Phaneesh Barwaria

[spirv] PyTorch AlexNet returns zeros on Apple M GPUs

3

### What happened? On running Alexnet infernece on macos with SHARK, we see the output tensor is zero. ### Steps to reproduce your issue mlir: `https://storage.googleapis.com/shark_tank/alexnet_torch/alexnet_torch.mlir` input: `https://storage.googleapis.com/shark_tank/alexnet_torch/inputs.npz` golden_output: `https://storage.googleapis.com/shark_tank/alexnet_torch/golden_out.npz`...

[spirv] PyTorch SqueezeNet returns NaN values on Apple M GPUs

6

### What happened? On comparing the results obtained from PyTorch with SHARK results, it was found that the difference in values is more than the acceptable tolerance range. Here's the...

[spirv] Inaccurate TF ConvBert result on Apple M GPUs

6

### What happened? On comparing the results obtained from TensorFlow with SHARK results the difference is more than the tolerance range. Following is the error message shown: ``` AssertionError: Not...

[M1 / M2] [Vulkan] Shader Init failed for MobileBert with PyTorch on SHARK

2

### What happened? It seems there is a buffer which was not correctly resolved causing the failure in shader init. Here's the std err output: ``` ----------------------------- Captured stderr call...

xfail vulkan unsupported op tests

1

Support for some arith ops have been removed from llvm -xfail all related test -add Mac Vulkan test to CI

use --device flag to choose single vulkan device

Just a check to see how many tests in vulkan actually fail.

llama2 - FP16 - IR generation

1

- fix combine mlir for llama2 for fp16 IR generation by keeping deduplicating constant variable names using appropriate suffix for first llama and second llama - needs to be tested...

Just uploading what I have till now [Gist](https://gist.github.com/PhaneeshB/761f75f5522d9f4a40ef949a328e93fe) of pytorch impl that I'm following to implement the OnnxToTorch lowering Additional Details - (also pasted as comment in gist) [Op Description](https://github.com/onnx/onnx/blob/main/docs/Operators.md#melweightmatrix)...

### What happened? For Torch-mlir bump in IREE #18169 Runtime failure - input OUT_OF_RANGE for test_maxpool_2d_uint8 on GPU NVidia Vulkan Passes in GPU Nvidia Cuda, CPU, GPU AMD (rocm +...