Added new providers, updated existing ones, and added new models and updated model lists

[kqlio67]

New providers added

• Added Allyfy provider with gpt3 model support
• Added GeminiProChat provider with gemini-pro model support
• Added Marsyoo provider with gpt-4o model support
• Added LiteIcoding provider with gpt-4o, gpt-4-turbo, claude-3-opus, claude-3-5-sonnet, gemini-pro models support
• Added FreeNetfly provider with gpt-3.5-turbo, gpt-4o-mini
• Added TeachAnything provider with llama-3-70b-instruct
• Added MagickPenAsk provider with gpt-4o
• Added MagickPenChat provider with gpt-4o-mini

Supported Functions Not Yet Implemented in Providers

Some providers have additional capabilities that are not yet integrated into their main implementation:

• Marsyoo:

  • Image upload support

• LiteIcoding:

  • Plugin support:
    • Browse Web
    • Google Search
    • DALL-E
    • Bing Search
    • Real Time
    • Script Execute
  • File handling:
    • Document upload
    • Image upload

Provider Updates and Improvements

ChatGot (formerly GeminiProChat)

Enhanced the structure and functionality of the ChatGot class:

• Renamed the class from GeminiProChat to ChatGot for better clarity and consistency
• Updated the default_model attribute from an empty string to gemini-pro
• Kept all other functionality intact, including the create_async_generator method and generate_signature function

Chatgpt4Online

Enhanced the structure and functionality of the Chatgpt4Online class:

• Added support for GPT-4 model
• Simplified the API interaction by removing the need for wpnonce and context_id extraction
• Updated the headers to include more specific browser and request information
• Implemented a more robust streaming response handling mechanism
• Removed the WebDriver dependency, simplifying the overall setup
• Updated the class attributes to better reflect the provider's capabilities
• Integrated the format_prompt helper function for consistent message formatting
• Improved error handling and response parsing for more reliable operation

Note: The provider supports downloading documents and pictures. This functionality will be added to the provider's code in a future update.

HuggingChat + HuggingFace

Enhanced the structure and functionality of the HuggingChat class:

• Updated the list of supported models:
  • Added meta-llama/Meta-Llama-3.1-70B-Instruct
  • Added meta-llama/Meta-Llama-3.1-405B-Instruct-FP8
  • Removed meta-llama/Meta-Llama-3-70B-Instruct
• Maintained the existing functionality and structure of the class
• Kept the same implementation for API interactions and response handling

Liaobots

Enhanced the structure and functionality of the Liaobots class:

• Updated the models dictionary with new and modified entries:

  • Added gpt-4o-mini-free and gpt-4o-mini models
  • Updated gpt-4-0613 model specifications
  • Added gpt-4-turbo model

• Removed several models that are no longer supported:

  • gpt-3.5-turbo
  • claude-3-opus-20240229
  • claude-3-opus-20240229-aws
  • claude-3-opus-100k-poe
  • claude-3-sonnet-20240229
  • claude-3-haiku-20240307
  • claude-2.1
  • claude-2.0
  • All Gemini models

• Removed unused class attributes and methods:

  • Removed supports_system_message, supports_gpt_35_turbo, supports_gpt_4, default_model
  • Removed model_aliases dictionary
  • Removed _auth_code and _cookie_jar attributes
  • Removed create_async_generator, get_model, is_supported, initialize_auth_code, and ensure_auth_code methods

• Simplified the class structure to focus on core functionality

PerplexityLabs

Enhanced the structure and functionality of the PerplexityLabs class:

• Updated the list of supported models:

  • Removed several models including claude-3-haiku, codellama, mistral-7b, llava, dbrx-instruct, and others
  • Added new models: gemma-2-9b-it, gemma-2-27b-it, nemotron-4-340b-instruct
  • Retained some existing models like llama-3 variants and mixtral-8x7b-instruct

• Removed the model_aliases dictionary, simplifying model management

• Maintained existing functionality for API interactions and response handling

• Kept the default_model as mixtral-8x7b-instruct

Pi

Enhanced the structure and functionality of the Pi class:

• Added default_model attribute set to pi
• Maintained existing functionality for API interactions and response handling
• No other significant changes to the class structure or methods

ReplicateHome

Enhanced the structure and functionality of the ReplicateHome class:

• Updated the default_model to stability-ai/stable-diffusion-3

• Revised the list of supported models:

  • For image generation:

    • Added stability-ai/stable-diffusion-3
    • Added bytedance/sdxl-lightning-4step
    • Added playgroundai/playground-v2.5-1024px-aesthetic
    • Removed stability-ai/sdxl and ai-forever/kandinsky-2.2

  • For text generation:

    • Added meta/meta-llama-3-70b-instruct
    • Added mistralai/mixtral-8x7b-instruct-v0.1
    • Added google-deepmind/gemma-2b-it
    • Removed meta/llama-2-70b-chat and mistralai/mistral-7b-instruct-v0.2

• Updated the versions dictionary with new model versions

• Modified the image_models and text_models sets to reflect the new model list

• Maintained existing functionality for API interactions and response handling

You

Enhanced the structure and functionality of the You class:

• Updated the list of supported models:

  • Added new models: gpt-4o-mini, claude-3.5-sonnet, llama-3.1-70b, gemini-1-5-flash, gemini-1-0-pro, dolphin-2.5
  • Removed models: claude-instant, zephyr
  • Reordered the list for better organization

• Removed the model_aliases dictionary

• Maintained existing functionality for API interactions and response handling

• Kept the default_model as gpt-3.5-turbo and default_vision_model as "agent"

Openai

Enhanced the structure and functionality of the Openai class:

• Added default_model attribute with an empty string value
• Maintained existing functionality for API interactions and response handling
• No other significant changes to the class structure or methods

OpenaiChat

Enhanced the structure and functionality of the OpenaiChat class:

• Updated the models list:

  • Removed gpt-3.5-turbo
  • Added gpt-4o-mini

• Modified the model_aliases dictionary

[a-alex-l]

So cool thanks.