bitext/Mistral-7B-Restaurants

TEXT GENERATIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:7BQuant:FP8Context Size:4kTool Calling:SupportedPublished:Aug 15, 2024License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The bitext/Mistral-7B-Restaurants model is a 7 billion parameter language model, fine-tuned from mistralai/Mistral-7B-Instruct-v0.2, specifically optimized for the Restaurants domain. Developed by Bitext, it excels at answering questions and assisting users with restaurant-related procedures. This model is designed as a specialized component for building chatbots and virtual assistants focused on the restaurant industry, leveraging hybrid synthetic data for its training.

Loading preview...

Model Overview

The bitext/Mistral-7B-Restaurants is a 7 billion parameter language model developed by Bitext, fine-tuned from the mistralai/Mistral-7B-Instruct-v0.2 base model. Its primary purpose is to serve the Restaurants domain, providing accurate and fast answers to user queries related to restaurant services and procedures. This model demonstrates Bitext's approach to creating verticalized enterprise models from general-purpose LLMs.

Key Capabilities

  • Domain-Specific Expertise: Highly specialized in restaurant-related interactions, including inquiries about menus, reservations, online food ordering, and events.
  • Chatbot and Virtual Assistant Foundation: Designed as a foundational component for building customer support chatbots and virtual assistants for the restaurant industry.
  • Data-Driven Training: Fine-tuned on the Bitext Restaurants Dataset, which includes 30 distinct restaurant-related intents, each with approximately 1000 examples.
  • Robust Architecture: Utilizes the MistralForCausalLM architecture with a LlamaTokenizer, maintaining the foundational strengths of the base Mistral model.

Intended Use Cases

This model is specifically recommended for:

  • Customer Support: Providing automated customer support for restaurants.
  • Information Retrieval: Answering questions about restaurant menus, operating hours, and services.
  • Transaction Assistance: Guiding users through reservation processes or online food orders.

Limitations

  • Domain Specificity: Performance may be suboptimal for questions outside the restaurant domain.
  • Potential Biases: Users should critically evaluate responses due to potential biases inherited from the training data.

Finetunes
1 models