update caffe with the latest master version from BLVC #4
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ted caffe target This is the first step towards "modern" IMPORTED-targets-only CMake setup. The find_package modules still need to be rewritten and upstreamed in form of config exports where possible.
Despite Caffe itself does not use OpenMP, explicitly linking to OpenMP should be done when one statically links to a BLAS library which uses OpenMP internally and does not provide proper CMake imported targets with proper dependencies (nobody this so far).
Rationale: these are duplicated in CMakeLists code, and they cannot be removed from there because many definitions need to be exported to the library clients. See issue #4625.
Benchmarking should not impact perf until timer is read
A bias/scaling can be applied wherever desired by defining the respective layers, and `ScaleLayer` can handle both as a memory optimization.
Document that Ubuntu 16.04 Requires CUDA 8
batch norm statistics are not learnable parameters subject to solver updates, so they must be shielded from the solver. `BatchNorm` layer now masks its statistics for itself by zeroing parameter learning rates instead of relying on the layer definition. n.b. declaring `param`s for batch norm layers is no longer allowed.
automatically strip old batch norm layer definitions including `param`
messages. the batch norm layer used to require manually masking its
state from the solver by setting `param { lr_mult: 0 }` messages for
each of its statistics. this is now handled automatically by the layer.
[examples] Fixed typos in examples/cpp_classification/readme
Batch Norm: Further Documentation and Simplified Definition
fix layerSetUp of scale_layer to not add bias blob when already present
[TravisCI] google/protobuf renamed the 3.0 branch
Ignore Visual Studio Metadata
slightly relax batch norm check
NV changed path to cuDNN
Deprecate WindowData layer type
Test for python forward and backward with start and end layer
[docs] groom Caffe site
Docker update to cuDNN 6
Explicit std::string to bp::object conversion
Handling destruction of empty Net objects
Rewrite crop layer GPU implementation
Downgrade boost requirement from 1.55 to 1.54
docs/debian guide: update compiler combination table
…brary cmake: rename libproto.a -> libcaffeproto.a
List branches in readme
….cpp to add two headers and Blockqueue with DataReader and Datum
|
EDITED: sorry, I confused the repo. I'll check it and merge it if everything works properly. Thanks! |
|
I have a main concern... Caffe removed the However, for what I could understand after taking a look to your code, you have simply copied again the old Since then we are using two different methods to parallelize the GPUs (the one the CPM files use vs. the one everything else in Caffe uses), is this completely thread safe or even single-thread safe? I am asking because I am not an expert in NCCL not Caffe data_reader, but keep using both sounds to me like a bug-prone idea... |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Hi all:
I have update the caffe with the latest master (2017/7/26) version from BLVC.
Would you please merge it?
Thanks.
Feng