summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2015-08-30Merge pull request #2998 from longjon/data-raceJeff Donahue1-1/+3
Fix a recently introduced race condition in DataLayer
2015-08-29Merge pull request #2894 from mfigurnov/fix-truncation-warningRonghang Hu1-1/+2
Fix truncation of value warning
2015-08-29fix GPU data raceJonathan L Long1-1/+3
Previously, the prefetch GPU -> top GPU and prefetch CPU -> prefetch GPU copies were launched concurrently in separate streams, allowing the next batch to be copied in before the current one is read. This patch explicitly synchronizes the prefetch -> top copy wrt the host, preventing the CPU -> GPU from being launched until its completion.
2015-08-27Merge pull request #2989 from jyegerlehner/embed-layer-compile-warningJeff Donahue1-1/+0
Fix EmbedLayer compiler warning for unused variable.
2015-08-27Fix EmbedLayer compiler warning for unused variable.J Yegerlehner1-1/+0
2015-08-27Add extra openblas search pathMatt Dawkins1-0/+2
2015-08-26Merge pull request #2944 from philkr/python_layer_paramJon Long2-1/+71
Give the python layer parameter/weight blobs.
2015-08-26Merge pull request #2970 from ronghanghu/spp-fixRonghang Hu2-7/+39
Fix SPPLayer top blob num and address `pyramid_height_ == 1`
2015-08-26Fix previous mistake on unimplemented top and address pyramid_height_==1 in ↵Ronghang Hu2-7/+39
SPPLayer also, do nothing in SPPLayer Reshape if already reshaped once and bottom size unchanged
2015-08-26Merge pull request #2964 from jyegerlehner/mvn-layer-fixesJeff Donahue3-8/+27
Fix MVNLayer
2015-08-26Merge pull request #2981 from maaskola/draw-deconvolutionJeff Donahue1-3/+3
Draw Deconvolution layers like Convolution layers
2015-08-26Draw Deconvolution layers like Convolution layersJonas Maaskola1-3/+3
2015-08-26MVNLayer fixes.J Yegerlehner3-8/+27
Fix the MVNLayer tests so they actually test what they claim. MVNLayer fixes: sum_multiplier_ sized correctly; backward gradient calculation. Gradient calculation per analysis of seanbell, found here: https://github.com/BVLC/caffe/issues/1938 Fixes according to review comments.
2015-08-25Merge pull request #2083 from jeffdonahue/tile-layerJeff Donahue5-1/+332
TileLayer
2015-08-25TileLayer: add CUDA kernelsJeff Donahue1-14/+39
2015-08-25Add TileLayerJeff Donahue5-1/+307
2015-08-25Merge pull request #2032 from jeffdonahue/embed-layerEvan Shelhamer7-4/+488
Embed layer for lookup table of one hot encodings
2015-08-25Merge pull request #2927 from jeffdonahue/improve-net-init-error-msgsEvan Shelhamer2-7/+33
improve net config and shape mismatch error messages
2015-08-25Merge pull request #2972 from jeffdonahue/concat-backward-fixEvan Shelhamer3-14/+25
[fix] properly backprop through ConcatLayer with propagate_down set
2015-08-25Python parameter test addedphilkr2-2/+60
2015-08-25Allow the python layer have weight/parameter blobs.philkr1-2/+14
2015-08-24Merge pull request #2963 from longjon/superfluous-toprotoJeff Donahue1-6/+0
Remove superfluous code in Net::ToProto
2015-08-24bugfix for ConcatLayer with propagate_down setJeff Donahue2-14/+16
if propagate_down[i] was set, offset_concat_axis was not correctly updated for subsequent bottoms i+1, i+2, ...
2015-08-24TestConcatLayer: add gradient check for bottom[1] only (to verifyJeff Donahue1-0/+9
propagate_down[0] == false correctness)
2015-08-22Merge pull request #2935 from rmanor/accuraciesRonghang Hu3-1/+134
Output accuracies per class.
2015-08-22Merge pull request #2253 from jyegerlehner/snapshot_on_signalRonghang Hu5-10/+268
Snapshot on signal
2015-08-23Output accuracies per class.Ran3-1/+134
Fixed case where number of samples in class can be zero. - Fixed ignore_label case, also added a test. - Two other fixes. Fixed lint errors. Small fix.
2015-08-22Add signal handler and early exit/snapshot to Solver.J Yegerlehner5-10/+268
Add signal handler and early exit/snapshot to Solver. Add signal handler and early exit/snapshot to Solver. Also check for exit and snapshot when testing. Skip running test after early exit. Fix more lint. Rebase on master. Finish rebase on master. Fixups per review comments. Redress review comments. Lint. Correct error message wording.
2015-08-21remove superfluous code in Net::ToProtoJonathan L Long1-6/+0
2015-08-20Merge pull request #2949 from jeffdonahue/deconv-gpu-backward-no-repeat-im2colJon Long1-1/+2
DeconvolutionLayer Backward_gpu fix: don't redo im2col
2015-08-20Merge pull request #2950 from lukeyeager/use-input_shapeJon Long11-44/+66
Use input_shape instead of input_dim in examples
2015-08-20Use input_shape instead of input_dim in examplesLuke Yeager11-44/+66
2015-08-20Merge pull request #2947 from BlGene/bdl_fixupJeff Donahue1-2/+1
BasePrefetchingDataLayer fixup.
2015-08-20DeconvolutionLayer Backward_gpu fix: don't redo im2colJeff Donahue1-1/+2
2015-08-20Merge pull request #2930 from lukeyeager/pycaffe-layer_type_listJeff Donahue4-11/+34
Expose LayerFactory::LayerTypeList in pycaffe
2015-08-20Expose LayerFactory::LayerTypeList in pycaffeLuke Yeager4-11/+34
Useful for validating NetParameters without crashing on SIGABRT
2015-08-20In BasePrefetchingDataLayer::Forward_cpu hanged top[0]->Reshape tomax argus1-2/+1
top[0]->ReshapeLike, in line with other calls.
2015-08-18Add information about how to get GPU topology from nvidia-smimhouston1-0/+2
2015-08-18Add some documentation on Multi-GPU supportmhouston1-0/+24
2015-08-16Merge pull request #2812 from philkr/python_loss_weightJon Long2-0/+14
Expose blob loss weight to python
2015-08-15Malloc at least 1 byte for MultiGPU P2PSync buffersRonghang Hu1-1/+2
2015-08-15Merge pull request #2931 from ronghanghu/fix-gpu-testRonghang Hu1-2/+13
Fix MultiGPU solver test crash
2015-08-14Fix MultiGPU solver test with TEST_GPUID != 1Ronghang Hu1-2/+13
This is a patch for multi-gpu testing issue (#2926). The problem fixed in this commit is that when calling make runtest with TEST_GPUID != 0 on a MultiGPU machine, solver tests will crash because gpu ids in multi-gpu tests doesn't match that of single GPU test.
2015-08-14Merge pull request #2928 from cypof/lib_make_targetJeff Donahue1-2/+4
New make target to only build the library.
2015-08-14New make target to only build the library.Cyprien Noel1-2/+4
2015-08-14[net] improve net config and shape mismatch error messagesJeff Donahue2-7/+33
2015-08-14Exposing blob loss weight to pythonphilkr2-0/+14
2015-08-14Merge pull request #2925 from flx42/fix_classification_exampleJeff Donahue1-0/+2
Make classification.bin support models with less than 5 classes
2015-08-14Make classification.bin support models with less than 5 classesFelix Abecassis1-0/+2
The example program would crash if the number of classes was less than 5, since it was still attempting to get the top 5 predictions. Close #2585
2015-08-14Merge pull request #2921 from buaaliyi/multi_gpuRonghang Hu1-2/+7
Destroy CUDA stream when finished
generated by cgit v1.2.3 (git 2.25.1) at 2024-10-16 17:34:18 +0000