Instructions to use sharryXR/asil-qwen35-9b-rl with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use sharryXR/asil-qwen35-9b-rl with Transformers:

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

pipe = pipeline("text-generation", model="sharryXR/asil-qwen35-9b-rl")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("sharryXR/asil-qwen35-9b-rl")
model = AutoModelForCausalLM.from_pretrained("sharryXR/asil-qwen35-9b-rl", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use sharryXR/asil-qwen35-9b-rl with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sharryXR/asil-qwen35-9b-rl"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sharryXR/asil-qwen35-9b-rl",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/sharryXR/asil-qwen35-9b-rl

SGLang

How to use sharryXR/asil-qwen35-9b-rl 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 "sharryXR/asil-qwen35-9b-rl" \
    --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": "sharryXR/asil-qwen35-9b-rl",
		"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 "sharryXR/asil-qwen35-9b-rl" \
        --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": "sharryXR/asil-qwen35-9b-rl",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use sharryXR/asil-qwen35-9b-rl with Docker Model Runner:
```
docker model run hf.co/sharryXR/asil-qwen35-9b-rl
```

ASIL Qwen3.5-9B RL

This model is part of the ASIL paper release (Findings of EMNLP 2026).
Project page: https://sharryxr.github.io/ASIL/
Code: https://github.com/sharryXR/ASIL

This repository contains the ASIL v0.1.0 paper release checkpoint for ASIL Qwen3.5-9B RL.

Release: v0.1.0
Selected checkpoint: global_step_140_actor_hf
Source path: /public/LLM_model_dataset/rl_temp/asil_operational_benchmark_a100_20260514_144617/results/rl/qwen35_9b_agentic_a800/qwen35_9b_rl_from_sft543_resume80_p100_307542_20260523_224030/checkpoints/global_step_140_actor_hf
Base/init checkpoint: /public/LLM_model_dataset/rl_temp/asil_sft_rl_a100_20260513_173133/results/sft_train/qwen35_9b_base_sft_v0_v1_6ep_20260520_000159/checkpoints/global_step_543; resumed from Verl global_step_80 in the first-stage RL run
Training data: rl_learnable_v4_320_80; 320 train / 80 valid task prompts
Prepared at: 2026-07-30T18:02:32+08:00

The repo root contains the HF-loadable checkpoint files (config.json, tokenizer files, generation_config.json, and *.safetensors). Training-only artifacts such as optimizer state, scheduler state, trainer state, logs, caches, wandb output, and credentials are intentionally excluded.

See checkpoint_metadata.json and SHA256SUMS for provenance and file checksums.