Functions — github.com/abetlen/llama-cpp-python

hub / github.com/abetlen/llama-cpp-python / functions

Functions1,336 in github.com/abetlen/llama-cpp-python

⨍ Functions 1,336◇ Types & classes 256↳ Endpoints 95

Functiondecorator

(f)

llama_cpp/llama_cpp_ext.py:26

Functiondecorator

(f: LlamaChatCompletionHandler)

llama_cpp/llama_chat_format.py:157

Methoddefault_params

Get the default llama_model_params.

llama_cpp/_internals.py:242

Methoddefault_params

Get the default llama_context_params.

llama_cpp/_internals.py:462

Methoddesc

(self)

llama_cpp/_internals.py:106

Functiondetokenize

( body: DetokenizeInputRequest, llama_proxy: LlamaProxy = Depends(get_llama_proxy), )

llama_cpp/server/app.py:591

Methoddetokenize

Detokenize the tokens into text. Args: tokens: The list of tokens to detokenize. prev_tokens: The list of previous to

llama_cpp/llama_tokenizer.py:29

Methoddetokenize

( self, tokens: List[int], prev_tokens: Optional[List[int]] = None, special: b

llama_cpp/llama_tokenizer.py:88

Methoddraft

( self, input_ids: np.ndarray, /, *, seq_id: int, max_tokens:

examples/server/server.py:1213

Methoddraft_many

( self, requests: Sequence[Tuple[np.ndarray, int, Optional[int]]], /, )

examples/server/server.py:1818

Methodend_pos

(self)

examples/server/server.py:3867

Methodeval_logits

(self)

llama_cpp/llama.py:589

Methodeval_token_count

(self)

examples/server/server.py:3955

Methodeval_tokens

(self)

llama_cpp/llama.py:585

Methodexact_checkpoints_only

(self)

examples/server/server.py:11752

Methodextend

( self, sequence_id: int, tokens: Sequence[int], position_increments: Optional

examples/server/server.py:1045

Methodfind_first

(s: str, substrs: List[str])

llama_cpp/llama_chat_format.py:3162

Methodfingerprint_file

(path: str)

examples/server/server.py:10433

Functionformat_alpaca

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1084

Functionformat_autotokenizer

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:770

Functionformat_baichuan

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1163

Functionformat_baichuan2

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1147

Functionformat_chatglm3

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1381

Functionformat_chatml

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1343

Functionformat_gemma

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1440

Functionformat_intel

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1253

Functionformat_llama2

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1050

Functionformat_llama3

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1067

Functionformat_mistral_instruct

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1360

Functionformat_mistrallite

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1294

Functionformat_oasst_llama

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1131

Functionformat_open_orca

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1267

Functionformat_openbuddy

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1179

Functionformat_openchat

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1398

Functionformat_phind

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1239

Functionformat_pygmalion

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1327

Functionformat_qwen

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1098

Functionformat_redpajama_incite

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1201

Functionformat_saiga

( messages: list[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1418

Functionformat_snoozy

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1217

Functionformat_tokenizer_config

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:817

Functionformat_zephyr

( messages: List[llama_types.ChatCompletionRequestMessage], **kwargs: Any, )

llama_cpp/llama_chat_format.py:1310

Methodfree

(self)

llama_cpp/server/model.py:68

Methodfree_batch

()

llama_cpp/_internals.py:485

Methodfree_ctx

()

llama_cpp/_internals.py:272

Methodfree_lora_adapter

()

llama_cpp/llama.py:439

Methodfree_model

()

llama_cpp/_internals.py:72

Methodfree_sampler

()

llama_cpp/_internals.py:682

Methodfree_wrapper

(sampler: llama_cpp.llama_sampler_p)

llama_cpp/_internals.py:654

Methodfrom_file

(cls, file: Union[str, Path], verbose: bool = True)

llama_cpp/llama_grammar.py:29

Methodfrom_ggml_file

(cls, path: str)

llama_cpp/llama_tokenizer.py:73

Methodfrom_pretrained

(cls, pretrained_model_name_or_path: str)

llama_cpp/llama_tokenizer.py:109

Methodfrom_pretrained

( cls, repo_id: str, filename: Optional[str], local_dir: Optional[Union[str, o

llama_cpp/llama_chat_format.py:3185

Functionfunctionary_chat_handler

( llama: llama.Llama, messages: List[llama_types.ChatCompletionRequestMessage], functions: Optiona

llama_cpp/llama_chat_format.py:1461

Functionfunctionary_v1_v2_chat_handler

( llama: llama.Llama, messages: List[llama_types.ChatCompletionRequestMessage], functions: Optiona

llama_cpp/llama_chat_format.py:1824

Functionget_chat_completion_handler

(name: str)

llama_cpp/llama_chat_format.py:150

Methodget_dot

()

llama_cpp/llama_grammar.py:541

Methodget_dot

()

examples/server/server.py:382

Methodget_embeddings

(self)

llama_cpp/_internals.py:351

Methodget_embeddings_ith

(self, i: int)

llama_cpp/_internals.py:354

Methodget_embeddings_seq

(self, seq_id: int)

llama_cpp/_internals.py:357

Functionget_event_publisher

( request: Request, inner_send_chan: MemoryObjectSendStream[typing.Any], body: CreateCompletionReq

llama_cpp/server/app.py:191

Functionget_llama_proxy

()

llama_cpp/server/app.py:73

Methodget_logits_ith

(self, i: int)

llama_cpp/_internals.py:348

Functionget_models

( llama_proxy: LlamaProxy = Depends(get_llama_proxy), )

llama_cpp/server/app.py:535

Methodget_recursive_refs

(ks, first_is_optional)

llama_cpp/llama_grammar.py:906

Methodget_recursive_refs

(keys: List[str], first_is_optional: bool)

examples/server/server.py:733

Methodget_route_handler

Defines custom route handler that catches exceptions and formats in OpenAI style error response

llama_cpp/server/errors.py:162

Methodget_seed

(self)

llama_cpp/_internals.py:854

Methodget_state_size

(self)

llama_cpp/_internals.py:315

Methodgrammar_accept_token

(self, grammar: LlamaGrammar, token: int)

llama_cpp/_internals.py:449

Functionguess_chat_format_from_gguf_metadata

(metadata: Dict[str, str])

llama_cpp/llama_chat_format.py:853

Methodhas_attention_budget

(self)

examples/server/server.py:11756

Functionhealthz

()

examples/server/server.py:16332

Functionhf_autotokenizer_to_chat_completion_handler

( pretrained_model_name_or_path: Union[str, os.PathLike[str]], )

llama_cpp/llama_chat_format.py:787

Functionhf_tokenizer_config_to_chat_completion_handler

( tokenizer_config: Dict[str, Any], add_generation_prompt: bool = True, )

llama_cpp/llama_chat_format.py:843

Methodjoin_seq

()

llama_cpp/llama_grammar.py:549

Methodjoin_sequence

()

examples/server/server.py:386

Methodkv_cache_seq_cp

(self, seq_id_src: int, seq_id_dst: int, p0: int, p1: int)

llama_cpp/_internals.py:303

Methodkv_cache_seq_keep

(self, seq_id: int)

llama_cpp/_internals.py:307

Methodkv_cache_seq_shift

(self, seq_id: int, p0: int, p1: int, shift: int)

llama_cpp/_internals.py:311

Methodlast

(self)

llama_cpp/_internals.py:620

Methodlength

(self)

examples/server/server.py:3951

Methodlength

(self)

examples/server/server.py:4140

Functionlist_models

()

examples/server/server.py:16316

Functionllama_adapter_get_alora_invocation_tokens

Get the invocation tokens if the current LoRA is an aLoRA.

llama_cpp/llama_cpp.py:2151

Functionllama_adapter_get_alora_n_invocation_tokens

Get the invocation token count if the current LoRA is an aLoRA.

llama_cpp/llama_cpp.py:2138

Functionllama_adapter_lora_free

(adapter: llama_adapter_lora_p, /)

llama_cpp/llama_cpp.py:2128

Functionllama_adapter_lora_init

( model: llama_model_p, path_lora: bytes, / )

llama_cpp/llama_cpp.py:2036

Functionllama_adapter_meta_count

Get the number of adapter metadata key/value pairs.

llama_cpp/llama_cpp.py:2069

Functionllama_adapter_meta_key_by_index

Get adapter metadata key name by index.

llama_cpp/llama_cpp.py:2086

Functionllama_adapter_meta_val_str

Get adapter metadata value as a string by key name.

llama_cpp/llama_cpp.py:2053

Functionllama_adapter_meta_val_str_by_index

Get adapter metadata value as a string by index.

llama_cpp/llama_cpp.py:2109

Functionllama_add_bos_token

(vocab: llama_vocab_p, /)

llama_cpp/llama_cpp.py:3787

Functionllama_add_eos_token

(vocab: llama_vocab_p, /)

llama_cpp/llama_cpp.py:3804

Functionllama_apply_adapter_cvec

( ctx: llama_context_p, data: CtypesPointerOrRef[ctypes.c_float], len: int, n_embd: int, i

llama_cpp/llama_cpp.py:2246

Functionllama_backend_free

Call once at the end of the program - currently only used for MPI

llama_cpp/llama_cpp.py:1314

Functionllama_backend_init

Initialize the llama + ggml backend Call once at the start of the program

llama_cpp/llama_cpp.py:1284

Functionllama_batch_free

Frees a batch of tokens allocated with llama_batch_init()

llama_cpp/llama_cpp.py:3033

Functionllama_batch_get_one

Return batch for single sequence of tokens NOTE: this is a helper function to facilitate transition to the new batch API - avoid using it

llama_cpp/llama_cpp.py:2988

← previous next →801–900 of 1,336, ranked by callers