import json import multiprocessing from re import compile, Match, Pattern from threading import Lock from functools import partial from typing import Callable, Coroutine, Iterator, List, Optional, Tuple, Union, Dict from typing_extensions import TypedDict, Literal import llama_cpp import anyio from anyio.streams.memory import MemoryObjectSendStream from starlette.concurrency import run_in_threadpool, iterate_in_threadpool from fastapi import Depends, FastAPI, APIRouter, Request, Response from fastapi.middleware.cors import CORSMiddleware from fastapi.responses import JSONResponse from fastapi.routing import APIRoute from pydantic import BaseModel, Field from pydantic_settings import BaseSettings from sse_starlette.sse import EventSourceResponse class Settings(BaseSettings): model: str = Field( description="The path to the model to use for generating completions." ) model_alias: Optional[str] = Field( default=None, description="The alias of the model to use for generating completions.", ) n_ctx: int = Field(default=2048, ge=1, description="The context size.") n_gpu_layers: int = Field( default=0, ge=0, description="The number of layers to put on the GPU. The rest will be on the CPU.", ) tensor_split: Optional[List[float]] = Field( default=None, description="Split layers across multiple GPUs in proportion.", ) rope_freq_base: float = Field(default=10000, ge=1, description="RoPE base frequency") rope_freq_scale: float = Field(default=1.0, description="RoPE frequency scaling factor") seed: int = Field( default=1337, description="Random seed. -1 for random." ) n_batch: int = Field( default=512, ge=1, description="The batch size to use per eval." ) n_threads: int = Field( default=max(multiprocessing.cpu_count() // 2, 1), ge=1, description="The number of threads to use.", ) f16_kv: bool = Field(default=True, description="Whether to use f16 key/value.") use_mlock: bool = Field( default=llama_cpp.llama_mlock_supported(), description="Use mlock.", ) use_mmap: bool = Field( default=llama_cpp.llama_mmap_supported(), description="Use mmap.", ) embedding: bool = Field(default=True, description="Whether to use embeddings.") low_vram: bool = Field( default=False, description="Whether to use less VRAM. This will reduce performance.", ) last_n_tokens_size: int = Field( default=64, ge=0, description="Last n tokens to keep for repeat penalty calculation.", ) logits_all: bool = Field(default=True, description="Whether to return logits.") cache: bool = Field( default=False, description="Use a cache to reduce processing times for evaluated prompts.", ) cache_type: Literal["ram", "disk"] = Field( default="ram", description="The type of cache to use. Only used if cache is True.", ) cache_size: int = Field( default=2 << 30, description="The size of the cache in bytes. Only used if cache is True.", ) vocab_only: bool = Field( default=False, description="Whether to only return the vocabulary." ) verbose: bool = Field( default=True, description="Whether to print debug information." ) host: str = Field(default="localhost", description="Listen address") port: int = Field(default=8000, description="Listen port") interrupt_requests: bool = Field( default=True, description="Whether to interrupt requests when a new request is received.", ) class ErrorResponse(TypedDict): """OpenAI style error response""" message: str type: str param: Optional[str] code: Optional[str] class ErrorResponseFormatters: """Collection of formatters for error responses. Args: request (Union[CreateCompletionRequest, CreateChatCompletionRequest]): Request body match (Match[str]): Match object from regex pattern Returns: Tuple[int, ErrorResponse]: Status code and error response """ @staticmethod def context_length_exceeded( request: Union[ "CreateCompletionRequest", "CreateChatCompletionRequest" ], match, # type: Match[str] # type: ignore ) -> Tuple[int, ErrorResponse]: """Formatter for context length exceeded error""" context_window = int(match.group(2)) prompt_tokens = int(match.group(1)) completion_tokens = request.max_tokens if hasattr(request, "messages"): # Chat completion message = ( "This model's maximum context length is {} tokens. " "However, you requested {} tokens " "({} in the messages, {} in the completion). " "Please reduce the length of the messages or completion." ) else: # Text completion message = ( "This model's maximum context length is {} tokens, " "however you requested {} tokens " "({} in your prompt; {} for the completion). " "Please reduce your prompt; or completion length." ) return 400, ErrorResponse( message=message.format( context_window, completion_tokens + prompt_tokens, prompt_tokens, completion_tokens, ), type="invalid_request_error", param="messages", code="context_length_exceeded", ) @staticmethod def model_not_found( request: Union[ "CreateCompletionRequest", "CreateChatCompletionRequest" ], match # type: Match[str] # type: ignore ) -> Tuple[int, ErrorResponse]: """Formatter for model_not_found error""" model_path = str(match.group(1)) message = f"The model `{model_path}` does not exist" return 400, ErrorResponse( message=message, type="invalid_request_error", param=None, code="model_not_found", ) class RouteErrorHandler(APIRoute): """Custom APIRoute that handles application errors and exceptions""" # key: regex pattern for original error message from llama_cpp # value: formatter function pattern_and_formatters: Dict[ "Pattern", Callable[ [ Union["CreateCompletionRequest", "CreateChatCompletionRequest"], "Match[str]", ], Tuple[int, ErrorResponse], ], ] = { compile( r"Requested tokens \((\d+)\) exceed context window of (\d+)" ): ErrorResponseFormatters.context_length_exceeded, compile( r"Model path does not exist: (.+)" ): ErrorResponseFormatters.model_not_found, } def error_message_wrapper( self, error: Exception, body: Optional[ Union[ "CreateChatCompletionRequest", "CreateCompletionRequest", "CreateEmbeddingRequest", ] ] = None, ) -> Tuple[int, ErrorResponse]: """Wraps error message in OpenAI style error response""" if body is not None and isinstance( body, ( CreateCompletionRequest, CreateChatCompletionRequest, ), ): # When text completion or chat completion for pattern, callback in self.pattern_and_formatters.items(): match = pattern.search(str(error)) if match is not None: return callback(body, match) # Wrap other errors as internal server error return 500, ErrorResponse( message=str(error), type="internal_server_error", param=None, code=None, ) def get_route_handler( self, ) -> Callable[[Request], Coroutine[None, None, Response]]: """Defines custom route handler that catches exceptions and formats in OpenAI style error response""" original_route_handler = super().get_route_handler() async def custom_route_handler(request: Request) -> Response: try: return await original_route_handler(request) except Exception as exc: json_body = await request.json() try: if "messages" in json_body: # Chat completion body: Optional[ Union[ CreateChatCompletionRequest, CreateCompletionRequest, CreateEmbeddingRequest, ] ] = CreateChatCompletionRequest(**json_body) elif "prompt" in json_body: # Text completion body = CreateCompletionRequest(**json_body) else: # Embedding body = CreateEmbeddingRequest(**json_body) except Exception: # Invalid request body body = None # Get proper error message from the exception ( status_code, error_message, ) = self.error_message_wrapper(error=exc, body=body) return JSONResponse( {"error": error_message}, status_code=status_code, ) return custom_route_handler router = APIRouter(route_class=RouteErrorHandler) settings: Optional[Settings] = None llama: Optional[llama_cpp.Llama] = None def create_app(settings: Optional[Settings] = None): if settings is None: settings = Settings() app = FastAPI( title="🦙 llama.cpp Python API", version="0.0.1", ) app.add_middleware( CORSMiddleware, allow_origins=["*"], allow_credentials=True, allow_methods=["*"], allow_headers=["*"], ) app.include_router(router) global llama llama = llama_cpp.Llama( model_path=settings.model, n_gpu_layers=settings.n_gpu_layers, tensor_split=settings.tensor_split, rope_freq_base=settings.rope_freq_base, rope_freq_scale=settings.rope_freq_scale, seed=settings.seed, f16_kv=settings.f16_kv, use_mlock=settings.use_mlock, use_mmap=settings.use_mmap, embedding=settings.embedding, logits_all=settings.logits_all, n_threads=settings.n_threads, n_batch=settings.n_batch, n_ctx=settings.n_ctx, last_n_tokens_size=settings.last_n_tokens_size, vocab_only=settings.vocab_only, verbose=settings.verbose, ) if settings.cache: if settings.cache_type == "disk": if settings.verbose: print(f"Using disk cache with size {settings.cache_size}") cache = llama_cpp.LlamaDiskCache(capacity_bytes=settings.cache_size) else: if settings.verbose: print(f"Using ram cache with size {settings.cache_size}") cache = llama_cpp.LlamaRAMCache(capacity_bytes=settings.cache_size) cache = llama_cpp.LlamaCache(capacity_bytes=settings.cache_size) llama.set_cache(cache) def set_settings(_settings: Settings): global settings settings = _settings set_settings(settings) return app llama_outer_lock = Lock() llama_inner_lock = Lock() def get_llama(): # NOTE: This double lock allows the currently streaming llama model to # check if any other requests are pending in the same thread and cancel # the stream if so. llama_outer_lock.acquire() release_outer_lock = True try: llama_inner_lock.acquire() try: llama_outer_lock.release() release_outer_lock = False yield llama finally: llama_inner_lock.release() finally: if release_outer_lock: llama_outer_lock.release() def get_settings(): yield settings async def get_event_publisher( request: Request, inner_send_chan: MemoryObjectSendStream, iterator: Iterator, ): async with inner_send_chan: try: async for chunk in iterate_in_threadpool(iterator): await inner_send_chan.send(dict(data=json.dumps(chunk))) if await request.is_disconnected(): raise anyio.get_cancelled_exc_class()() if settings.interrupt_requests and llama_outer_lock.locked(): await inner_send_chan.send(dict(data="[DONE]")) raise anyio.get_cancelled_exc_class()() await inner_send_chan.send(dict(data="[DONE]")) except anyio.get_cancelled_exc_class() as e: print("disconnected") with anyio.move_on_after(1, shield=True): print( f"Disconnected from client (via refresh/close) {request.client}" ) raise e model_field = Field(description="The model to use for generating completions.", default=None) max_tokens_field = Field( default=16, ge=1, description="The maximum number of tokens to generate." ) temperature_field = Field( default=0.8, ge=0.0, le=2.0, description="Adjust the randomness of the generated text.\n\n" + "Temperature is a hyperparameter that controls the randomness of the generated text. It affects the probability distribution of the model's output tokens. A higher temperature (e.g., 1.5) makes the output more random and creative, while a lower temperature (e.g., 0.5) makes the output more focused, deterministic, and conservative. The default value is 0.8, which provides a balance between randomness and determinism. At the extreme, a temperature of 0 will always pick the most likely next token, leading to identical outputs in each run.", ) top_p_field = Field( default=0.95, ge=0.0, le=1.0, description="Limit the next token selection to a subset of tokens with a cumulative probability above a threshold P.\n\n" + "Top-p sampling, also known as nucleus sampling, is another text generation method that selects the next token from a subset of tokens that together have a cumulative probability of at least p. This method provides a balance between diversity and quality by considering both the probabilities of tokens and the number of tokens to sample from. A higher value for top_p (e.g., 0.95) will lead to more diverse text, while a lower value (e.g., 0.5) will generate more focused and conservative text.", ) stop_field = Field( default=None, description="A list of tokens at which to stop generation. If None, no stop tokens are used.", ) stream_field = Field( default=False, description="Whether to stream the results as they are generated. Useful for chatbots.", ) top_k_field = Field( default=40, ge=0, description="Limit the next token selection to the K most probable tokens.\n\n" + "Top-k sampling is a text generation method that selects the next token only from the top k most likely tokens predicted by the model. It helps reduce the risk of generating low-probability or nonsensical tokens, but it may also limit the diversity of the output. A higher value for top_k (e.g., 100) will consider more tokens and lead to more diverse text, while a lower value (e.g., 10) will focus on the most probable tokens and generate more conservative text.", ) repeat_penalty_field = Field( default=1.1, ge=0.0, description="A penalty applied to each token that is already generated. This helps prevent the model from repeating itself.\n\n" + "Repeat penalty is a hyperparameter used to penalize the repetition of token sequences during text generation. It helps prevent the model from generating repetitive or monotonous text. A higher value (e.g., 1.5) will penalize repetitions more strongly, while a lower value (e.g., 0.9) will be more lenient.", ) presence_penalty_field = Field( default=0.0, ge=-2.0, le=2.0, description="Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.", ) frequency_penalty_field = Field( default=0.0, ge=-2.0, le=2.0, description="Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.", ) mirostat_mode_field = Field( default=0, ge=0, le=2, description="Enable Mirostat constant-perplexity algorithm of the specified version (1 or 2; 0 = disabled)", ) mirostat_tau_field = Field( default=5.0, ge=0.0, le=10.0, description="Mirostat target entropy, i.e. the target perplexity - lower values produce focused and coherent text, larger values produce more diverse and less coherent text", ) mirostat_eta_field = Field( default=0.1, ge=0.001, le=1.0, description="Mirostat learning rate" ) class CreateCompletionRequest(BaseModel): prompt: Union[str, List[str]] = Field( default="", description="The prompt to generate completions for." ) suffix: Optional[str] = Field( default=None, description="A suffix to append to the generated text. If None, no suffix is appended. Useful for chatbots.", ) max_tokens: int = max_tokens_field temperature: float = temperature_field top_p: float = top_p_field mirostat_mode: int = mirostat_mode_field mirostat_tau: float = mirostat_tau_field mirostat_eta: float = mirostat_eta_field echo: bool = Field( default=False, description="Whether to echo the prompt in the generated text. Useful for chatbots.", ) stop: Optional[Union[str, List[str]]] = stop_field stream: bool = stream_field logprobs: Optional[int] = Field( default=None, ge=0, description="The number of logprobs to generate. If None, no logprobs are generated.", ) presence_penalty: Optional[float] = presence_penalty_field frequency_penalty: Optional[float] = frequency_penalty_field logit_bias: Optional[Dict[str, float]] = Field(None) logprobs: Optional[int] = Field(None) # ignored or currently unsupported model: Optional[str] = model_field n: Optional[int] = 1 best_of: Optional[int] = 1 user: Optional[str] = Field(default=None) # llama.cpp specific parameters top_k: int = top_k_field repeat_penalty: float = repeat_penalty_field logit_bias_type: Optional[Literal["input_ids", "tokens"]] = Field(None) model_config = { "json_schema_extra": { "examples": [ { "prompt": "\n\n### Instructions:\nWhat is the capital of France?\n\n### Response:\n", "stop": ["\n", "###"], } ] } } def make_logit_bias_processor( llama: llama_cpp.Llama, logit_bias: Dict[str, float], logit_bias_type: Optional[Literal["input_ids", "tokens"]], ): if logit_bias_type is None: logit_bias_type = "input_ids" to_bias: Dict[int, float] = {} if logit_bias_type == "input_ids": for input_id, score in logit_bias.items(): input_id = int(input_id) to_bias[input_id] = score elif logit_bias_type == "tokens": for token, score in logit_bias.items(): token = token.encode("utf-8") for input_id in llama.tokenize(token, add_bos=False): to_bias[input_id] = score def logit_bias_processor( input_ids: List[int], scores: List[float], ) -> List[float]: new_scores = [None] * len(scores) for input_id, score in enumerate(scores): new_scores[input_id] = score + to_bias.get(input_id, 0.0) return new_scores return logit_bias_processor @router.post( "/v1/completions", ) async def create_completion( request: Request, body: CreateCompletionRequest, llama: llama_cpp.Llama = Depends(get_llama), ) -> llama_cpp.Completion: if isinstance(body.prompt, list): assert len(body.prompt) <= 1 body.prompt = body.prompt[0] if len(body.prompt) > 0 else "" exclude = { "n", "best_of", "logit_bias", "logit_bias_type", "user", } kwargs = body.model_dump(exclude=exclude) if body.logit_bias is not None: kwargs['logits_processor'] = llama_cpp.LogitsProcessorList([ make_logit_bias_processor(llama, body.logit_bias, body.logit_bias_type), ]) iterator_or_completion: Union[llama_cpp.Completion, Iterator[ llama_cpp.CompletionChunk ]] = await run_in_threadpool(llama, **kwargs) if isinstance(iterator_or_completion, Iterator): # EAFP: It's easier to ask for forgiveness than permission first_response = await run_in_threadpool(next, iterator_or_completion) # If no exception was raised from first_response, we can assume that # the iterator is valid and we can use it to stream the response. def iterator() -> Iterator[llama_cpp.CompletionChunk]: yield first_response yield from iterator_or_completion send_chan, recv_chan = anyio.create_memory_object_stream(10) return EventSourceResponse( recv_chan, data_sender_callable=partial( # type: ignore get_event_publisher, request=request, inner_send_chan=send_chan, iterator=iterator(), ) ) else: return iterator_or_completion class CreateEmbeddingRequest(BaseModel): model: Optional[str] = model_field input: Union[str, List[str]] = Field(description="The input to embed.") user: Optional[str] = Field(default=None) model_config = { "json_schema_extra": { "examples": [ { "input": "The food was delicious and the waiter...", } ] } } @router.post( "/v1/embeddings", ) async def create_embedding( request: CreateEmbeddingRequest, llama: llama_cpp.Llama = Depends(get_llama) ): return await run_in_threadpool( llama.create_embedding, **request.model_dump(exclude={"user"}) ) class ChatCompletionRequestMessage(BaseModel): role: Literal["system", "user", "assistant"] = Field( default="user", description="The role of the message." ) content: str = Field(default="", description="The content of the message.") class CreateChatCompletionRequest(BaseModel): messages: List[ChatCompletionRequestMessage] = Field( default=[], description="A list of messages to generate completions for." ) max_tokens: int = max_tokens_field temperature: float = temperature_field top_p: float = top_p_field mirostat_mode: int = mirostat_mode_field mirostat_tau: float = mirostat_tau_field mirostat_eta: float = mirostat_eta_field stop: Optional[List[str]] = stop_field stream: bool = stream_field presence_penalty: Optional[float] = presence_penalty_field frequency_penalty: Optional[float] = frequency_penalty_field logit_bias: Optional[Dict[str, float]] = Field(None) # ignored or currently unsupported model: Optional[str] = model_field n: Optional[int] = 1 user: Optional[str] = Field(None) # llama.cpp specific parameters top_k: int = top_k_field repeat_penalty: float = repeat_penalty_field logit_bias_type: Optional[Literal["input_ids", "tokens"]] = Field(None) model_config = { "json_schema_extra": { "examples": [ { "messages": [ ChatCompletionRequestMessage( role="system", content="You are a helpful assistant." ).model_dump(), ChatCompletionRequestMessage( role="user", content="What is the capital of France?" ).model_dump(), ] } ] } } @router.post( "/v1/chat/completions", ) async def create_chat_completion( request: Request, body: CreateChatCompletionRequest, llama: llama_cpp.Llama = Depends(get_llama), settings: Settings = Depends(get_settings), ) -> llama_cpp.ChatCompletion: exclude = { "n", "logit_bias", "logit_bias_type", "user", } kwargs = body.model_dump(exclude=exclude) if body.logit_bias is not None: kwargs['logits_processor'] = llama_cpp.LogitsProcessorList([ make_logit_bias_processor(llama, body.logit_bias, body.logit_bias_type), ]) iterator_or_completion: Union[llama_cpp.ChatCompletion, Iterator[ llama_cpp.ChatCompletionChunk ]] = await run_in_threadpool(llama.create_chat_completion, **kwargs) if isinstance(iterator_or_completion, Iterator): # EAFP: It's easier to ask for forgiveness than permission first_response = await run_in_threadpool(next, iterator_or_completion) # If no exception was raised from first_response, we can assume that # the iterator is valid and we can use it to stream the response. def iterator() -> Iterator[llama_cpp.ChatCompletionChunk]: yield first_response yield from iterator_or_completion send_chan, recv_chan = anyio.create_memory_object_stream(10) return EventSourceResponse( recv_chan, data_sender_callable=partial( # type: ignore get_event_publisher, request=request, inner_send_chan=send_chan, iterator=iterator(), ) ) else: return iterator_or_completion class ModelData(TypedDict): id: str object: Literal["model"] owned_by: str permissions: List[str] class ModelList(TypedDict): object: Literal["list"] data: List[ModelData] @router.get("/v1/models") async def get_models( settings: Settings = Depends(get_settings), ) -> ModelList: assert llama is not None return { "object": "list", "data": [ { "id": settings.model_alias if settings.model_alias is not None else llama.model_path, "object": "model", "owned_by": "me", "permissions": [], } ], }