|
Download README.md from usmar/undertow: direct link, hf CLI and curl.
- Browser
- Download file 1.36 kB
-
https://huggingface.co/spaces/usmar/undertow/resolve/main/README.md
- Command line
-
hf download hf://spaces/usmar/undertow/README.md
-
curl -L -o README.md https://huggingface.co/spaces/usmar/undertow/resolve/main/README.md
1.36 kB
| title: undertow | |
| emoji: π | |
| colorFrom: blue | |
| colorTo: gray | |
| sdk: static | |
| pinned: false | |
| license: mit | |
| short_description: Four sentiment corpora, every model tested on all four | |
| # undertow | |
| A sentiment classifier reports one number: accuracy on a test set drawn from the same | |
| corpus it trained on. **undertow** is about the other twelve. | |
| Four corpora β IMDB, SST-2, Yelp, Amazon β two classifiers, and every model evaluated on | |
| every domain. A bag of words and a linear probe on frozen DistilBERT features are nearly | |
| tied on the test set everyone reports (86.0% vs 87.1%). On the three domains nobody | |
| reports, the probe wins by 3.5 points and carries 29% less domain-shift penalty. Picking | |
| the method on in-domain accuracy gets it wrong on two of the four corpora. | |
| This page runs the TF-IDF arm live in your browser: type a review and the four models, | |
| identical in every way except which 6,000 reviews they read, each give you a verdict. | |
| Nothing is sent anywhere and no answer is precomputed β the JavaScript is a tested port of | |
| the Python original, held to it at 1e-9. | |
| Corpora: [IMDB](https://ai.stanford.edu/~amaas/data/sentiment/) (Maas et al., 2011), | |
| [SST-2](https://gluebenchmark.com/tasks) (Socher et al., 2013), Yelp and Amazon polarity | |
| (Zhang et al., 2015). | |
| Method, controls and the full write-up: <https://github.com/UsmarHaider/undertow> | |