Skip to content
Language models

Sijo2.0-1.7B

a compact model for Linux terminals and debugging

2 min readupdated

Sijo-1.7B v2.0 is a lightweight 1.73B-parameter language model by MyJellyLab built for Linux terminal tasks, Python bug hunting, and step-by-step reasoning. It outputs ready-to-run Bash commands, fixed code, XML tool calls, and chain-of-thought reasoning inside ... tags. Fine-tuned with LoRA (a lightweight adapter for adjusting model weights) on a curated dataset mix designed to prevent catastrophic forgetting, it is a solid fit for developers and hobbyists who want a local assistant for the console and basic coding.

Key features

  • 24-layer hybrid architecture adapted via standard FP16 LoRA (r=16, alpha=32, NEFTune=5.0).
  • Supports structured reasoning inside ... tags and native tool calling using XML .
  • Trained on a mix of 7,500 SFT (supervised fine-tuning) examples and 800 DPO (direct preference optimization) pairs with anti-forgetting protection.
  • Accurately recognizes when no tools are connected and answers general non-coding prompts without unnecessary refusals.

What you can do

  • Plug it into a local CLI assistant to generate Bash commands and automate routine Linux terminal tasks.
  • Use it to spot bugs in Python scripts and perform multi-step code debugging.
  • Deploy it as a lightweight agentic model to invoke external utilities via XML tags.

Example

The model passed 6 out of 7 benchmark scenarios tested by the author: it properly handles empty lists during Python bug fixes, builds Linux CLI commands wrapped in XML calls, self-corrects across multi-step tasks, and honestly reports when no tools are available.

Read moreDetails
Eval results (FP16): 6 of 7 test suites passed (PASS); the model identity check was flagged as REVIEW.
Datasets used: Nemotron-Terminal-Corpus, NL2SH-ALFA, Open-SWE-Traces, When2Call, Grounded-QA, Tulu-3, and others.
Language: English.
Discuss Telegram 0 181

Will it run?

Runs in F16

Fits without quantisation.

  • F16weights 3.2 GB · ready files4.2 GB
  • FP8 / INT8recommendedweights 1.6 GB2.6 GB
  • 4-bitweights 0.9 GB1.9 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 2060 · F16 — from RTX 2060

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

Language models: with a 8192-token context.

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

Run it on Massed Compute

A ready script for the machine we advise above: paste it and the model downloads and starts.

  1. 1. Rent A30

    $0.26 an hour with the coupon. Pick the Ubuntu image when you create the machine.

    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.

    # myjellylabs/Sijo2.0-1.7B — A30, FP8 / INT8
    # NVIDIA driver: CUDA 13 needs 580+. If older or missing, the current nvidia-open is installed (keep yours: KEEP_DRIVER=1)
    DRV=$(nvidia-smi --query-gpu=driver_version --format=csv,noheader 2>/dev/null | head -1 | cut -d. -f1)
    case "$DRV" in ""|*[!0-9]*) DRV=0 ;; esac
    if [ "$DRV" -lt 580 ] && [ -z "${KEEP_DRIVER:-}" ]; then
      . /etc/os-release && \
      curl -fsSLO "https://developer.download.nvidia.com/compute/cuda/repos/ubuntu${VERSION_ID//./}/x86_64/cuda-keyring_1.1-1_all.deb" && \
      sudo dpkg -i cuda-keyring_1.1-1_all.deb && rm -f cuda-keyring_1.1-1_all.deb && \
      sudo apt-get update && \
      sudo DEBIAN_FRONTEND=noninteractive apt-get install -y nvidia-open || \
      { echo "Could not install the NVIDIA driver: see the error above. To keep the current one, run with KEEP_DRIVER=1."; exit 1; }
      echo "NVIDIA driver updated. Reboot the machine (sudo reboot) and run this same script again."
      exit 0
    fi
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv venv ~/vllm --python 3.12 && source ~/vllm/bin/activate
    uv pip install vllm
    vllm serve myjellylabs/Sijo2.0-1.7B --max-model-len 8192 --quantization fp8
    Add more LoRAs to Sijo2.0-1.7B
  3. 3. Use it

    The server listens on port 8000 and speaks the OpenAI API: http://localhost:8000/v1 on the machine. From your computer: ssh -L 8000:localhost:8000 user@machine-IP.

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.

  • MTP (multi-token prediction)0.34.0from November 16, 2026new release

    1.5–2× generation speed

    If the model has MTP layers (DeepSeek-V3/R1, Qwen3-Next, GLM, MiMo and others), it predicts several tokens per step itself.

    How to enable: vLLM / SGLang: --speculative-config '{"method": "mtp", "num_speculative_tokens": 1}'. Check the model card.

    Earlier generations (2)
    • EAGLE-3 — Speculative decoding with a small draft head. For models without their own MTP layers.
    • EAGLE-2 / Medusa — Earlier draft-head generations. Still work if that is all you find.
  • FP8 KV cache0.34.0from November 16, 2026new release

    2× more context in the same memory

    Keeps the attention cache in 8 bits, so long contexts fit where they did not.

    How to enable: vLLM: --kv-cache-dtype fp8; llama.cpp: -ctk q8_0 -ctv q8_0.

  • vLLM / SGLang0.34.0from November 16, 2026new release

    many times more tokens/s on a server

    Server engines: batching, FlashAttention, FP8 and speculative decoding out of the box.

    How to enable: vllm serve <model> — an OpenAI-compatible API.

    Earlier generations (1)
    • llama.cpp — GGUF models on any GPU, Mac or even CPU; some layers can stay in RAM.

Model family

New posts about versions of Sijo2.0-1.7B (quants, finetunes, LoRAs) will show up under the bell.

LoRAs and adapters 2All on Hugging Face

About the model

myjellylabs/Sijo2.0-1.7B
Author
myjellylabs
Category
Language models
Released
Posted

On NeyroLearn

Hugging Face

1810

Ask AI about this model

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

Comments