* Patch clip model for ONNX compatibility Changes to use INT32 for tokenization, since ONNX doesn't yet support ArgMax(INT64) Use explicit dimension for norm * Add compatibility fix for torch 1.7 |
||
|---|---|---|
| .. | ||
| __init__.py | ||
| bpe_simple_vocab_16e6.txt.gz | ||
| clip.py | ||
| model.py | ||
| simple_tokenizer.py | ||