Instructions to use deepkick/qwen3-4b-advanced-sft-v0-merged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use deepkick/qwen3-4b-advanced-sft-v0-merged with Transformers:

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

pipe = pipeline("text-generation", model="deepkick/qwen3-4b-advanced-sft-v0-merged")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("deepkick/qwen3-4b-advanced-sft-v0-merged")
model = AutoModelForMultimodalLM.from_pretrained("deepkick/qwen3-4b-advanced-sft-v0-merged")
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 deepkick/qwen3-4b-advanced-sft-v0-merged with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "deepkick/qwen3-4b-advanced-sft-v0-merged"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepkick/qwen3-4b-advanced-sft-v0-merged",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/deepkick/qwen3-4b-advanced-sft-v0-merged

SGLang

How to use deepkick/qwen3-4b-advanced-sft-v0-merged 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 "deepkick/qwen3-4b-advanced-sft-v0-merged" \
    --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": "deepkick/qwen3-4b-advanced-sft-v0-merged",
		"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 "deepkick/qwen3-4b-advanced-sft-v0-merged" \
        --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": "deepkick/qwen3-4b-advanced-sft-v0-merged",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use deepkick/qwen3-4b-advanced-sft-v0-merged with Docker Model Runner:
```
docker model run hf.co/deepkick/qwen3-4b-advanced-sft-v0-merged
```

qwen3-4b-advanced-sft-v0-merged / README.md

deepkick

Upload merged model (SFT v0)

785397e verified 4 months ago

preview code

raw

history blame contribute delete

1.21 kB

metadata

base_model: Qwen/Qwen3-4B-Instruct-2507
datasets:
  - u-10bei/sft_alfworld_trajectory_dataset_v5
language:
  - en
license: apache-2.0
library_name: transformers
pipeline_tag: text-generation
tags:
  - agent
  - alfworld
  - sft
  - merged

qwen3-4b-advanced-sft-v0-merged

This repository provides a merged model fine-tuned from Qwen/Qwen3-4B-Instruct-2507.

Dataset: u-10bei/sft_alfworld_trajectory_dataset_v5
Method: LoRA SFT (merged into base model)
Max sequence length: 2048
Epochs: 2
Learning rate: 2e-06
LoRA: r=64, alpha=128

vLLM Compatibility

LoRA adapter has been merged
No tokenizer vocabulary modification
Intended for AgentBench Advanced evaluation

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM

model_id = "deepkick/qwen3-4b-advanced-sft-v0-merged"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto"
)

Data / License Notes

Dataset: u-10bei/sft_alfworld_trajectory_dataset_v5 Please refer to the dataset page for license and terms.
Base model: Qwen/Qwen3-4B-Instruct-2507 Please comply with the base model's original terms of use.