Skip to content
Other

sva-qlora-all

QLoRA adapter for generating SystemVerilog Assertions from RTL

1 min readupdated

This is a QLoRA (quantized low-rank adaptation for fine-tuning language models) adapter built on top of the popular Qwen/Qwen2.5-Coder-7B-Instruct model. Trained on the full set of generated RTL-SVA pairs (the All variant), it automatically generates SystemVerilog Assertions (SVA) directly from Register Transfer Level (RTL) source code. Hardware developers and chip verification engineers can use it to speed up writing verification code.

Key features

  • Built on the Qwen2.5-Coder-7B-Instruct architecture
  • Uses QLoRA adaptation tailored to a specialized domain task
  • Trained on a dataset of RTL-SVA pairs
  • Focused on automating chip verification

What you can do

  • Use the adapter in SystemVerilog digital circuit verification projects
  • Automate SVA generation from existing RTL modules
  • Integrate the model into hardware-software development pipelines

Example

Pair the base Qwen/Qwen2.5-Coder-7B-Instruct model with this PEFT adapter to generate SVA from your RTL code, following the instructions and materials in the project repository.

Read moreDetails
Library: peft
Base model: Qwen/Qwen2.5-Coder-7B-Instruct
Training: QLoRA on all generated RTL-SVA pairs (All variant)
Source code and dataset: [https://anonymous.4open.science/r/veri2-CD05/](https://anonymous.4open.science/r/veri2-CD05/)
Discuss Telegram 0 3

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 15 GB · ready files17 GB
  • FP8 / INT8recommendedweights 7.7 GB9.0 GB
  • 4-bitweights 4.6 GB5.6 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 3080 · BF16 — from RTX 4090

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

This is a LoRA: counted together with its base model Qwen2.5-Coder-7B-Instruct.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # veri2-anon/sva-qlora-all
    # ===== Tokens: paste yours between the quotes =====
    export HF_TOKEN=""        # huggingface.co/settings/tokens, for gated models
    # ==========================================
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download Qwen/Qwen2.5-Coder-7B-Instruct --local-dir ~/models/Qwen2.5-Coder-7B-Instruct
    hf download veri2-anon/sva-qlora-all --local-dir ~/models/sva-qlora-all
    # The model is in ~/models. How to run it is in the model card (link above).
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

Requires the base model

Qwen/Qwen2.5-Coder-7B-Instruct

This model is a LoRA / adapter. Download the base model first, then load this one on top.

  1. Qwen2.5-7Bfamily root
  2. Qwen2.5-Coder-7B
  3. Qwen2.5-Coder-7B-Instruct
  4. sva-qlora-all

New posts about versions of Qwen2.5-7B (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 1All on Hugging Face
Merges 5All on Hugging Face

About the model

veri2-anon/sva-qlora-all
Author
veri2-anon
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

30

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments