diff options
Diffstat (limited to 'g4f/Provider/HuggingFace.py')
| -rw-r--r-- | g4f/Provider/HuggingFace.py | 37 |
1 files changed, 12 insertions, 25 deletions
diff --git a/g4f/Provider/HuggingFace.py b/g4f/Provider/HuggingFace.py index a3741196..586e5f5f 100644 --- a/g4f/Provider/HuggingFace.py +++ b/g4f/Provider/HuggingFace.py @@ -1,38 +1,24 @@ from __future__ import annotations + import json from aiohttp import ClientSession, BaseConnector + from ..typing import AsyncResult, Messages from .base_provider import AsyncGeneratorProvider, ProviderModelMixin from .helper import get_connector from ..errors import RateLimitError, ModelNotFoundError from ..requests.raise_for_status import raise_for_status + +from .HuggingChat import HuggingChat + class HuggingFace(AsyncGeneratorProvider, ProviderModelMixin): url = "https://huggingface.co/chat" working = True needs_auth = True supports_message_history = True - default_model = "meta-llama/Meta-Llama-3.1-70B-Instruct" - models = [ - 'meta-llama/Meta-Llama-3.1-70B-Instruct', - 'meta-llama/Meta-Llama-3.1-405B-Instruct-FP8', - 'CohereForAI/c4ai-command-r-plus', - 'mistralai/Mixtral-8x7B-Instruct-v0.1', - 'NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO', - '01-ai/Yi-1.5-34B-Chat', - 'mistralai/Mistral-7B-Instruct-v0.3', - 'microsoft/Phi-3-mini-4k-instruct', - ] - - model_aliases = { - "llama-3.1-70b": "meta-llama/Meta-Llama-3.1-70B-Instruct", - "llama-3.1-405b": "meta-llama/Meta-Llama-3.1-405B-Instruct-FP8", - "command-r-plus": "CohereForAI/c4ai-command-r-plus", - "mixtral-8x7b": "mistralai/Mixtral-8x7B-Instruct-v0.1", - "mixtral-8x7b": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", - "yi-1.5-34b": "01-ai/Yi-1.5-34B-Chat", - "mistral-7b": "mistralai/Mistral-7B-Instruct-v0.3", - "phi-3-mini-4k": "microsoft/Phi-3-mini-4k-instruct", - } + default_model = HuggingChat.default_model + models = HuggingChat.models + model_aliases = HuggingChat.model_aliases @classmethod def get_model(cls, model: str) -> str: @@ -76,7 +62,7 @@ class HuggingFace(AsyncGeneratorProvider, ProviderModelMixin): } if api_key is not None: headers["Authorization"] = f"Bearer {api_key}" - + params = { "return_full_text": False, "max_new_tokens": max_new_tokens, @@ -84,7 +70,7 @@ class HuggingFace(AsyncGeneratorProvider, ProviderModelMixin): **kwargs } payload = {"inputs": format_prompt(messages), "parameters": params, "stream": stream} - + async with ClientSession( headers=headers, connector=get_connector(connector, proxy) @@ -106,6 +92,7 @@ class HuggingFace(AsyncGeneratorProvider, ProviderModelMixin): yield chunk else: yield (await response.json())[0]["generated_text"].strip() + def format_prompt(messages: Messages) -> str: system_messages = [message["content"] for message in messages if message["role"] == "system"] question = " ".join([messages[-1]["content"], *system_messages]) @@ -114,4 +101,4 @@ def format_prompt(messages: Messages) -> str: for idx, message in enumerate(messages) if message["role"] == "assistant" ]) - return f"{history}<s>[INST] {question} [/INST]"
\ No newline at end of file + return f"{history}<s>[INST] {question} [/INST]" |