Models factory
The Models factory class is a flexible way to create and manage a model configuration with its initialization settings.
After defining it in a JSON file named 'models.json', you can create the model by simply referring to its name, for example:
The first line calls Models.setup() to initialize the Models factory with the folder where model files and configs ("models.json" and "formats.json") are located.
The second line calls Models.create() to create a model from the name "openai:gpt-4-development". In this case we created a remote model, but we could as well create a local model based in a GGUF file.
The names should be in the format "provider:model_name" and Sibila currently supports the following providers:
| Provider | Type | Creates object of type |
|---|---|---|
| llamacpp | Local GGUF model | LlamaCppModel |
| anthropic | Remote model | AnthropicModel |
| fireworks | Remote model | FireworksModel |
| groq | Remote model | GroqModel |
| mistral | Remote model | MistralModel |
| openai | Remote model | OpenAIModel |
| together | Remote model | TogetherModel |
The name part, after the "provider:", must either be:
- A remote model name, like "gpt-4": "openai:gpt-4"
- For llamacpp, a local model name, defined in a models.json file, like "openchat": "llamacpp:openchat"
- Also for llamacpp, name can be the actual filename of a model in the "models" folder: "llamacpp:openchat-3.5-1210.Q4_K_M.gguf".
Although you can use filenames as model names, it's generally a better idea, for continued use, to create an entry in the "models.json" file - this allows future model replacement to be much easier.
See Managing models to learn how to register these model names.