Instructions to use techwithsergiu/Qwen3.5-text-9B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use techwithsergiu/Qwen3.5-text-9B with Transformers:

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

pipe = pipeline("text-generation", model="techwithsergiu/Qwen3.5-text-9B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("techwithsergiu/Qwen3.5-text-9B")
model = AutoModelForCausalLM.from_pretrained("techwithsergiu/Qwen3.5-text-9B")
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

vLLM

How to use techwithsergiu/Qwen3.5-text-9B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "techwithsergiu/Qwen3.5-text-9B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "techwithsergiu/Qwen3.5-text-9B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/techwithsergiu/Qwen3.5-text-9B

SGLang

How to use techwithsergiu/Qwen3.5-text-9B 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 "techwithsergiu/Qwen3.5-text-9B" \
    --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": "techwithsergiu/Qwen3.5-text-9B",
		"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 "techwithsergiu/Qwen3.5-text-9B" \
        --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": "techwithsergiu/Qwen3.5-text-9B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use techwithsergiu/Qwen3.5-text-9B with Docker Model Runner:
```
docker model run hf.co/techwithsergiu/Qwen3.5-text-9B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Qwen3.5-text-9B

Text-only bf16 derivative of Qwen/Qwen3.5-9B.

The visual tower (vision encoder, image merger, video preprocessor) has been removed. All text-backbone weights are identical to the original — no retraining, no weight changes, no quality loss for text tasks.

Primary use-case: intermediate model for GGUF conversion or CPU-side f16 merge after LoRA training. For direct fine-tuning use techwithsergiu/Qwen3.5-text-9B-bnb-4bit.

What was changed

Visual tower removed: visual, image_newline, patch_embed, and related keys stripped from safetensors shards
config.json updated: architectures → Qwen3_5ForCausalLM, vision_config removed
tokenizer_config.json and chat_template.jinja: image/video branches stripped from the Jinja2 chat template — prevents tokenizer errors when no image is provided
Vision-specific sidecar files omitted (preprocessor_config.json, processor_config.json, video_preprocessor_config.json)
All text weights remain at bf16

Model family

Model	Type	Base model
Qwen/Qwen3.5-9B	f16 · VLM · source	—
techwithsergiu/Qwen3.5-9B-bnb-4bit	BNB NF4 · VLM	Qwen/Qwen3.5-9B
techwithsergiu/Qwen3.5-text-9B	bf16 · text-only	Qwen/Qwen3.5-9B
techwithsergiu/Qwen3.5-text-9B-bnb-4bit	BNB NF4 · text-only	Qwen3.5-text-9B
techwithsergiu/Qwen3.5-text-9B-GGUF	GGUF quants	Qwen3.5-text-9B

Removing the visual tower saves ~0.19 GB (0.8B), ~0.62 GB (2B / 4B), or ~0.85 GB (9B). The relative saving is larger for smaller models.

Inference

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_name = "techwithsergiu/Qwen3.5-text-9B"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    dtype=torch.bfloat16,
    device_map="auto",
)

messages = [{"role": "user", "content": "What is the capital of Romania?"}]

# Thinking OFF — direct answer
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
    enable_thinking=False,
)
inputs = tokenizer(text, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=256)
response = tokenizer.decode(
    outputs[0][inputs["input_ids"].shape[1]:],
    skip_special_tokens=True,
)
print(response)

# Thinking ON — chain-of-thought before the answer
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
    enable_thinking=True,
)
inputs = tokenizer(text, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=1024)
response = tokenizer.decode(
    outputs[0][inputs["input_ids"].shape[1]:],
    skip_special_tokens=True,
)
print(response)

Fine-tuning

This model is an intermediate artifact — not a direct training target. For fine-tuning, use techwithsergiu/Qwen3.5-text-9B-bnb-4bit which is the BNB-quantized version of this model.

Training pipeline (QLoRA · Unsloth · TRL): github.com/techwithsergiu/qwen-qlora-train

Pipeline diagram

Conversion

Converted using qwen35-toolkit — a Python toolkit for BNB quantization, visual tower removal, verification and HF Hub publishing of Qwen3.5 models.

Acknowledgements

Based on Qwen/Qwen3.5-9B by the Qwen Team. If you use this model in research, please cite the original:

@misc{qwen3.5,
    title  = {{Qwen3.5}: Towards Native Multimodal Agents},
    author = {{Qwen Team}},
    month  = {February},
    year   = {2026},
    url    = {https://qwen.ai/blog?id=qwen3.5}
}