Instructions to use vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone with Transformers:

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

pipe = pipeline("text-generation", model="vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone

SGLang

How to use vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone 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 "vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone" \
    --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": "vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone",
		"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 "vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone" \
        --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": "vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone with Docker Model Runner:
```
docker model run hf.co/vtava/Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone
```

Qwen3.5-0.8B-PDelta3-CLVR-Local32-Standalone / load_test.py

vtava

Upload validated TinyCeNN standalone release

e9f1128 verified 17 days ago

Raw History Blame Contribute Delete

1.14 kB

	from pathlib import Path
	import json
	import torch
	from load_model import load_model

	root = Path(__file__).resolve().parent
	meta = json.loads((root / "standalone_config.json").read_text())
	probe = json.loads((root / "validation_probe.json").read_text())
	model, tokenizer = load_model(root)
	classes = [m.__class__.__name__ for m in model.modules()]
	assert meta["expected_class"] in classes, f"Missing custom class: {meta['expected_class']}"
	import tinycenn_lm
	runtime = Path(tinycenn_lm.__file__).resolve()
	assert str(runtime).startswith(str(root)), f"External TinyCeNN runtime used: {runtime}"
	ids = tokenizer(probe["prompt"], return_tensors="pt").input_ids.to(next(model.parameters()).device)
	with torch.inference_mode():
	out = model(input_ids=ids, use_cache=False, return_dict=True)
	assert torch.isfinite(out.logits).all()
	actual = int(out.logits[0, -1].argmax())
	assert actual == int(probe["next_token_id"]), (actual, probe["next_token_id"])
	print("STANDALONE_RELOAD_PASS")
	print("architecture=", meta["expected_class"])
	print("layers=", meta["layers"])
	print("runtime=", runtime)
	print("next_token=", tokenizer.decode([actual]))