yamatazen/Qwen3-HereticLM-4B

TEXT GENERATIONConcurrent Unit Cost:1Model Size:4BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jul 11, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

yamatazen/Qwen3-HereticLM-4B is a 4 billion parameter language model created by yamatazen, built upon the Qwen3 architecture. This model is a merge of two pre-trained Qwen3-4B models, specifically designed for general language tasks. It utilizes the SLERP merge method to combine the strengths of its constituent models, offering a balanced performance for various applications.

Loading preview...

Model Overview

The yamatazen/Qwen3-HereticLM-4B is a 4 billion parameter language model developed by yamatazen, based on the Qwen3 architecture. It was created using the mergekit tool, specifically employing the SLERP merge method.

Merge Details

This model is a composite of two distinct Qwen3-4B models:

  • yamatazen/Qwen3-4B-HereticCPT
  • heretic-org/Qwen3-4B-Instruct-2507-heretic

The merging process used a t parameter of 0.5, indicating an equal weighting between the two base models. The configuration specified bfloat16 for both the merge and output data types, ensuring efficient processing.

Key Characteristics

  • Architecture: Qwen3-based.
  • Parameter Count: 4 billion parameters.
  • Merge Method: SLERP, combining two specialized Qwen3-4B models.
  • Purpose: Designed to leverage the combined capabilities of its merged components for general language understanding and generation tasks.

Intended Use

This model is suitable for developers looking for a 4B parameter model that integrates the characteristics of two different Qwen3-4B variants, potentially offering a more generalized or robust performance profile compared to its individual constituents.