gpt4free

公开

关注 0 Fork 0 Star 1

XFEstudio/gpt4free

refactor: Refactor image generation parameters handling

- Updated the `PollinationsAI` class in `g4f/Provider/PollinationsAI.py`: - Changed `aspect_ratio` parameter handling to conditionally use default "1:1" if not specified. - Enhanced media handling by introducing `media` parameter in `_generate_image` method. - Updated parameter processing in `_generate_image_async` method for `model == "gptimage"`. - Updated `Api` class in `g4f/api/__init__.py`: - Simplified handling of `credentials` for `config.api_key`. - Updated `Images` class in `g4f/client/__init__.py`: - Added `download_media` parameter to `_process_image_response` method. - Enhanced `_process_image_response` method to conditionally download media based on `download_media` flag. - Updated `_process_image_response` method in `Images` class in `g4f/client/__init__.py`: - Enhanced handling of media response based on `download_media` flag. - Updated `is_valid_media` function in `g4f/image/__init__.py`: - Added typing annotations for clarity. - Updated `AnyProvider` class in `g4f/providers/any_provider.py`: - Improved handling of `api_key` dictionary to set `extra_body["api_key"]`. - Updated `IterListProvider` class in `g4f/providers/retry_provider.py`: - Enhanced handling of `model` and `api_key` parameters. - Updated `BaseProvider` class in `g4f/providers/types.py`: - Added `create_function` and `async_create_function` methods. - Updated `BaseRetryProvider` class in `g4f/providers/types.py`: - Enhanced handling of `model` and `api_key` parameters in provider iteration.

963c3a58

hlohaus <983577+hlohaus@users.noreply.github.com>

提交于 2025-06-12 14:23

父提交 2befef9

代码差异

7 个文件 +93 -26

Modified g4f/Provider/PollinationsAI.py +20 -8

@@ -252,7 +252,7 @@ class PollinationsAI(AsyncGeneratorProvider, ProviderModelMixin):

252 252 extra_body: dict = None,

253 253 # Image generation parameters

254 254 prompt: str = None,

255 aspect_ratio: str = "1:1",

255 aspect_ratio: str = None,

256 256 width: int = None,

257 257 height: int = None,

258 258 seed: Optional[int] = None,

@@ -294,6 +294,7 @@ class PollinationsAI(AsyncGeneratorProvider, ProviderModelMixin):

294 294 async for chunk in cls._generate_image(

295 295 model=model,

296 296 prompt=format_media_prompt(messages, prompt),

297 media=media,

297 298 proxy=proxy,

298 299 aspect_ratio=aspect_ratio,

299 300 width=width,

@@ -347,6 +348,7 @@ class PollinationsAI(AsyncGeneratorProvider, ProviderModelMixin):

347 348 cls,

348 349 model: str,

349 350 prompt: str,

351 media: MediaListType,

350 352 proxy: str,

351 353 aspect_ratio: str,

352 354 width: int,

@@ -362,20 +364,30 @@ class PollinationsAI(AsyncGeneratorProvider, ProviderModelMixin):

362 364 api_key: str,

363 365 timeout: int = 120

364 366 ) -> AsyncResult:

365 if model == "gptimage":

366 n = 1

367 params = use_aspect_ratio({

368 "width": width,

369 "height": height,

367 params = {

370 368 "model": model,

371 369 "nologo": str(nologo).lower(),

372 370 "private": str(private).lower(),

373 371 "enhance": str(enhance).lower(),

374 372 "safe": str(safe).lower(),

375 }, aspect_ratio)

373 }

374 if model == "gptimage":

375 n = 1

376 # Only remote images are supported

377 image = [item[0] for item in media if isinstance(item[0], str) and item[0].startswith("http")]

378 params = {

379 **params,

380 "image": ",".join(image) if image else "",

381 }

382 else:

383 params = use_aspect_ratio({

384 "width": width,

385 "height": height,

386 **params

387 }, "1:1" if aspect_ratio is None else aspect_ratio)

376 388 query = "&".join(f"{k}={quote_plus(str(v))}" for k, v in params.items() if v is not None)

377 389 encoded_prompt = prompt

378 if model == "gptimage" and aspect_ratio != "1:1":

390 if model == "gptimage" and aspect_ratio is not None:

379 391 encoded_prompt = f"{encoded_prompt} aspect-ratio: {aspect_ratio}"

380 392 encoded_prompt = quote_plus(encoded_prompt)[:2048-len(cls.image_api_endpoint)-len(query)-8]

381 393 url = f"{cls.image_api_endpoint}prompt/{encoded_prompt}?{query}"

Modified g4f/api/__init__.py +2 -4

@@ -425,7 +425,7 @@ class Api:

425 425 try:

426 426 if config.provider is None:

427 427 config.provider = AppConfig.provider if provider is None else provider

428 if credentials is not None and credentials.credentials != "secret":

428 if config.api_key is None and credentials is not None and credentials.credentials != "secret":

429 429 config.api_key = credentials.credentials

430 430

431 431 conversation = None

@@ -618,9 +618,7 @@ class Api:

618 618 provider=config.provider if provider is None else provider,

619 619 prompt=config.input,

620 620 audio=filter_none(voice=config.voice, format=config.response_format, language=config.language),

621 **filter_none(

622 api_key=api_key,

623 )

621 api_key=api_key,

624 622 )

625 623 if isinstance(response.choices[0].message.content, AudioResponse):

626 624 response = response.choices[0].message.content.data

Modified g4f/client/__init__.py +14 -4

@@ -479,6 +479,7 @@ class Images:

479 479 response,

480 480 model,

481 481 provider_name,

482 kwargs.get("download_media", True),

482 483 response_format,

483 484 proxy

484 485 )

@@ -531,7 +532,7 @@ class Images:

531 532 urls.extend(item.urls)

532 533 if not urls:

533 534 return None

534 alt = getattr(items[0], "alt", items[0].options.get("text"))

535 alt = getattr(items[0], "alt", "")

535 536 return MediaResponse(urls, alt, items[0].options)

536 537

537 538 def create_variation(

@@ -580,13 +581,21 @@ class Images:

580 581 if error is not None:

581 582 raise error

582 583 raise NoMediaResponseError(f"No media response from {provider_name}")

583 return await self._process_image_response(response, model, provider_name, response_format, proxy)

584 return await self._process_image_response(

585 response,

586 model,

587 provider_name,

588 kwargs.get("download_media", True),

589 response_format,

590 proxy

591 )

584 592

585 593 async def _process_image_response(

586 594 self,

587 595 response: MediaResponse,

588 596 model: str,

589 597 provider: str,

598 download_media: bool,

590 599 response_format: Optional[str] = None,

591 600 proxy: str = None

592 601 ) -> ImagesResponse:

@@ -609,9 +618,10 @@ class Images:

609 618 images = await asyncio.gather(*[get_b64_from_url(image) for image in response.get_list()])

610 619 else:

611 620 # Save locally for None (default) case

612

            images = await copy_media(response.get_list(), response.get("cookies"), response.get("headers"), proxy, response.alt)

621 if download_media or response.get("cookies"):

622

                images = await copy_media(response.get_list(), response.get("cookies"), response.get("headers"), proxy, response.alt)

613 623 images = [Image.model_construct(url=image, revised_prompt=response.alt) for image in images]

614

624

615 625 return ImagesResponse.model_construct(

616 626 created=int(time.time()),

617 627 data=images,

Modified g4f/image/__init__.py +2 -2

@@ -14,7 +14,7 @@ try:

14 14 except ImportError:

15 15 has_requirements = False

16 16

17 from ..typing import ImageType, Union, Image

17 from ..typing import ImageType, Image

18 18 from ..errors import MissingRequirementsError

19 19

20 20 EXTENSIONS_MAP: dict[str, str] = {

@@ -107,7 +107,7 @@ def is_data_an_media(data, filename: str = None) -> str:

107 107 return is_accepted_format(data)

108 108 return is_data_uri_an_image(data)

109 109

110 def is_valid_media(data, filename: str = None) -> str:

110 def is_valid_media(data: ImageType = None, filename: str = None) -> str:

111 111 if is_valid_audio(data, filename):

112 112 return True

113 113 if filename:

Modified g4f/providers/any_provider.py +3 -2

@@ -319,8 +319,9 @@ class AnyProvider(AsyncGeneratorProvider, ProviderModelMixin):

319 319 extra_providers = []

320 320 if isinstance(api_key, dict):

321 321 for provider in api_key:

322 if provider in __map__ and __map__[provider] not in MAIN_PROVIERS:

323 extra_providers.append(__map__[provider])

322 if api_key.get(provider):

323 if provider in __map__ and __map__[provider] not in MAIN_PROVIERS:

324 extra_providers.append(__map__[provider])

324 325 for provider in MAIN_PROVIERS + extra_providers:

325 326 if provider.working:

326 327 if not model or model in provider.get_models() or model in provider.model_aliases:

Modified g4f/providers/retry_provider.py +16 -6

@@ -53,11 +53,16 @@ class IterListProvider(BaseRetryProvider):

53 53

54 54 for provider in self.get_providers(stream and not ignore_stream, ignored):

55 55 self.last_provider = provider

56 debug.log(f"Using {provider.__name__} provider")

57 yield ProviderInfo(**provider.get_dict(), model=model if model else getattr(provider, "default_model"))

56 if not model:

57 model = getattr(provider, "default_model", None)

58 model = provider.model_aliases.get(model, model) if hasattr(provider, "model_aliases") else model

59 debug.log(f"Using {provider.__name__} provider with model {model}")

60 yield ProviderInfo(**provider.get_dict(), model=model)

58 61 extra_body = kwargs.copy()

59 62 if isinstance(api_key, dict):

60 extra_body["api_key"] = api_key.get(provider.get_parent())

63 api_key = api_key.get(provider.get_parent())

64 if api_key:

65 extra_body["api_key"] = api_key

61 66 try:

62 67 response = provider.create_function(model, messages, stream=stream, **extra_body)

63 68 for chunk in response:

@@ -92,11 +97,16 @@ class IterListProvider(BaseRetryProvider):

92 97

93 98 for provider in self.get_providers(stream and not ignore_stream, ignored):

94 99 self.last_provider = provider

95 debug.log(f"Using {provider.__name__} provider" + (f" and {model} model" if model else ""))

96 yield ProviderInfo(**provider.get_dict(), model=model if model else getattr(provider, "default_model"))

100 if not model:

101 model = getattr(provider, "default_model", None)

102 model = provider.model_aliases.get(model, model) if hasattr(provider, "model_aliases") else model

103 debug.log(f"Using {provider.__name__} provider with model {model}")

104 yield ProviderInfo(**provider.get_dict(), model=model)

97 105 extra_body = kwargs.copy()

98 106 if isinstance(api_key, dict):

99 extra_body["api_key"] = api_key.get(provider.get_parent())

107 api_key = api_key.get(provider.get_parent())

108 if api_key:

109 extra_body["api_key"] = api_key

100 110 if conversation is not None and hasattr(conversation, provider.__name__):

101 111 extra_body["conversation"] = JsonConversation(**getattr(conversation, provider.__name__))

102 112 try:

Modified g4f/providers/types.py +36 -0

@@ -42,6 +42,42 @@ class BaseProvider(ABC):

42 42 def get_parent(cls) -> str:

43 43 return getattr(cls, "parent", cls.__name__)

44 44

45 @abstractmethod

46 def create_function(

47 *args,

48 **kwargs

49 ) -> CreateResult:

50 """

51 Create a function to generate a response based on the model and messages.

53 Args:

54 model (str): The model to use.

55 messages (Messages): The messages to process.

56 stream (bool): Whether to stream the response.

58 Returns:

59 CreateResult: The result of the creation.

60 """

61 raise NotImplementedError()

63 @staticmethod

64 def async_create_function(

65 *args,

66 **kwargs

67 ) -> CreateResult:

68 """

69 Asynchronously create a function to generate a response based on the model and messages.

71 Args:

72 model (str): The model to use.

73 messages (Messages): The messages to process.

74 stream (bool): Whether to stream the response.

76 Returns:

77 CreateResult: The result of the creation.

78 """

79 raise NotImplementedError()

45 81 class BaseRetryProvider(BaseProvider):

46 82 """

47 83 Base class for a provider that implements retry logic.