fxmarty comments

Results 334 comments of


                                            fxmarty

Error onnx_proto.TensorProto.FLOAT8E4M3FN: float8e4m3fn

Thank you @medphisiker for the details.

Add SAM bettertransformer support

The gain would be marginal as

WIP: end-to-end ONNX export and inference for stable diffusion

> Inference through CPUExecutionProvider yields garbage is due to a bug in FusedConv in ONNX Runtime, tracked in https://github.com/microsoft/onnxruntime/issues/14500 > Memory usage for a single-batch inference with CUDAExecutionProvider is huge...

WIP: end-to-end ONNX export and inference for stable diffusion

`torch.jit.trace` is pretty much unusable with deep loop: https://github.com/pytorch/pytorch/issues/93943 I'll just go on with torch.jit.scrit.

SeamlessM4T support

Hi we'll be waiting for transformers to add it, feel free to ping me again @xenova

SeamlessM4T support

Thank you @bil-ash, adding it to my todos!

Support transformers 4.42

will need to patch mistral https://github.com/huggingface/transformers/pull/31696

Support transformers 4.42

Failing tests are unrelated

Clip vision model onnx export

@contrebande-labs nothing I believe!

[ONNX export] Musicgen for text-to-audio

Hi @kanger45 @MaiZhiHao @zeke-john https://github.com/huggingface/optimum/pull/1779 is merged, which exports Musicgen in several parts to generate audio samples conditioned on a text prompt (Reference: https://huggingface.co/docs/transformers/model_doc/musicgen#text-conditional-generation). This uses the decoder KV cache....