Ruonan Wang

Can T2I-Adapter be used for SD-2.x？

5

I wonder whether T2I-Adapter could be used for SD-2.x ?

MKL FP16 GEMM crash on MTL iGPU

4

# Summary I found on MTL iGPU, if I call FP16 gemm of onemkl (no matter using OneAPI 2024.0 or 2024.2), the program will crash, and if I call it...

## Description ### 1. Why the change? ### 2. User API changes ### 3. Summary of the change ### 4. How to test? - [ ] N/A - [ ]...

Failed to run on Intel GPUs

2

# Prerequisites Please answer the following questions for yourself before submitting an issue. - [x] I am running the latest code. Development is very rapid so there are no tagged...

## Description ### 1. Why the change? - Remove some unused code - Simplify `padded_causal_mask` logic, but don't see obvious performance change ### 2. User API changes No change. ###...

aten::bmm op is much more slower in float16 for llm rest token generation

5

### Describe the bug Hi all, when I do some profile with open-llama-3b on Arc A770, I found in float16, aten::bmm becomes extramely slower compared to float32 (111.4ms vs 22.5ms)....