| GPU coords pruning error fix, not enough capacity | 5 年前 |
| Templated coords map & sparse tensor quantization mode Squashed commit of the following: commit da5a1bbc6177813466b92c813926d077f5f1398f Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Fri Apr 10 01:15:12 2020 -0700 sparse tensor with automatic feature average for duplicate coords commit 853488855afa225c2ff83947a902118dd0f90c29 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Wed Apr 8 01:55:47 2020 -0700 coords map instantiations and fixes commit 28fbde92165f8441e47fbc3b5c60821e2f74eea5 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Sat Apr 4 01:26:36 2020 -0700 Coords MapType commit 3df3a03f45cfcbd4aea58cadd26b8e65f8f9e521 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Thu Apr 2 18:43:45 2020 -0700 restructure coordsmap init commit 496d8f84ee829e050bdd34209c23fed7e293a7d6 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Thu Apr 2 14:50:28 2020 -0700 quantization | 6 年前 |
| kernel region to_gpu with shared_ptr | 5 年前 |
| broadcast cpu/gpu | 5 年前 |
| replace at::cuda::getCurrentCUDASparseHandle with custom func (fix #308) (#315) * force initialize cusparse handle * replace all at::cuda::getCurrentCUDASparseHandle with custom func * change log | 5 年前 |
| more thrust exception handling (#357) | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| interpolation cpu/cuda | 5 年前 |
| pytorch 1.8.1 support with gpu_storage | 5 年前 |
| conv with emtpy tensor (fix #384) | 4 年前 |
| conv with emtpy tensor (fix #384) | 4 年前 |
| hotfix: fix stride map with input tensor stride (fix #343) | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| gpu convolution | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| kernel region WIP: kernel_region region test gpu kernel region | 5 年前 |
| union map | 5 年前 |
| Fix Darwin&CPU build (#438) * Fix C++17 for build * Mac CPU library fixes * Convert long to int64_t to fix clang build | 4 年前 |
| Generative Convolution Transpose cpu/gpu | 5 年前 |
| more thrust exception handling (#357) | 5 年前 |
| hotfix: fix stride map with input tensor stride (fix #343) | 5 年前 |
| explicit coordinatemapkey hash | 5 年前 |
| Fix Darwin&CPU build (#438) * Fix C++17 for build * Mac CPU library fixes * Convert long to int64_t to fix clang build | 4 年前 |
| orgin map thrust transform to direact kernel call | 5 年前 |
| tfield max pool cpu/gpu | 5 年前 |
| direct max pool | 5 年前 |
| dispatcher cleanup, pytorch requirement update | 5 年前 |
| convolution transpose cpu/gpu | 5 年前 |
| global max fix for batchsize 1 | 5 年前 |
| global max fix for batchsize 1 | 5 年前 |
| pytorch 1.8.1 support with gpu_storage | 5 年前 |
| more thrust exception handling (#357) | 5 年前 |
| remove tfield out_key | 5 年前 |
| coo_spmm to return correct shape for nnz==0 (fix #383) | 4 年前 |
| interpolation cpu/cuda | 5 年前 |
| more thrust exception handling (#357) | 5 年前 |
| origin kernel map decomposition to kernel_map.hpp | 5 年前 |
| add cross-shaped kernel support (#437) Co-authored-by: karbo lei <karbo@lei.online> | 4 年前 |
| direct max pool | 5 年前 |
| direct max pool | 5 年前 |
| transposed pool cpu/gpu | 5 年前 |
| replace at::cuda::getCurrentCUDASparseHandle with custom func (fix #308) (#315) * force initialize cusparse handle * replace all at::cuda::getCurrentCUDASparseHandle with custom func * change log | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| interpolation cpu/cuda | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| BLAS configuration, minor Makefile updates Squashed commit of the following: commit de81c762e27c1ca1620e5de02e01531bf61cf2a7 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Mon Dec 30 01:20:07 2019 -0800 cxx optional commit 7f8e9d33673d7e09db2b9624891a3f6dc077bf53 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Sun Dec 29 02:25:09 2019 -0800 update gitignore commit c3f41650544bc9a20a58f738eb6cea28bbff8b40 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Sun Dec 29 02:20:16 2019 -0800 openblas compilation fix commit 2e17bac4cb883f42d1ec767fad7b0f4b996622ed Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Sun Dec 29 02:02:09 2019 -0800 fix minor compilation error commit 6d1be0eecd129d3789bedc4e1fce119da020894e Author: Chris Choy <chrischoy@Hyuns-MacBook-Pro.local> Date: Sun Dec 29 01:46:44 2019 -0800 gitignore update commit d39c7ffeecccfbe5bff812bccf7d026486dac4c9 Author: Chris Choy <chrischoy@ai.stanford.edu> Date: Sun Dec 29 01:44:13 2019 -0800 BLAS support commit 94f3b9969bcd06e54f1f1a8b64781d6db32225aa Author: Chris Choy <chrischoy@Hyuns-MacBook-Pro.local> Date: Sun Dec 29 00:20:51 2019 -0800 mac build | 6 年前 |
| more thrust exception handling (#357) | 5 年前 |
| gradcheck update, convolution mode | 5 年前 |
| Fix Darwin&CPU build (#438) * Fix C++17 for build * Mac CPU library fixes * Convert long to int64_t to fix clang build | 4 年前 |
| more thrust exception handling (#357) | 5 年前 |
| direct max pool | 5 年前 |
| direct max pool | 5 年前 |
| interpolation cpu/cuda | 5 年前 |
| pruning gpu | 5 年前 |
| pruning exception handling, 0 keep | 5 年前 |
| Fix Darwin&CPU build (#438) * Fix C++17 for build * Mac CPU library fixes * Convert long to int64_t to fix clang build | 4 年前 |
| interpolation cpu/cuda | 5 年前 |
| docker build fix | 3 年前 |
| spmm average | 5 年前 |
| hotfix: fix stride map with input tensor stride (fix #343) | 5 年前 |
| tfield sparse_insert, caching inverse mapping | 5 年前 |
| Fix Darwin&CPU build (#438) * Fix C++17 for build * Mac CPU library fixes * Convert long to int64_t to fix clang build | 4 年前 |