beniz/deepdetect

Deep Learning API and Server

tinco

@tinco:matrix.org

https://user-images.githubusercontent.com/1381301/66535560-d3422200-eace-11e9-9123-5535d469db19.png

Emmanuel Benazera

@beniz

what we'd do with DD is detection with refinedet_512 then apply a segmenter to every object, using a chain, but that's different than detectron2 that does both in a single pass.

tinco

@tinco:matrix.org

[m]

ah right

so why did you deprecate detectron2, was it not used a lot or does it perform less well than the alternatives more most use cases?

Emmanuel Benazera

@beniz

because pytorch did basically deprecate caffe2

tinco

@tinco:matrix.org

[m]

ahh alright, thanks for catching me up haha :D

Emmanuel Benazera

@beniz

Also, very don't see may semantic segmentation models with our customers. I believe this is due to labeling costs. We like to automate the labeling steps ;)

tinco

@tinco:matrix.org

[m]

we're paying students haha, and our customers are actually paying for the labeling already, we're trying to optimize the process

Emmanuel Benazera

@beniz

sure

tinco

@tinco:matrix.org

[m]

so what's your business model, do you sell consultancy around deepdetect, or is it a tool you use to implement machine learning at your clients?

Emmanuel Benazera

@beniz

yes, we are mostly a service company, we serve large corps mostly on complex problems, when there's no product on the shelves, or when there's not much litterature on whether a problem can be solved with ML/DL/RL. DD is the tool that embeds everything we have solved, and that goes into production.

tinco

@tinco:matrix.org

[m]

very cool, thanks for sharing!

Emmanuel Benazera

@beniz

no worries, we've got several requests for yolo models recently, so they might make it into the framework soon.

As for semantic seg, the path for us will be through our torch C++ backend, here again depending on requests and usage.

you can open issues on github for feature requests

tinco

@tinco:matrix.org

[m]

hey, so I just noticed that yolov5 is in pytorch, does that mean the model could just be dropped through a model repository in deep detect, or will there be some code needed as well?

Emmanuel Benazera

@beniz

almost... we've looked at it recently, and the ultralytics repo has code that makes it a bit more tricky, typically there's a bbox filtering step that they did put outside the model, which is weird, and that would need to recoded, that's for the detail.

we've got the request several times now, so we'll try to have an answer to yolov5 :)

tinco

@tinco:matrix.org

[m]

there's no python in deepdetect at all is there? if there was it would be a cool feature to have python based plugins that you could use to preprocess/post process data and add support for little things like that, though of course that's a never ending story with native extensions and such

Emmanuel Benazera

@beniz

it's full C++ yes, there's a python client.

Ananya Chaturvedi

@ananyachat

Hi, I am having trouble in following the instructions on the quickstart page of deep detect. I am using the option "build from source (Ubuntu 18.04 TLS)".

At the step with cmake command after moving to the folder /deepdetect/build, I am getting an error that "Building with tensorflow AND torch can't be build together". I am getting this error no matter what backend option I choose.

P.S.: I have macbook, so in order to use linux on my laptop, I am using a virtual linux instance created by my company for me.

Can someone please help me with this?

above is the screenshot of the error message I am getting

Emmanuel Benazera

@beniz

Hi, can you share your exact cmake call and the log that follows please ?

Ananya Chaturvedi

@ananyachat

Hi, this is the cmake call which gave the above error message:

cmake .. -DUSE_SIMSEARCH=ON -DUSE_CPU_ONLY=ON -DUSE_TF=ON -DCUDA_USE_STATIC_CUDA_RUNTIME=OFF -DUSE_CAFFE=OFF

I also tried with the GPU computation:

cmake .. -DUSE_SIMSEARCH=ON -DUSE_CUDNN=ON -DUSE_TF=ON -DCUDA_USE_STATIC_CUDA_RUNTIME=OFF -DUSE_CAFFE=OFF

Still getting the same error message.

Emmanuel Benazera

@beniz

I've just tested it and this line works fine for me. However, we discourage using our Tensorflow build. It's basically deprecated as everything has transitioned to Pytorch C++.

Ananya Chaturvedi

@ananyachat

I tried that too

Emmanuel Benazera

@beniz

then your problem is elsewhere

what's your cmake version ? cmake --version

Ananya Chaturvedi

@ananyachat

it is 3.14

do you think it could be because I using a linux instance on a macbook instead of an actual linux os?

Louis Jean

@Bycob

USE_TORCH, USE_TF etc are persistent between cmake calls. Did you try to switch them on the same command line, e.g
cmake .. -DUSE_SIMSEARCH=ON -DUSE_CPU_ONLY=ON -DCUDA_USE_STATIC_CUDA_RUNTIME=OFF -DUSE_CAFFE=OFF -DUSE_TORCH=ON -DUSE_TF=OFF

Emmanuel Benazera

@beniz

or rm -rf of the build dir yes...

Louis Jean

@Bycob

Normally cmake displays a list of all the options, but it may not be the case if there is an error during configure step...

Something like this

-- RELEASE:               OFF
-- WARNING:               ON
-- USE_COMMAND_LINE:      ON
-- USE_JSON_API:          ON
-- USE_HTTP_SERVER:       ON
-- USE_HTTP_SERVER_OATPP: ON
-- BUILD_SPDLOG:          ON
-- BUILD_PROTOBUF:        ON
-- BUILD_TESTS:           ON
-- BUILD_TOOLS:           OFF
-- USE_CAFFE:             OFF
-- USE_CAFFE_CPU_ONLY:    OFF
-- USE_CAFFE_DEBUG:       OFF
-- USE_CAFFE2:            OFF
-- USE_CAFFE2_CPU_ONLY:   OFF
-- USE_TORCH:             ON
-- USE_TORCH_CPU_ONLY:    OFF
-- USE_SIMSEARCH:         OFF
-- USE_TF:                OFF
-- USE_TF_CPU_ONLY:       OFF
-- USE_NCNN:              OFF
-- USE_HDF5:              ON
-- USE_TENSORRT:          OFF
-- USE_DLIB:              OFF
-- USE_DLIB_CPU_ONLY:     OFF
-- USE_ANNOY:             OFF
-- USE_FAISS:             ON
-- USE_FAISS_CPU_ONLY:    OFF
-- USE_CUDNN:             OFF
-- USE_XGBOOST:           OFF
-- USE_XGBOOST_CPU_ONLY:  ON
-- USE_TSNE:              OFF
-- USE_BOOST_BACKTRACE:   ON

Ananya Chaturvedi

@ananyachat

USE_TORCH, USE_TF etc are persistent between cmake calls. Did you try to switch them on the same command line, e.g
cmake .. -DUSE_SIMSEARCH=ON -DUSE_CPU_ONLY=ON -DCUDA_USE_STATIC_CUDA_RUNTIME=OFF -DUSE_CAFFE=OFF -DUSE_TORCH=ON -DUSE_TF=OFF

Okay this suggestion by @BynaryCobweb gave a different error message that "Could NOT find CUDNN (missing: CUDNN_LIBRARY CUDNN_INCLUDE_DIR)". So may be all the problem is there because I have not installed cudnn.

I'll try doing that

thanks

Emmanuel Benazera

@beniz

you don't need CUDNN if you have no GPU

For a CPU only build, this should work fine :

cmake .. -DUSE_CAFFE=OFF -DUSE_TORCH=ON -DBUILD_TESTS=OFF -DUSE_CPU_ONLY=ON -DUSE_HTTP_SERVER_OATPP=ON -DUSE_HTTP_SERVER=OFF -DBUILD_SPDLOG=ON

you may want to completely wipe your build/ dir and start a build from scratch.

Ananya Chaturvedi

@ananyachat

oh okay

I'll try to start a build from scratch

thanks

It worked!

thanks a ton

Emmanuel Benazera

@beniz

you're welcome

Where communities thrive

People

Repo info

Activity