Uploaded finetuned model

  • Developed by: csdnkey
  • License: apache-2.0
  • Finetuned from model : unsloth/qwen2-vl-7b-instruct-unsloth-bnb-4bit

This qwen2_vl model was trained 2x faster with Unsloth and Huggingface's TRL library.

Model description:

  1. It is fine-tuned based on qwen2.5-VL-7B through OZON product card rules, and only the language module is fine-tuned.
  2. The purpose is to optimize the product title, comply with the OZON product card title rules, and generate titles that are friendly to OZON search engines and buyers.
  3. Prompt word example: Optimize the product title according to OZON rules and the provided product categories, and output it as a Russian title. Category: [Jewelry], Brand: [BeTop] Original title: "Exotic retro fashion multi-layer tassel anklet for women, metal chain instep chain, cos court style foot ornaments"
Downloads last month
-
Safetensors
Model size
2B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Dataset used to train csdnkey/Qwen2-VL-7B-Instruct-OZON-0319