search

facebookresearch / fastText

Library for fast text representation and classification.
https://fasttext.cc/
MIT License
25.93k stars 4.72k forks source link

Didn't have enough time to train once #1163

Closed rzsgrt closed 3 years ago

rzsgrt commented 3 years ago

Hi, i'm using latest fasttext (0.9.2) and use autotune for supervised classification. my train data around 1 million and my val data around 1K. Other argument i use : autotuneModelSize = "20M" and autotuneDuration = 600 It seems that i only have one best score and then at the end i got error RuntimeError: Didn't have enough time to train once: please increaseautotune-duration. I already increase autotuneDuration, but still got same error. Is it because train size?

rzsgrt commented 3 years ago

Turns out that i makes mistake when create train data