harlynmivra/Xing4.0-29B-A4B-Minutes

Xing4.0-29B-A4B-Minutes

Xing4.0-29B-A4B-Minutes is a supervised fine-tuned (SFT) language model based on Xing4.0-29B-A4B, designed for generating structured meeting minutes from conversational or meeting transcription text.

Model Description

The model is fine-tuned on datasets constructed from dialogue and meeting transcripts paired with corresponding summaries or meeting minutes.

• Training method: Supervised Fine-Tuning (SFT)
• Base model: XingChen-AGI/Xing4.0-29B-A4B
• Model type: Causal Language Model
• Language: Chinese
• Task: Summarization
• SFT context length: 8192

Intended Uses

The model is primarily intended for generating structured meeting minutes and summaries from conversational or meeting transcription text.

Given a meeting transcript, the model can generate the following types of information:

• Title: Generate a concise title that summarizes the main topic of the meeting.
• Keywords: Extract representative keywords or key topics from the meeting content.
• Abstract: Provide a concise summary covering the main discussion and overall conclusions.
• Key Points: Identify and organize the major discussion points, important information, and key conclusions.
• Action Items: Extract follow-up tasks, responsible parties, deadlines, or other actionable items when such information is explicitly mentioned in the transcript.

Typical input may include single/multi-speaker meeting transcripts, dialogue records, or other conversational text.

The model is intended to transform lengthy and potentially redundant transcription content into concise, structured, and readable meeting minutes.

Citation

XingChen-AGI/Xing4.0-29B-A4B