Downloads · 30 days
0
scihigh2026-subtask1-qwen7b-qlora
Yakk99/scihigh2026-subtask1-qwen7b-qlora
scihigh2026-subtask1-qwen7b-qlora is a summarization model from Yakk99. Use it when you need a shorter version of a longer text. The card lists the license as apache-2.0.
QLoRA adapter (4-bit NF4 base, LoRA r=16 on attention+MLP).
Access
Public
Updated Aug 12, 2026
Repo size
92.2 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors80.8 MB · 88%
At a glance
- Task
- Summarization
- License
- apache-2.0
- Access
- Public
- Created
- Aug 11, 2026
- Updated
- Aug 12, 2026
- SHA
- 2349cc2c
Try a prompt
Base models
Model card
From the Hugging Face model README
SciHigh-2026 Subtask 1 — Qwen2.5-7B QLoRA (pilot)
QLoRA adapter (4-bit NF4 base, LoRA r=16 on attention+MLP).
Fine-tuned for the FIRE 2026 SciHigh shared task, Subtask 1: generating research highlights from scientific paper abstracts (MixSub corpus).
- Input format:
<paper title> | <abstract>(corpus punctuation-stripped style; abstracts repaired via DOI-verified Semantic Scholar recovery) - Output: 3-5 highlight sentences, ~55 words, greedy decoding
- Training: one epoch on 15,960 pairs, RTX 4090
- Validation ROUGE-L F1 (plain LCS, no stemming): 27.55
Repository contents
| file | purpose |
|---|---|
train_final.py | standalone training recipe that produced this model |
train_final.ipynb | notebook version of the same recipe |
infer_test.py | test-set submission CSV generation |
final/ | full data pipeline: abstract recovery, title fetch, dataset build, submission validation |
Data pipeline summary
~40% of corpus abstracts are truncated mid-sentence. Each row was matched to
its ScienceDirect PII (100% verified join), resolved to a DOI via Elsevier's
keyless API, and its complete abstract recovered from Semantic Scholar under
a label-free >=90%-token-overlap acceptance filter (~95% of truncated rows
repaired). Paper titles were fetched for 100% of rows and prepended to
inputs. See final/README.md for the full pipeline and negative-results
summary. Trained 2026-08-10.
- Task
- Summarization
- License
- apache-2.0
- Created
- Aug 11, 2026
- Updated
- Aug 12, 2026