adpretko/armv8mac_to_x86_qwen25coder_3p0b_full

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:3.1BQuant:BF16Ctx Length:32kPublished:Mar 25, 2026License:otherArchitecture:Transformer Warm

The adpretko/armv8mac_to_x86_qwen25coder_3p0b_full model is a 3.1 billion parameter language model fine-tuned from Qwen/Qwen2.5-Coder-3B-Instruct. This model specializes in code generation and translation tasks, specifically focusing on converting armv8mac code to x86 architecture. It leverages a 32768 token context length to handle complex code structures and is optimized for developers working with cross-architecture code migration.

Loading preview...

Model Overview

The adpretko/armv8mac_to_x86_qwen25coder_3p0b_full is a specialized 3.1 billion parameter language model, fine-tuned from the robust Qwen/Qwen2.5-Coder-3B-Instruct base. Its primary function is to facilitate the conversion of code from the armv8mac architecture to x86.

Key Capabilities

  • Code Translation: Specifically trained on multiple armv8mac_to_x86 datasets (000 through 006) to perform cross-architecture code translation.
  • Code Generation: Inherits strong code generation capabilities from its Qwen2.5-Coder base.
  • Context Handling: Features a substantial 32768 token context window, enabling it to process and understand larger code blocks and complex programming logic during translation.

Training Details

The model was trained with a learning rate of 2e-05 over 0.5 epochs, utilizing a cosine learning rate scheduler with a 0.03 warmup ratio. It employed an AdamW optimizer and a total batch size of 8, distributed across multiple GPUs.

Intended Use Cases

This model is particularly well-suited for developers and engineers who need to migrate or adapt existing armv8mac codebase to x86 environments. Its fine-tuned nature suggests improved performance on these specific translation tasks compared to general-purpose code models.