Alelcv27/Llama3.2-3B-DELLA-Math-Code

TEXT GENERATIONConcurrency Cost:1Model Size:3.2BQuant:BF16Ctx Length:32kPublished:Apr 19, 2026Architecture:Transformer Cold

Alelcv27/Llama3.2-3B-DELLA-Math-Code is a 3.2 billion parameter language model based on the Llama 3.2 architecture, created by Alelcv27. This model is a merge of specialized base models, utilizing the DELLA merge method to enhance its capabilities in mathematical reasoning and code generation. With a substantial 32768-token context length, it is optimized for tasks requiring strong performance in both numerical problem-solving and programming contexts.

Loading preview...

Overview

Alelcv27/Llama3.2-3B-DELLA-Math-Code is a 3.2 billion parameter language model built upon the Llama 3.2 architecture. Developed by Alelcv27, this model distinguishes itself through its unique construction: it is a merge of pre-trained language models specifically designed to excel in mathematical and coding domains. The merging process utilized the advanced DELLA merge method, combining specialized base models to create a unified model with enhanced capabilities in these areas.

Key Capabilities

  • Enhanced Mathematical Reasoning: The model integrates a base model focused on mathematical tasks, suggesting improved performance in numerical problem-solving and logical deduction related to mathematics.
  • Proficient Code Generation: By incorporating a base model trained for coding, it is expected to demonstrate strong capabilities in generating, understanding, and debugging code across various programming languages.
  • Extended Context Window: With a context length of 32768 tokens, it can process and generate longer sequences of text, which is particularly beneficial for complex coding projects or multi-step mathematical problems.

Good For

  • Mathematical Applications: Ideal for tasks requiring accurate calculations, formula generation, or solving mathematical word problems.
  • Software Development: Suitable for developers needing assistance with code generation, refactoring, or understanding complex codebases.
  • Educational Tools: Can be leveraged in tools designed to teach or assist with programming and mathematics due to its specialized training.

Finetunes
1 models