Upload fp16/q4f16 decoder ONNX weights w/ float32 inputs_embeds

#15
by Xenova HF staff - opened
onnx/decoder_model_merged_fp16.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:18906c3a7f15ae8176e0208184510ebacb8f65c66f7e80e89226918a267c147a
3
- size 1342510363
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6ec36a518b896dfc6448c4343898d0ea5109702677d34cea3919fba074044d1
3
+ size 1342510427
onnx/decoder_model_merged_q4f16.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:73e9722bd19fe5b4174d8b6eabe28d590d3a0cbd45cb0b4165e15b084aa0187c
3
- size 965031413
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d74ec46083829ddb18f58fcceb358d2ba58d2a1320bdab431c32e4d2896981d
3
+ size 965031477