:house_with_garden: Fast & easy transfer learning for NLP. Harvesting language models for the industry. Focus on Question Answering.
1.73k
stars
247
forks
source link
Allow more flexible args for model/tokenizer loading (incl. proxies etc) #809
Closed
tholor closed 3 years ago
The first step towards implementing https://github.com/deepset-ai/haystack/issues/997
We need to pass some args like proxies or resume_download all the way down from the inferencer to HF's
from_pretrained()