Instructions to use scienceverse/bibr-extract-qwen35-4b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use scienceverse/bibr-extract-qwen35-4b with Transformers:

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

pipe = pipeline("image-text-to-text", model="scienceverse/bibr-extract-qwen35-4b")
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("scienceverse/bibr-extract-qwen35-4b")
model = AutoModelForMultimodalLM.from_pretrained("scienceverse/bibr-extract-qwen35-4b", 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 scienceverse/bibr-extract-qwen35-4b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "scienceverse/bibr-extract-qwen35-4b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "scienceverse/bibr-extract-qwen35-4b",
		"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/scienceverse/bibr-extract-qwen35-4b

SGLang

How to use scienceverse/bibr-extract-qwen35-4b 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 "scienceverse/bibr-extract-qwen35-4b" \
    --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": "scienceverse/bibr-extract-qwen35-4b",
		"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 "scienceverse/bibr-extract-qwen35-4b" \
        --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": "scienceverse/bibr-extract-qwen35-4b",
		"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 scienceverse/bibr-extract-qwen35-4b with Docker Model Runner:
```
docker model run hf.co/scienceverse/bibr-extract-qwen35-4b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

bibr-extract-qwen35-4b

This is the structured-extraction student model used by bibr. It answers bibr's extraction prompts (statistical equations, reference segmentation, funding / contributions / affiliations, …) in JSON. The model is Qwen/Qwen3.5-4B fine-tuned on answers from a larger teacher model, with the LoRA merged into full bf16 weights.

GGUF builds (Q8_0, Q4_K_M) are in scienceverse/bibr-extract-qwen35-4b-GGUF.

Training

Base model: Qwen/Qwen3.5-4B (Apache-2.0), revision 851bf6e
Method: LoRA (r=16, alpha=32, dropout 0) on every language-model linear layer (200 modules: attention q/k/v/o, linear-attention in_proj_qkv / in_proj_z / out_proj, MLP gate/up/down). 1 epoch, lr 1e-4, 161 optimizer steps, loss on the assistant answer only.
Merge: the adapter is merged into the bf16 base weights (W + 2.0·BA), bitwise identical to peft merge_and_unload. The vision tower and the MTP head are unchanged from the base. The MTP head was not trained, so speculative decoding with it may accept fewer tokens.
Training data: 1,113 answers by Qwen3.8-27B (the teacher) to bibr extraction prompts on 181 CC BY papers from PubMed Central.

Evaluation

These scores come from bibr's extraction harness of 9 Oct 2026 (bibr 9621739, v5 scorer, gate192 + ci60), with the unmerged LoRA served in bf16 on vLLM:

model	passes / 252
Qwen3.8-27B (teacher)	223
bibr-extract-qwen35-4b	216
bibr-extract-qwen35-2b	207

Usage notes

Use the model with thinking off, which is how it was trained and scored. The chat template here defaults to thinking off (the upstream Qwen3.5-4B template defaults to on). enable_thinking=True still works when passed. llama-server turns thinking on whenever a template supports it, so run it with --reasoning off or send "chat_template_kwargs": {"enable_thinking": false}.

Downloads last month: 45

Safetensors

Model size

5B params

Tensor type

BF16

F32

Model tree for scienceverse/bibr-extract-qwen35-4b

Base model

Qwen/Qwen3.5-4B-Base

Finetuned

Qwen/Qwen3.5-4B

Finetuned

(1180)

this model

Quantizations

4 models