Commit History

Set kv_cache_dtype + torch_dtype to float32 (matches actual ONNX weights)
c167f01
verified

shreyask commited on

Drop dangling model.onnx ref from use_external_data_format
c524ad2
verified

shreyask commited on

Patch ONNX layout + config for Transformers.js (onnx/model_q4)
aabf777
verified

shreyask commited on

Upload config.json
ee6691d
verified

shreyask commited on