HenryJJ/dolphin-2.6-mistral-7b-dpo-orca-v3

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Jan 14, 2024License:apache-2.0Architecture:Transformer Open Weights Cold

HenryJJ/dolphin-2.6-mistral-7b-dpo-orca-v3 is a 7 billion parameter auto-regressive language model, fine-tuned by HenryJJ using DPO from cognitivecomputations/dolphin-2.6-mistral-7b. It was trained for 1200 steps on the Intel/orca_dpo_pairs dataset with a 1024 context window. This model is optimized for instruction following and conversational tasks, utilizing a ChatML prompt format.

Loading preview...

Model Overview

HenryJJ/dolphin-2.6-mistral-7b-dpo-orca-v3 is a 7 billion parameter English language model, fine-tuned by HenryJJ. It is based on the Llama 2 transformer architecture and was developed using Direct Preference Optimization (DPO) from the cognitivecomputations/dolphin-2.6-mistral-7b base model. The training utilized the Intel/orca_dpo_pairs dataset, conducted over 1200 steps with a 1024 token context window.

Key Capabilities

  • Instruction Following: Enhanced through DPO training on the Orca dataset, making it proficient in understanding and executing user instructions.
  • Conversational AI: Designed to be a helpful AI assistant, responding to user prompts effectively.
  • ChatML Format: Employs the ChatML prompt format, ensuring compatibility with systems expecting this structure, with <|im_end|> mapping to token_id 2 for EOS.

Good For

  • Applications requiring a responsive and compliant AI assistant.
  • Scenarios where clear instruction following is critical.
  • Developers working with ChatML-formatted prompts and seeking a 7B parameter model for English language tasks.