Huang Haiduo


Stride of TorchSparse Conv3d is no effect?

2

Hello, I have a problem about Stride of TorchSparse Conv3d. Through experiments, I verify that stride(2) does not have scale influence on the output coordinates. In other words, STRIDE does...

A bug may have been resolved.

2

There is a bug [here](https://github.com/megvii-research/Sparsebit/blob/d8ebc7998034d18b12e126abc66cb9b00848a687/sparsebit/quantization/quant_model.py#L178), I thought of a simple way to fix it, which is applicable to QAT of ViT. ``` elif 'input_quantizer.scale' in dict(_module.state_dict()).keys(): _module.input_quantizer.set_fake_fused() # 有bug, quant_state会来回切....

https://github.com/hahnyuan/PTQ4ViT/blob/b42c790c2811e7835d435a9f9a6c64a079a7483a/quant_layers/matmul.py#L324 Is it because the effect is not good?

I'm looking forward to it that "add explanation and illustration for Cuda kernel"!!! ):

Thank you for your excellent work. 1. I think [**"scores_normalized = scores.log_softmax(dim=-1)"**](https://github.com/voidism/DoLa/blob/11b73b74ec1a72216e3c97c587177d72d8288f8f/dola.py#L113C8-L113C56) is redundant because **“scores”** has already been passed through log_softmax (i.e., "final_logits = final_logits.log_softmax(dim=-1)"). 2. When you fix...

[Feature] TTT implementation is not the original paper’s concept, but merely data augmentation

6

### Checklist - [x] 1. If the issue you raised is not a feature but a question, please raise a discussion at https://github.com/sgl-project/SpecForge/discussions/new/choose Otherwise, it will be closed. - [x]...