Instructions to use yujie111222/tc-4b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use yujie111222/tc-4b with Transformers:

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

pipe = pipeline("image-text-to-text", model="yujie111222/tc-4b")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("yujie111222/tc-4b")
model = AutoModelForMultimodalLM.from_pretrained("yujie111222/tc-4b", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use yujie111222/tc-4b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "yujie111222/tc-4b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "yujie111222/tc-4b",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/yujie111222/tc-4b

SGLang

How to use yujie111222/tc-4b 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 "yujie111222/tc-4b" \
    --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": "yujie111222/tc-4b",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "yujie111222/tc-4b" \
        --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": "yujie111222/tc-4b",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use yujie111222/tc-4b with Docker Model Runner:
```
docker model run hf.co/yujie111222/tc-4b
```

tc-4b — Torchcast Decision 4B, continued (step-49)

Continued training of torchcast-ai/torchcast-decision-4b v1.3.1 (base: Qwen3.5-4B) with LoRA (r=32, alpha=64, lr 1e-4, 49 optimizer steps, weights merged), on ~94k admitted rows from a license-filtered, stratified sample of DecisionBank.

base_model: Qwen3.5-4B (via torchcast-ai/torchcast-decision-4b v1.3.1)
Training data: DecisionBank-260k subset (100k sampled under a license allowlist — apache/mit/bsd/cc-by/cc0/unlicense/public-domain; no JevBench public-split items used)
Serving: identical to the base model — bundled torchcast_decision/ package + decision_config.json; full-attention layers are noncausal; follow the base model serving instructions. Generic causal-attention runtimes (vLLM, llama.cpp) do not reproduce the evaluated interface.
Local reference (JevBench 231 public items, v1.6 scorer): Intelligence 79.81 / Calibration 91.26 / Capability 85.54 (base v1.3.1: 74.93 / 87.82 / 81.37). Gains concentrate in the score question type (distribution quality, expected-value distance), not argmax accuracy.

Licence & attribution

Apache License 2.0. Derived from torchcast-ai/torchcast-decision-4b (Apache-2.0, © Torchcast AI), itself a fine-tune of Qwen3.5-4B (Apache-2.0, © 2026 Alibaba Cloud). See LICENSE, NOTICE, and the bundled torchcast_decision/ package's own licence files. merge_manifest.json documents the per-file SHA-256 of the merge.

Downloads last month: 26

Safetensors

Model size

5B params

Tensor type

BF16

Model tree for yujie111222/tc-4b

Base model

Qwen/Qwen3.5-4B-Base

Finetuned

Qwen/Qwen3.5-4B

Finetuned

torchcast-ai/torchcast-decision-4b

Adapter

(2)

this model