[All Seq2Seq model + CLM models that can be used with EncoderDecoder] Add cross-attention weights to outputs#8071
Merged
patrickvonplaten merged 8 commits intohuggingface:masterfrom ysgit:output-crossattentionNov 6, 2020
+653-85
Commits
Commits on Nov 2, 2020
- committed
Yossi Synett
Commits on Nov 6, 2020
- authored
- committed
- committed
- committed
- committed
- committed
- committed
