35b9fff | soumith | 03 December 2016, 00:33:27 UTC | fix half type numerics issue in SpatialFractionalMaxPooling | 03 December 2016, 00:33:27 UTC |
b3b9a88 | Soumith Chintala | 02 December 2016, 02:29:10 UTC | Merge pull request #389 from gchanan/gradOutputShapeChecks Improve fix/shape checks for Temporal and Spatial modules | 02 December 2016, 02:29:10 UTC |
81ca160 | Gregory Chanan | 01 December 2016, 23:42:55 UTC | Remove unnecessary shape checks in Spatial Pooling modules. Checks comparing input image sizes to kernel sizes are superseded by output size checks. | 01 December 2016, 23:49:53 UTC |
0d22308 | Gregory Chanan | 29 November 2016, 20:19:56 UTC | Add gradOutput shape checks in temporal modules. | 01 December 2016, 23:49:48 UTC |
bb8e462 | Gregory Chanan | 29 November 2016, 19:28:25 UTC | Fix spacing in SpatialDilatedMaxPooling. | 01 December 2016, 23:49:41 UTC |
1ddca5f | Soumith Chintala | 01 December 2016, 17:05:02 UTC | Merge pull request #385 from gchanan/autoConfigLua Add newContiguous calls that have been removed from lua. | 01 December 2016, 17:05:02 UTC |
227bf2d | Soumith Chintala | 01 December 2016, 17:04:28 UTC | Merge pull request #386 from gchanan/VolumetricDilatedMaxPoolingTest Generate valid image sizes for VolumetricDilatedMaxPooling tests. | 01 December 2016, 17:04:28 UTC |
e1d4be2 | Gregory Chanan | 01 December 2016, 17:01:43 UTC | Generate valid image sizes for VolumetricDilatedMaxPooling tests. | 01 December 2016, 17:03:21 UTC |
92eac42 | Gregory Chanan | 30 November 2016, 21:35:07 UTC | Add newContiguous calls that have been removed from lua. | 30 November 2016, 21:58:22 UTC |
c26131c | Soumith Chintala | 28 November 2016, 19:50:59 UTC | Merge pull request #383 from gchanan/TemporalShapeCheck Improve error messages/shape check in TemporalMaxPooling. | 28 November 2016, 19:50:59 UTC |
fdb589f | Gregory Chanan | 28 November 2016, 19:46:26 UTC | Improve error messages/shape check in TemporalMaxPooling. | 28 November 2016, 19:46:26 UTC |
ac7390f | Soumith Chintala | 28 November 2016, 06:04:46 UTC | Merge pull request #382 from BTNC/win use local modified select_compute_arch.cmake for msvc | 28 November 2016, 06:04:46 UTC |
5a80d2e | Rui Guo | 28 November 2016, 06:02:21 UTC | use local modified select_compute_arch.cmake for msvc | 28 November 2016, 06:02:21 UTC |
1793771 | Soumith Chintala | 26 November 2016, 16:55:08 UTC | Merge pull request #381 from fmassa/shapecheck_fix Fix shapeCheck in Spatial Pooling modules | 26 November 2016, 16:55:08 UTC |
72208fd | Francisco Massa | 26 November 2016, 16:41:59 UTC | Fix shapeCheck in Spatial Pooling modules | 26 November 2016, 16:41:59 UTC |
b923c23 | soumith | 22 November 2016, 21:23:16 UTC | added shape checks for SpatialAveragePooling | 22 November 2016, 21:23:16 UTC |
f4391ed | Soumith Chintala | 21 November 2016, 17:11:38 UTC | Merge pull request #379 from torch/local SpatialConvolutionLocal uses baddbmm | 21 November 2016, 17:11:38 UTC |
f94eb59 | soumith | 21 November 2016, 17:10:26 UTC | SpatialConvolutionLocal uses baddbmm | 21 November 2016, 17:10:26 UTC |
50224b7 | Soumith Chintala | 17 November 2016, 22:53:48 UTC | Merge pull request #378 from gchanan/nnLinearHalf Test for nn.Linear with CudaHalfTensor and CudaDoubleTensor. | 17 November 2016, 22:53:48 UTC |
cea3855 | Gregory Chanan | 17 November 2016, 17:09:41 UTC | Test for nn.Linear with CudaHalfTensor and CudaDoubleTensor. | 17 November 2016, 22:39:12 UTC |
4422d79 | Soumith Chintala | 16 November 2016, 18:35:11 UTC | Merge pull request #377 from gchanan/checkContiguous Add contiguous checks / auto contiguous | 16 November 2016, 18:35:11 UTC |
90b75f1 | Gregory Chanan | 14 November 2016, 22:16:04 UTC | Call newContiguous for tensors that are required to be contiguous. Also add tests to verify that non-contiguous tensors are handled correctly. | 16 November 2016, 17:50:11 UTC |
e94d55c | Gregory Chanan | 15 November 2016, 22:31:36 UTC | Small test fixes for contiguous tests. | 15 November 2016, 22:31:36 UTC |
7972f85 | Gregory Chanan | 15 November 2016, 21:02:58 UTC | Handle contiguousness and improve shape checks in SpatialAdaptiveMaxPooling, SpatialUpSamplingNearest, and TemporalConvolution. | 15 November 2016, 22:17:45 UTC |
8732111 | Gregory Chanan | 15 November 2016, 17:11:39 UTC | SpatialSubSampling contiguous check. | 15 November 2016, 22:16:48 UTC |
2c57125 | Gregory Chanan | 14 November 2016, 22:10:09 UTC | VolumetricConvolution check contiguous. | 15 November 2016, 22:15:55 UTC |
f9ada9c | Soumith Chintala | 15 November 2016, 20:31:07 UTC | Merge pull request #376 from apaszke/thcunn_header Remove duplicate entries and add optional marks in THCUNN.h | 15 November 2016, 20:31:07 UTC |
b5ef7a5 | Adam Paszke | 15 November 2016, 20:22:14 UTC | Remove duplicate entries and add optional marks in THCUNN.h | 15 November 2016, 20:22:14 UTC |
941fc47 | Soumith Chintala | 12 November 2016, 16:21:02 UTC | Merge pull request #374 from howard0su/master Reference HalfTensor when it exist | 12 November 2016, 16:21:02 UTC |
fcd7b5e | Howard Su | 12 November 2016, 15:35:58 UTC | Don't reference HalfTensor when it doesn't exist | 12 November 2016, 15:35:58 UTC |
7cff722 | Soumith Chintala | 12 November 2016, 00:00:54 UTC | Merge pull request #373 from gchanan/sc2 more improvments on error messages and shape checks. | 12 November 2016, 00:00:54 UTC |
202cb36 | Gregory Chanan | 11 November 2016, 23:46:44 UTC | Remove comment. | 11 November 2016, 23:46:44 UTC |
2017698 | Gregory Chanan | 11 November 2016, 22:37:40 UTC | Fix SpatialDilatedMaxPooling shape check. In nn, indices are 3d, but they are 4d in cunn. | 11 November 2016, 23:43:54 UTC |
e8de684 | Gregory Chanan | 11 November 2016, 23:32:45 UTC | more improvements on error messages and shape checks. | 11 November 2016, 23:43:49 UTC |
499b900 | Soumith Chintala | 11 November 2016, 22:55:05 UTC | Merge pull request #372 from BTNC/win change to compile with msvc && ptrdiff_t replacement | 11 November 2016, 22:55:05 UTC |
44b6eb7 | Rui Guo | 11 November 2016, 05:33:36 UTC | change to work on windows && ptrdiff_t replacement | 11 November 2016, 05:33:36 UTC |
97e90b9 | Soumith Chintala | 11 November 2016, 00:51:26 UTC | Merge pull request #371 from gchanan/assertSameGPU Rename assertSameGPU_generic to assertSameGPU. | 11 November 2016, 00:51:26 UTC |
dc7443c | Gregory Chanan | 10 November 2016, 23:38:33 UTC | Rename assertSameGPU_generic to assertSameGPU. Also remove old assertSameGPU since there is no longer both generic and non-generic support. | 10 November 2016, 23:40:41 UTC |
b835dfa | Soumith Chintala | 10 November 2016, 23:35:22 UTC | Merge pull request #370 from gchanan/sizeCheckErrorMessages Improving error messages in nn. | 10 November 2016, 23:35:22 UTC |
195a178 | Gregory Chanan | 09 November 2016, 21:11:06 UTC | Improving error messages in nn. Differences from nn equivalent: 1) No changes to VolumetricConvolutionMM, which doesn't exist in cunn. 2) No changes to HardShrink, which doesn't exist in cunn. 3) LookupTable doesn't verify that all inputs are within range. | 10 November 2016, 23:12:35 UTC |
6cfce89 | Soumith Chintala | 10 November 2016, 21:29:00 UTC | Merge pull request #369 from gchanan/halfMulDiv Remove mul/div from THCHalfAutoNumerics | 10 November 2016, 21:29:00 UTC |
d5ab117 | Gregory Chanan | 10 November 2016, 20:12:29 UTC | Remove mul/div from THCHalfAutoNumerics as they've been moved to THCNumerics. | 10 November 2016, 20:13:41 UTC |
f27dd4f | Soumith Chintala | 10 November 2016, 17:12:25 UTC | Merge pull request #368 from gchanan/halfGenericIfdef Protect half operations with CUDA_HALF_TENSOR with generic modules. | 10 November 2016, 17:12:25 UTC |
c2fb607 | Gregory Chanan | 10 November 2016, 16:57:33 UTC | Protect half operations with CUDA_HALF_TENSOR with generic modules. | 10 November 2016, 16:59:23 UTC |
09d1a71 | Soumith Chintala | 10 November 2016, 14:57:46 UTC | Merge pull request #365 from SYSTRAN/master more robust support of small batches when input/output dimensions differ | 10 November 2016, 14:57:46 UTC |
8d5027f | Soumith Chintala | 08 November 2016, 23:18:01 UTC | Merge pull request #346 from gchanan/pointwiseGeneric Add generic support for pointwise modules | 08 November 2016, 23:18:01 UTC |
27479c3 | Gregory Chanan | 08 November 2016, 21:48:48 UTC | Remove non-generic support for modules. | 08 November 2016, 21:49:30 UTC |
604d6ff | Gregory Chanan | 04 November 2016, 18:57:08 UTC | Allow wider test tolerances for: 1) Size of half numbers 2) Convolution weight/bias 3) BatchNormalization | 08 November 2016, 21:47:01 UTC |
b0444c9 | Gregory Chanan | 03 November 2016, 16:22:27 UTC | ArgCheck that dilation parameters are > 0 and ensure tests pick dilation parameters > 0. | 08 November 2016, 21:46:52 UTC |
9fd7a88 | Gregory Chanan | 01 November 2016, 20:49:07 UTC | Add generic support for LookupTable. In some cases, does not do accumulation as accreal. | 08 November 2016, 21:46:48 UTC |
92958e4 | Gregory Chanan | 01 November 2016, 16:47:26 UTC | Add generic support for TemporalConvolution. Has increased tolerance for backward weight/bias like other Convolution modules. | 08 November 2016, 21:46:45 UTC |
86ecad3 | Gregory Chanan | 01 November 2016, 16:24:36 UTC | Add generic support for VolumetricFullConvolution, VolumetricDilatedConvolution. Has increased tolerance for backward weight/bias like other Convolution modules. | 08 November 2016, 21:46:33 UTC |
010336f | Gregory Chanan | 01 November 2016, 15:24:31 UTC | Add generic support for VolumetricConvolution. Uses the higher tolerances for weight/bias that are used for SpatialConvolution modules. | 08 November 2016, 21:07:35 UTC |
a270ee2 | Gregory Chanan | 31 October 2016, 20:12:07 UTC | Add generic support for VolumetricReplicationPadding. | 08 November 2016, 21:07:35 UTC |
7578ac1 | Gregory Chanan | 31 October 2016, 19:48:38 UTC | Add generic support for VolumetricAveragePooling. | 08 November 2016, 21:07:35 UTC |
ebb9d64 | Gregory Chanan | 31 October 2016, 17:00:47 UTC | Add generic support for VolumetricMaxPooling, VolumetricMaxUnpooling, VolumetricDilatedMaxPooling. | 08 November 2016, 21:07:35 UTC |
1c82c32 | Gregory Chanan | 31 October 2016, 17:51:41 UTC | Add generic support for TemporalMaxPooling. | 08 November 2016, 21:07:35 UTC |
aa18682 | Gregory Chanan | 08 November 2016, 21:06:52 UTC | Rebase BatchNormalization. | 08 November 2016, 21:06:52 UTC |
ba27398 | Gregory Chanan | 28 October 2016, 15:42:01 UTC | Add support for L1Cost. Changes thrust::reduce to trust::transform_reduce in order to be able to do summation at accreal precision. | 08 November 2016, 21:01:06 UTC |
bedeebf | Gregory Chanan | 26 October 2016, 21:42:28 UTC | Add generic support for SparseLinear. We don't support SparseLInear with fp16 because of lack of cusparseHcsrmm (or equivalent Ex function) until CUDA 8.0. | 08 November 2016, 21:01:06 UTC |
f43f9aa | Gregory Chanan | 26 October 2016, 20:47:03 UTC | Add generic support for DistKLDivCriterion. | 08 November 2016, 21:01:06 UTC |
44ec4ed | Gregory Chanan | 26 October 2016, 20:23:15 UTC | Add generic support for ClassNLLCriterion. | 08 November 2016, 21:01:06 UTC |
a01d7c6 | Gregory Chanan | 26 October 2016, 18:10:55 UTC | Add generic support for BCECriterion. Test skips comparing vs lua version for half type, because hdot is not currently implemented in cutorch. | 08 November 2016, 21:01:06 UTC |
125992e | Gregory Chanan | 25 October 2016, 20:08:50 UTC | Add generic support for L1SmoothCriterion. | 08 November 2016, 21:01:06 UTC |
92e335c | Gregory Chanan | 25 October 2016, 19:44:15 UTC | Add generic support for MultiLabelMarginCriterion. | 08 November 2016, 21:01:06 UTC |
2904d10 | Gregory Chanan | 25 October 2016, 17:22:45 UTC | Add generic support for MultiMarginCriterion. Accumulation is done at accreal precision and changes target tensor indexing to THCIndexTensor. | 08 November 2016, 21:01:06 UTC |
b707ed6 | Gregory Chanan | 24 October 2016, 20:47:48 UTC | Add generic support for MSECriterion. | 08 November 2016, 21:01:06 UTC |
0b3a0b0 | Gregory Chanan | 24 October 2016, 19:48:42 UTC | Add generic support for SoftMarginCriterion. | 08 November 2016, 21:01:06 UTC |
d4b4a58 | Gregory Chanan | 24 October 2016, 19:17:14 UTC | Add generic support for MarginCriterion. | 08 November 2016, 21:01:06 UTC |
8e23e5b | Gregory Chanan | 24 October 2016, 17:30:03 UTC | Add generic support for AbsCriterion. | 08 November 2016, 21:01:06 UTC |
b838e36 | Gregory Chanan | 19 October 2016, 22:17:17 UTC | Fix spacing in SpatialDilatedMaxPooling. | 08 November 2016, 21:01:06 UTC |
f5d6f30 | Gregory Chanan | 19 October 2016, 20:26:45 UTC | More leeway for convolution backward weight/bias. | 08 November 2016, 21:01:06 UTC |
5fb9b7b | Gregory Chanan | 18 October 2016, 16:05:06 UTC | Generic support for SpatialFullConvolution and SpatialDilatedConvolution. Uses matrix multiple for matrix vector multiply for half (no matrix vector implementation exists). | 08 November 2016, 21:01:06 UTC |
1a435b0 | Gregory Chanan | 17 October 2016, 22:46:22 UTC | Add generic support for SpatialFractionalMaxPooling. | 08 November 2016, 21:01:06 UTC |
be33326 | Gregory Chanan | 17 October 2016, 21:57:58 UTC | Generic support for SpatialConvolutionMM. Still need Hgemv. | 08 November 2016, 21:01:06 UTC |
1aad64f | Gregory Chanan | 17 October 2016, 19:50:44 UTC | Add generic support for SpatialConvolutionLocal. | 08 November 2016, 21:01:06 UTC |
c138aca | Gregory Chanan | 17 October 2016, 18:48:10 UTC | Add generic support for SpatialUpSamplingBilinear. Math is done at accreal precision. At real precision, forward pass fails, but backward passes. We do backward pass at accreal precision for consistency. | 08 November 2016, 21:01:06 UTC |
7e92dd2 | Gregory Chanan | 17 October 2016, 17:13:04 UTC | Add generic support for SpatialUpSamplingNearest. Accumulates as AccType. | 08 November 2016, 21:01:06 UTC |
7b191ae | Gregory Chanan | 17 October 2016, 16:42:27 UTC | Add generic support for SpatialReplicationPadding. | 08 November 2016, 21:01:06 UTC |
9e8056f | Gregory Chanan | 17 October 2016, 16:21:18 UTC | Add generic support for SpatialReflectionPooling. | 08 November 2016, 21:01:06 UTC |
0d15053 | Gregory Chanan | 14 October 2016, 23:06:21 UTC | Add generic support for SpatialSubSampling. Half types fail on backward, probably because we don't consistently accumulate in accreal. This is difficult because gradInput is accumulated directly (either with atomicAdd or not) rather than in another variable. | 08 November 2016, 21:01:06 UTC |
c077b3f | Gregory Chanan | 17 October 2016, 15:54:36 UTC | Generic support for SpatialCrossMapLRN Removed the C-linkage for a couple of functions because they are now generic -- not sure if they were used by anyone outside. | 08 November 2016, 21:01:06 UTC |
2232736 | Gregory Chanan | 14 October 2016, 20:58:55 UTC | Add generic support for SpatialAveragePooling. | 08 November 2016, 21:01:06 UTC |
b3b79b4 | Gregory Chanan | 14 October 2016, 20:38:22 UTC | Add generic support for SpatialAdaptiveMaxPooling. | 08 November 2016, 21:01:06 UTC |
8b64c92 | Gregory Chanan | 13 October 2016, 20:56:38 UTC | Use THCIndexTensors more generally. | 08 November 2016, 21:01:06 UTC |
be236c8 | Gregory Chanan | 13 October 2016, 15:46:10 UTC | Use indices for SpatialAdaptiveMaxPooling indices. | 08 November 2016, 21:01:06 UTC |
6994ae7 | Gregory Chanan | 13 October 2016, 15:16:03 UTC | Add generic support for SpatialMaxUnpooling. | 08 November 2016, 21:01:05 UTC |
30ccba8 | Gregory Chanan | 13 October 2016, 14:53:02 UTC | Fix tests | 08 November 2016, 21:01:05 UTC |
0c0e3d8 | Gregory Chanan | 12 October 2016, 22:52:52 UTC | Add generic support for SpatialMaxPooling. Also fix tests for SpatialDilatedMaxPooling. | 08 November 2016, 21:01:05 UTC |
3141686 | Gregory Chanan | 11 October 2016, 16:43:09 UTC | Get SpatialDilatedMaxPooling generic working with long tensors as index. Does as much math as possible in accreal to try to suss out why CudaHalfTensor fails. | 08 November 2016, 21:01:05 UTC |
e3d7d12 | Gregory Chanan | 10 October 2016, 20:06:59 UTC | Add generic support for SpatialDilatedMaxPooling. | 08 November 2016, 21:01:05 UTC |
3c89443 | Gregory Chanan | 07 October 2016, 22:16:47 UTC | Add generic support for SpatialClassNLLCriterion. | 08 November 2016, 21:01:05 UTC |
c08d781 | Gregory Chanan | 10 October 2016, 14:18:02 UTC | Remove fastExpIfAvail and benchmarking from functional tests. Also fix broken IFNDEF and test whitespace. | 08 November 2016, 21:01:05 UTC |
5d0c877 | Gregory Chanan | 06 October 2016, 23:07:53 UTC | Reorganize THCHalfAutoNumerics. | 08 November 2016, 21:01:05 UTC |
d892d1f | Gregory Chanan | 06 October 2016, 22:49:33 UTC | Iterate pointwise tests over all supported tensor types. | 08 November 2016, 21:01:05 UTC |
09acb86 | Gregory Chanan | 05 October 2016, 17:04:01 UTC | Add generic support for RReLU. | 08 November 2016, 21:01:05 UTC |
c6f67e1 | Gregory Chanan | 06 October 2016, 22:40:02 UTC | Add generic support for PReLU. This is the first instance of functions that take a lua number but are not reals in C. So, instead of automatically converting lua numbers in the half case, we parse the function definitions to find the argument positions to convert. | 08 November 2016, 21:01:05 UTC |
76232cb | Gregory Chanan | 10 October 2016, 16:32:27 UTC | fix logsoftmax | 08 November 2016, 21:01:05 UTC |
e65e1cf | Gregory Chanan | 10 October 2016, 16:07:08 UTC | Add generic support for LogSoftMax. | 08 November 2016, 21:01:05 UTC |
63b9beb | Gregory Chanan | 04 October 2016, 16:33:17 UTC | Add generic support for SoftMax. Math is done at accreal precision (e.g. for half, math is done at float precision). Originally code called __expf, which doesn't have a double equivalent; we call exp instead of converting down. | 08 November 2016, 21:01:05 UTC |