adpretko/riscv_to_armv8_qwen25coder_3p0b_full

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

The adpretko/riscv_to_armv8_qwen25coder_3p0b_full model is a 3.1 billion parameter language model, fine-tuned from Qwen/Qwen2.5-Coder-3B-Instruct. It specializes in code translation, specifically from RISC-V to ARMv8 architectures, leveraging a 32768-token context length. This model is designed for developers working on cross-architecture code migration and optimization tasks. Its primary strength lies in its specialized training on multiple riscv_to_armv8 datasets, making it highly effective for this specific code transformation.

Loading preview...

Model Overview

The adpretko/riscv_to_armv8_qwen25coder_3p0b_full is a specialized 3.1 billion parameter language model, fine-tuned from the robust Qwen2.5-Coder-3B-Instruct base. This model is engineered for a very specific and critical task: translating code from the RISC-V instruction set architecture to ARMv8. It leverages a substantial context window of 32768 tokens, enabling it to process and understand larger blocks of code during translation.

Key Capabilities

  • RISC-V to ARMv8 Code Translation: The model's core capability is its proficiency in converting RISC-V assembly or low-level code constructs into their ARMv8 equivalents. This is achieved through fine-tuning on a series of dedicated datasets (riscv_to_armv8_000 through riscv_to_armv8_006).
  • Large Context Window: With a 32K token context length, it can handle more complex code snippets and maintain better contextual understanding during translation tasks.

Training Details

The model was trained with a learning rate of 2e-05, a batch size of 1 (accumulated to 8), and utilized a cosine learning rate scheduler with a 0.03 warmup ratio over 0.5 epochs. The training was performed using Transformers 4.46.1 and Pytorch 2.5.1+cu121.

Ideal Use Cases

  • Cross-Architecture Development: Developers working on projects that require porting or optimizing code between RISC-V and ARMv8 platforms.
  • Embedded Systems: Useful for embedded systems development where code needs to be adapted for different processor architectures.
  • Research & Prototyping: Facilitates rapid prototyping and research into automated code translation for these specific architectures.