kuntalkumarpal/r2c

Recognition to Cognition Networks

r2c

Recognition to Cognition Networks

Environment Creation

• conda create -n r2c10 python=3.6

• interactive -N 1 -n 20 -p cidsegpu1 -q cidsegpu1_other_res --gres=gpu:4 -t 0-7:00:00

• source activate r2c10

• module load cuda/9.0.176

• export LD_LIBRARY_PATH=/packages/7x/cuda/9.0.176:$LD_LIBRARY_PATH

• conda install numpy setuptools cmake cffi tqdm pyyaml scipy ipython mkl mkl-include cython typing h5py pandas nltk spacy numpydoc scikit-learn jpeg

• pip install -r allennlp-requirements.txt

• conda uninstall PyYAML

• pip install -r allennlp-requirements.txt

• pip install --no-deps allennlp==0.8.0

• python -m spacy download en_core_web_sm

• conda install pytorch torchvision cudatoolkit=9.0 -c pytorch

Checks

• torch.version
  '1.1.0'

• torchvision.version
  '0.3.0'

• Currently Loaded Modulefiles:
  1. cudnn/7.0 2) anaconda3/5.3.0 3) cuda/9.0.176

Adding other file (Follow rowan's data folder README)

• Image and annots files addition

1. Get the *.h5 files into the data folder
2. Get the vcr1images into the data folder
3. Get the 3 jsonl files into the data folder

Execution

• module load cuda/9.0.176
• export LD_LIBRARY_PATH=/packages/7x/cuda/9.0.176:$LD_LIBRARY_PATH
• export PYTHONPATH=~/r2c-latest/r2c
• cd r2c/models/
• python train.py -params multiatt/default.json -folder saves/flagship_answer

Current Issue/Status :

• Running for 1 hr and then giving memory error
  Warning ⚠️ (Need to solve this issue)

/home/kkpal/.conda/envs/r2cw/lib/python3.6/site-packages/torch/nn/_reduction.py:46: UserWarning: size_average and reduce args will be deprecated, please use reduction='mean' instead. warnings.warn(warning.format(ret))

/home/kkpal/.conda/envs/r2cw/lib/python3.6/site-packages/torch/nn/modules/rnn.py:525: RuntimeWarning: RNN module weights are not part of single contiguous chunk of memory. This means they need to be compacted at every call, possibly greatly increasing memory usage. To compact weights again call flatten_parameters(). self.num_layers, self.dropout, self.training, self.bidirectional)