gemma4-4b-sci

Early-stage research experiment. Trained for 600 steps on 30K examples. Expect hallucinations and factual errors.

gemma4-4b-sci is a scientific-domain fine-tune of Gemma 4 E4B via QLoRA on 30,000 examples from OpenSciLM/OS_Train_Data and SciRIFF. Inspired by OpenScholar — this is a generation-only model without a retrieval pipeline.

Model Description

• Developed by: Michele Banfi
• Base model: unsloth/gemma-4-E4B-it
• Method: QLoRA (4-bit) + SFT via Unsloth, language layers only (vision encoder frozen)
• Training: 600 steps, 30K examples (15K OS_Train_Data + 15K SciRIFF), NVIDIA RTX 5090
• License: Gemma Terms of Use

Model Sources

• Repository: https://github.com/michelebanfi/gemma-4-finetuning
• Evaluation: ScholarQABench
• Ollama: ollama run hf.co/linosium/gemma4-4b-sci

Quick Start

ollama run hf.co/linosium/gemma4-4b-sci

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model = AutoModelForCausalLM.from_pretrained("linosium/gemma4-4b-sci", torch_dtype=torch.bfloat16, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("linosium/gemma4-4b-sci")

messages = [{"role": "user", "content": "Explain the role of CRISPR-Cas9 in gene editing."}]
input_ids = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(input_ids, max_new_tokens=512)[0][input_ids.shape[1]:], skip_special_tokens=True))

Evaluation

ScholarQABench — draft results, 600-step run. Tier 1 uses gold paper contexts (fair comparison). Tier 2 has no retrieval, so citation scores are 0 by design.

Tier 1 — single-paper tasks

Task	Metric	gemma4-4b-sci	OpenScholar-8B
SciFact (208)	Accuracy	77.9%	76.4%
PubMedQA (843)	Accuracy	81.5%	76.0%
QASA (1375)	ROUGE-L	20.9	23.0
SciFact	Citation F1	0.0	68.9
PubMedQA	Citation F1	0.0	43.6
QASA	Citation F1	4.3	56.3

Correctness matches or exceeds OpenScholar-8B (2× the parameters) at 600 steps. Citation gap is entirely due to the missing retrieval pipeline.

Tier 2 — synthesis tasks (no retrieval)

Task	Citation F1	OpenScholar-8B
ScholarQA-CS (110)	0.0	47.9
ScholarQA-Bio (1451)	0.0	42.8
ScholarQA-Neuro (1308)	0.0	50.8

Citation

@article{asai2024openscholar,
  title   = {OpenScholar: Synthesizing Scientific Literature with Retrieval-Augmented LMs},
  author  = {Asai, Akari and others},
  journal = {Nature},
  year    = {2024},
  url     = {https://allenai.org/blog/nature-openscilm}
}