Yujin Wang
Yujin Wang
## ❓ Questions and Help #### What is your question? Hey! I want to use different optimizers for my params in model. However, I found that when call `__init__()` function...
I have distilled a hubert via superb interface. I want to use this ckpt in fairseq finetune/decode stage. How can I do it?
## ❓ Questions and Help ### Before asking: 1. search the issues. 2. search the docs. #### What is your question? In wav2vec2.0 and hubert, the config `task.normalize` is set...
## Problem I have modified the tutorial code of `https://github.com/openai/triton/blob/main/python/tutorials/03-matrix-multiplication.py` to test int8 GEMM of Triton. Source code is as follows: ```python """ Matrix Multiplication ===================== In this tutorial, you...
Does anyone have reproduce the LoRA result of roberta-base? I found the reproduction result of LoRA cannot achieve the result that paper has claimed. e.g.: Paper claimed that the RTE...
I noticed that in the newest commit, you mentioned that all the results are implemented based on the [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness), so can you show how to evaluate the model using this...
 The default code only provide the per-tensor quantization for smoothquant model: 
For QLoRA models, we do not need to update the $\mathbf{W}$, so the buffered activation of $\mathbf{A}$ is useless. It is suggested not to save $\mathbf{A}$ in `ctx` to save...
# Prerequisites Before submitting your issue, please ensure the following: - [x] I am running the latest version of PowerInfer. Development is rapid, and as of now, there are no...
Hello, it seems that the calculator cannot record the operation num of non-linear operations like layernorm and softmax, can you support it?