chinhnc/Quyet-1.0-Large
chinhnc/Quyet-1.0-Large is a 31.3 billion parameter decision model developed by Chinh Nguyen, built on the Gemma-4-31B-it architecture with a LoRA fine-tune. It specializes in analyzing text, JSON, or conversations to make calibrated decisions by picking options for typed questions (choice, score, noul) and returning probabilities. This model is optimized for decision-making tasks, supporting inputs up to 6,000 tokens within an 8,000-token prompt, and is tuned for both English and Vietnamese.
Loading preview...
Quyet-1.0-Large: A Calibrated Decision Model
Quyet-1.0-Large, developed by Chinh Nguyen, is a 31.3 billion parameter model specifically designed for calibrated decision-making. Built upon the google/gemma-4-31B-it base model with a merged LoRA fine-tune, it processes various input states (text, JSON, conversations) and answers one or more typed questions by selecting options and providing confidence probabilities.
Key Capabilities
- Specialized Decision-Making: Unlike general-purpose LLMs, Quyet-1.0-Large is fine-tuned to act as a decision engine, providing structured outputs for specific questions.
- Typed Questions: Supports
choice(pick one label),score(ordered scale), andnoul(true/false) question types. - Calibrated Probabilities: Returns confidence scores and probabilities for each decision, enabling more nuanced applications.
- Multilingual Support: Primarily tuned for English and Vietnamese, with functional support for other languages.
- Efficient Input Handling: Processes states up to 6,000 tokens within an 8,000-token prompt, suitable for complex scenarios.
Good For
- Automated Customer Support: Classifying customer intent, urgency, or sentiment from messages.
- Structured Data Extraction: Extracting specific decisions or classifications from unstructured text or JSON.
- Business Process Automation: Automating decision points in workflows based on textual inputs.
- Risk Assessment: Evaluating conditions and making calibrated judgments based on provided information.