SD2 1045 α merge llm engine to the pp

[juandavidcruzgomez]

• Moving the base of LLM Engine inside PP. Some LLMEngine tests, for instantiation and completion calling
• Test embedding with mocked call
• Chat functions in llmengine
• Remove monitoring from edenai_apis to gain performance
• missing args in chat
• Update llm engine error mapping
• Chat/Multimodal in llm engine
• Update openai
• Update anthropic
• Update : Mistral/Google/Cohere/Meta
• [Fix] system prompt message
• add litellm models 'on the fly'
• LLM Tests (Embeddings is not workign in this commit)
• Use a pydantic model for representing the Litellm models
• Add togetherai with LLM Engine
• [Feat] update all text feature using llm engine
• update providers
• Clean providers
• Image & translation feature on llm engine
• bring back text generation for openai
• Clean providers
• change the name to avoid collisions
• Circular reference
• Update llm_engine imports
• Update llm_engine imports
• Update llm_engine imports
• Update the path for the clientsd
• Update the path for the clients
• Test a new file: models.json is not being included in the package
• Test a new file: models.json is not being included in the package
• Update the manifest duh
• Update the models
• test for llm engine
• Update manifest
• Update manifest
• [Fix] providers output
• Fix embeddings signature
• Fix XAI anonymization version
• Fix deepseek chat signature
• Fix sentiment analysis signature
• Fix mistral chat output
• Update manifest
• XAI info file update
• Migrate deepseek to llm engine
• [Fix] Stream response in llm engine
• Migrate replicate
• Tool tests
• Remove print
• [Fix] drop params in embeddings
• Add together_ai template
• Update anonymization dataclass to be more flexible on LLMs
• Stop cannot be a list but a string
• Stop cannot be a list but a string
• Stop cannot be a list but a string
• Stop cannot be a list but a string
• Stop cannot be a list but a string
• Stop cannot be a list but a string
• Add togetherai