_n_layer :: CLlamaModel -> IO Int32
- c_llama_model_n_head :: CLlamaModel -> IO Int32
- c_llama_model_n_head_kv :: CLlamaModel -> IO Int32
- c_llama_get_model :: CLlamaContext -> IO CLlamaModel
- c_llama_model_rope_freq_scale_train :: CLlamaModel -> IO CFloat
- c_llama_vocab_type :: CLlamaVocab -> IO CInt
- c_llama_model_meta_val_str :: CLlamaModel -> CString -> CString -> CSize -> IO CInt
- c_llama_model_meta_count :: CLlamaModel -> IO CInt
- c_llama_model_meta_key_by_index :: CLlamaModel -> CInt -> CString -> CSize -> IO CInt
- c_llama_model_meta_val_str_by_index :: CLlamaModel -> CInt -> CString -> CSize -> IO CInt
- c_llama_model_desc :: CLlamaModel -> CString -> CSize -> IO CInt
- c_llama_model_size :: CLlamaModel -> IO Word64
- c_llama_model_chat_template :: CLlamaModel -> CString -> IO CString
- c_llama_model_n_params :: CLlamaModel -> IO Word64
- c_llama_model_has_encoder :: CLlamaModel -> IO CBool
- c_llama_model_has_decoder :: CLlamaModel -> IO CBool
- c_llama_model_decoder_start_token :: CLlamaModel -> IO LlamaToken
- c_llama_model_is_recurrent :: CLlamaModel -> IO CBool
- c_llama_model_quantize :: CString -> CString -> CLlamaModelQuantizeParams -> IO Word32
- c_llama_model_quantize_default_params :: IO CLlamaModelQuantizeParams
Documentation
p_llama_free :: FinalizerPtr CLlamaContext Source #
Foreign pointer to the C function for freeing the context.
p_llama_model_free :: FinalizerPtr CLlamaModel Source #
Foreign pointer to the C function for freeing the model.
c_llama_free :: CLlamaContext -> IO () Source #
c_llama_model_free :: CLlamaModel -> IO () Source #
c_llama_model_load_from_splits :: Ptr CString -> CSize -> CLlamaModelParams -> IO CLlamaModel Source #
c_llama_model_rope_type_into :: CLlamaModel -> Ptr CInt -> IO () Source #
Get RoPE type from model
c_llama_model_n_ctx_train :: CLlamaModel -> IO Int32 Source #
Get training context size
c_llama_model_n_embd :: CLlamaModel -> IO Int32 Source #
Get embedding dimension
c_llama_model_n_layer :: CLlamaModel -> IO Int32 Source #
Get number of layers
c_llama_model_n_head :: CLlamaModel -> IO Int32 Source #
Get number of heads
c_llama_model_n_head_kv :: CLlamaModel -> IO Int32 Source #
Get number of key/value heads
c_llama_get_model :: CLlamaContext -> IO CLlamaModel Source #
Get model from context
c_llama_vocab_type :: CLlamaVocab -> IO CInt Source #
c_llama_model_meta_val_str :: CLlamaModel -> CString -> CString -> CSize -> IO CInt Source #
LLAMA_API int32_t llama_model_meta_val_str(const struct llama_model * model, const char * key, char * buf, size_t buf_size);
c_llama_model_meta_count :: CLlamaModel -> IO CInt Source #
LLAMA_API int32_t llama_model_meta_count(const struct llama_model * model);
c_llama_model_meta_key_by_index :: CLlamaModel -> CInt -> CString -> CSize -> IO CInt Source #
LLAMA_API int32_t llama_model_meta_key_by_index(const struct llama_model * model, int32_t i, char * buf, size_t buf_size);
c_llama_model_meta_val_str_by_index :: CLlamaModel -> CInt -> CString -> CSize -> IO CInt Source #
LLAMA_API int32_t llama_model_meta_val_str_by_index(const struct llama_model * model, int32_t i, char * buf, size_t buf_size);
c_llama_model_desc :: CLlamaModel -> CString -> CSize -> IO CInt Source #
LLAMA_API int32_t llama_model_desc(const struct llama_model * model, char * buf, size_t buf_size);
c_llama_model_size :: CLlamaModel -> IO Word64 Source #
LLAMA_API uint64_t llama_model_size(const struct llama_model * model);
c_llama_model_chat_template :: CLlamaModel -> CString -> IO CString Source #
LLAMA_API const char * llama_model_chat_template(const struct llama_model * model, const char * name);
c_llama_model_n_params :: CLlamaModel -> IO Word64 Source #
LLAMA_API uint64_t llama_model_n_params(const struct llama_model * model);
c_llama_model_has_encoder :: CLlamaModel -> IO CBool Source #
LLAMA_API bool llama_model_has_encoder(const struct llama_model * model);
c_llama_model_has_decoder :: CLlamaModel -> IO CBool Source #
LLAMA_API bool llama_model_has_decoder(const struct llama_model * model);
c_llama_model_decoder_start_token :: CLlamaModel -> IO LlamaToken Source #
LLAMA_API llama_token llama_model_decoder_start_token(const struct llama_model * model);
c_llama_model_is_recurrent :: CLlamaModel -> IO CBool Source #
LLAMA_API bool llama_model_is_recurrent(const struct llama_model * model);
c_llama_model_quantize :: CString -> CString -> CLlamaModelQuantizeParams -> IO Word32 Source #
LLAMA_API uint32_t llama_model_quantize( const char * fname_inp, const char * fname_out, const llama_model_quantize_params * params);