pytorch

d51f6de9 - [quant][core][feature] Implement index_put for quantized CUDA tensors (#85685)

[quant][core][feature] Implement index_put for quantized CUDA tensors (#85685) Summary: - Add new cuda test for quantized index_put - Add determinsitc test for CPU and CUDA quantized index_put - Add in QuantizedCUDA implementation for index_put - wrote new `index_put_kernel_quantized_cuda` - CUDA index_put determinstic implemented in `index_put_with_sort_kernel_quantized` I think quantize_val<scalar_t> is not CUDA compatible, because of the reliance on std::numeric_limits. Might be something useful to add in the future? Test Plan: ``` python test/test_quantization.py -k test_qtensor_index_put ``` Reviewers: Subscribers: Tasks: Tags: quant Pull Request resolved: https://github.com/pytorch/pytorch/pull/85685 Approved by: https://github.com/dzdang