GigaChat Model
AI-generatedOverview
This node integrates with the GigaChat language model API from Sberbank, allowing users to perform chat-based language model operations. It is useful for generating text completions, conversational AI, and other natural language processing tasks using the specified GigaChat model. Users can select the model and configure parameters such as maximum tokens, temperature, top-p sampling, and repetition penalty to control the behavior and creativity of the generated responses.
Use Case Examples
- Generating conversational responses in a chatbot application.
- Creating text completions for content generation.
- Experimenting with different model parameters to fine-tune output style and length.
Properties
| Name | Meaning |
|---|---|
| Имя модели | Specifies which GigaChat model to use for the chat operation. The options are dynamically loaded from the API. |
| Макс. токенов | Sets the maximum number of tokens the model can generate in the response. |
| Температура | Controls the randomness of the model's output. Higher values produce more random results. |
| Top P | Controls nucleus sampling probability to limit token selection to a subset with cumulative probability top_p. |
| Repetition Penalty | Penalizes repeated tokens to reduce repetition in the generated text. |
Output
JSON
model- The name of the GigaChat model used for the request.response- The generated text or chat response from the GigaChat model.
Dependencies
- Requires an API key credential for GigaChat API access, provided via an authorization bearer token.
Troubleshooting
- Common issues include invalid or missing API credentials, resulting in authorization errors.
- Incorrect model names or unavailable models may cause request failures.
- Setting extreme values for parameters like temperature or max tokens may lead to unexpected or poor-quality outputs.
Links
- GigaChat Node Disclaimers - Important disclaimers and usage notes for the unofficial GigaChat node.