Leshem Choshen<p>I don’t train from scratch, I use RoBERTa🧐<br />Wait…<br />Why not cross-encoder/stsb-roberta?facebook/muppet-roberta?</p><p>We automatically identify the best models on 🤗(periodically)</p><p>Just pick the best one<br />and finetune on your task</p><p><a href="https://ibm.github.io/model-recycling/" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="">ibm.github.io/model-recycling/</span><span class="invisible"></span></a></p><p><a href="https://sigmoid.social/tags/NLProc" class="mention hashtag" rel="tag">#<span>NLProc</span></a> <a href="https://sigmoid.social/tags/MachineLearning" class="mention hashtag" rel="tag">#<span>MachineLearning</span></a> <a href="https://sigmoid.social/tags/finetuning" class="mention hashtag" rel="tag">#<span>finetuning</span></a> <a href="https://sigmoid.social/tags/intertraining" class="mention hashtag" rel="tag">#<span>intertraining</span></a> <a href="https://sigmoid.social/tags/huggingFace" class="mention hashtag" rel="tag">#<span>huggingFace</span></a> <a href="https://sigmoid.social/tags/SoTA" class="mention hashtag" rel="tag">#<span>SoTA</span></a></p>