Instructions to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Use Docker

docker model run hf.co/simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

LM Studio
Jan

vLLM

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

SGLang

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with Ollama:
```
ollama run hf.co/simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M
```
Unsloth Desktop

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with Docker Model Runner:
```
docker model run hf.co/simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M
```

Lemonade

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Run and chat with the model

lemonade run user.poc-simorg-coder-30b-a3b-qwen3-lora-v0.1-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 / MANIFEST.md

miladhabibi

Add files using upload-large-folder tool

6014f33 verified 3 months ago

preview code

Raw

History Blame Contribute Delete

4.56 kB

Hugging Face publish manifest

Complete list of files in this directory for upload to Hugging Face.

Regenerate staging (hardlinks + copies):

uv run python scripts/prepare_huggingface_publish.py

Upload entire folder (~183 GB) — use upload-large-folder:

$env:HF_XET_HIGH_PERFORMANCE = "1"
uv run hf upload-large-folder simorg-platform/poc-simorg-coder-30b-a3b-qwen3-lora-v0.1 .\huggingface\ `
  --repo-type model --num-workers 8

Do not use plain hf upload for the full folder — HF will warn it may fail. See docs/publish.md.

Root files (required)

File	Purpose
`README.md`	Model card (YAML front matter + usage)
`LICENSE`	Apache License 2.0 (from Qwen base model)
`NOTICE`	Derivative-work attribution
`.gitattributes`	Git LFS rules for `.safetensors` and `.gguf`
`MANIFEST.md`	This file list

`safetensors/` — merged full model (~60 GB)

All files from merged LoRA + base weights. No scripts, no tokens.

File
`config.json`
`generation_config.json`
`model.safetensors.index.json`
`model-00001-of-00013.safetensors`
`model-00002-of-00013.safetensors`
`model-00003-of-00013.safetensors`
`model-00004-of-00013.safetensors`
`model-00005-of-00013.safetensors`
`model-00006-of-00013.safetensors`
`model-00007-of-00013.safetensors`
`model-00008-of-00013.safetensors`
`model-00009-of-00013.safetensors`
`model-00010-of-00013.safetensors`
`model-00011-of-00013.safetensors`
`model-00012-of-00013.safetensors`
`model-00013-of-00013.safetensors`
`tokenizer.json`
`tokenizer_config.json`
`vocab.json`
`merges.txt`
`added_tokens.json`
`special_tokens_map.json`
`chat_template.jinja`

Excluded: training scripts, workspace/ cache, HF tokens.

`gguf/` — quantized models (~121 GB total)

File	Approx. size	Notes
`qwen3-coder-simorg-Q3_K_M.gguf`	~14 GB	Smallest
`qwen3-coder-simorg-Q4_K_S.gguf`	~16 GB
`qwen3-coder-simorg-Q4_K_M.gguf`	~17 GB	Recommended
`qwen3-coder-simorg-Q5_K_M.gguf`	~20 GB
`qwen3-coder-simorg-Q6_K.gguf`	~23 GB
`qwen3-coder-simorg-Q8_0.gguf`	~30 GB

Excluded: qwen3-coder-simorg-f16.gguf (local intermediate only; use safetensors/ for full weights). Legacy qwen-simorg-f16.gguf (older Qwen2.5 build).

`lora/` — PEFT adapter (~1–2 GB)

File	Purpose
`adapter_config.json`	LoRA configuration
`adapter_model.safetensors`	LoRA weights
`tokenizer.json`	Tokenizer
`tokenizer_config.json`
`vocab.json`
`merges.txt`
`added_tokens.json`
`special_tokens_map.json`
`chat_template.jinja`	Inference chat template

Excluded: model.safetensors (Ollama-only duplicate), auto-generated PEFT README.md, scripts.

`training-data/` — SFT dataset

File	Entries	Format
`train-global.json`	131	`[{"question": "...", "answer": "..."}, ...]`

Add future datasets here before re-running prepare_huggingface_publish.py.

`ollama/` — Modelfiles (6 files)

File	GGUF reference
`Modelfile-qwen-3-Q3-K-M`	`../gguf/qwen3-coder-simorg-Q3_K_M.gguf`
`Modelfile-qwen-3-Q4-K-S`	`../gguf/qwen3-coder-simorg-Q4_K_S.gguf`
`Modelfile-qwen-3-Q4-K-M`	`../gguf/qwen3-coder-simorg-Q4_K_M.gguf`
`Modelfile-qwen-3-Q5-K-M`	`../gguf/qwen3-coder-simorg-Q5_K_M.gguf`
`Modelfile-qwen-3-Q6-K`	`../gguf/qwen3-coder-simorg-Q6_K.gguf`
`Modelfile-qwen-3-Q8-0`	`../gguf/qwen3-coder-simorg-Q8_0.gguf`

Paths are relative — run ollama create from the ollama/ directory after clone.

Security checklist

Before upload, confirm:

No HF_TOKEN or hf_... secrets in any staged file
No .env or credential files
No Python/shell scripts in this folder
adapter_config.json only references public base model ID

Total upload size (approximate)

Section	Size
`safetensors/`	~60 GB
`gguf/`	~121 GB
`lora/`	~1–2 GB
`training-data/`	<1 MB
`ollama/`	<1 MB
Total	~183 GB

Use uv run hf upload or Git LFS; stable connection recommended.

Hugging Face publish manifest

Root files (required)

safetensors/ — merged full model (~60 GB)

gguf/ — quantized models (~121 GB total)

lora/ — PEFT adapter (~1–2 GB)

training-data/ — SFT dataset

ollama/ — Modelfiles (6 files)

Security checklist

Total upload size (approximate)

`safetensors/` — merged full model (~60 GB)

`gguf/` — quantized models (~121 GB total)

`lora/` — PEFT adapter (~1–2 GB)

`training-data/` — SFT dataset

`ollama/` — Modelfiles (6 files)