Azure99/blossom-v3_1-mistral-7b
Azure99/blossom-v3_1-mistral-7b is a 7 billion parameter conversational large language model, fine-tuned by Azure99 on a mixed dataset including Orca, Wizard, Chat, and Math data, based on the Mistral-7B-v0.1 architecture. It offers robust general capabilities and context comprehension, with a focus on conversational AI and mathematical reasoning. This model is particularly strong in English and Chinese, making it suitable for multi-turn dialogue and instruction-following tasks.
Loading preview...
Blossom-v3.1-mistral-7b Overview
Blossom-v3.1-mistral-7b is a 7 billion parameter conversational large language model developed by Azure99, built upon the Mistral-7B-v0.1 pre-trained model. It is specifically fine-tuned for robust general capabilities and strong context comprehension in conversational AI.
Key Capabilities
- Conversational AI: Excels in multi-turn dialogue and instruction-following, trained on a diverse dataset including Orca, Wizard, and Chat data.
- Mathematical Reasoning: Incorporates a dedicated 2K Blossom math reasoning dataset in its second training stage, enhancing its ability to handle mathematical queries.
- Multilingual Support: Benefits from high-quality Chinese and English datasets, making it effective for both languages, though a Baichuan2-based variant is recommended for purely Chinese scenarios.
- Efficient Training: Underwent a two-stage training process, starting with 100K single-turn instruction datasets (Wizard, Orca) for 1 epoch, followed by 3 epochs on a mix of math reasoning, multi-turn chat, and sampled data.
Good For
- Developing chatbots and conversational agents requiring strong general understanding.
- Applications needing robust context comprehension in dialogue.
- Tasks involving mathematical reasoning and problem-solving.
- Use cases requiring a balance of English and Chinese language capabilities in a conversational setting.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.