Model Overview

jenny08311/affine-test-4 is a 32 billion parameter language model, developed by jenny08311, that was created through a merge of pre-trained models. This model utilizes the TIES merge method and is built upon the Qwen/Qwen3-32B base architecture.

Merge Details

This model integrates components from two specific models:

• gurand/Affine-5CFL2YaBrJZCUSPBTjcDcTUSbnrm3UtAgKRsTU2KRcu9nvyR
• gurand/Affine-5CrMoVRmR8yP69Kh4iyrELehGYzUh3t7Q9hYVZUSjJA3VqDV
  The merging process involved specific parameter configurations for density and weight across MLP and self-attention layers, indicating a tailored approach to combine the strengths of the merged models. The configuration specifies bfloat16 as the dtype and includes int8_mask and normalize parameters.

Use Cases

Given its foundation on Qwen3-32B and the TIES merge method, this model is suitable for a broad range of general-purpose language generation and understanding tasks. Its 32B parameter count and 32768 token context length suggest capabilities for handling complex prompts and generating detailed responses.