ULMFiT

fast.ai · 2018 · Transfer Learning · USA / Ireland

A to Z

A · App: No
B · Built By: fast.ai
C · CEO & Politics: [UNKNOWN]
D · Date: 2018
E · Expertise: Intermediate+
F · Free Tier: Yes
G · Generates: Text only
H · Deployment: Cloud + Local (open weights)
I · Inputs: Text
J · Jurisdiction: USA / Ireland
K · Key Staffers: Jeremy Howard (Lead author), Sebastian Ruder (Co-author)
L · Languages: English-primary
M · Model Type: Transfer Learning
N · Safety Issues: None reported
O · Open Source: Yes — open weights
P · Pricing: Free (open source in the fastai library, Apache 2.0 license)
Q · AIDb Rating: [UNKNOWN]
R · What Makes It Special: ULMFiT (Universal Language Model Fine-tuning), Jeremy Howard and Sebastian Ruder's transfer-learning method for NLP from their ACL 2018 paper: it fine-tunes a pretrained language model for any NLP task, instead of the task-specific changes and training from scratch NLP then needed. They reported it beat the state of the art on six text classification tasks, cutting error 18-24% on most, and matched training from scratch on 100x more data with 100 labeled examples; models and code are open.
S · Stock / Status: Research / open-source project
T · Training Cutoff: [UNKNOWN]
U · Users: [UNKNOWN]
V · Voice Chat: No
W · Weights: Downloadable
X · X Handle: [UNKNOWN]
Y · Year Founded: [UNKNOWN]
Z · Bottom Line: Open-weights Transfer Learning from fast.ai, released 2018.

Data completeness 77% · Last verified 2026-05-12 · Primary source ↗

“ULMFiT.” AIDb — The AI Database. Verified 2026-05-12. https://theaidatabase.com/m/ulmfit.html