Created by: shoegazerstella

I am using the available docker image. After fixing this issue I have this error when trying sh run_train.sh

grep: warning: GREP_OPTIONS is deprecated; please use an alias or script
-----------  Configuration Arguments -----------
augment_conf_path: conf/augmentation.config
batch_size: 4
dev_manifest: data/tiny/manifest.tiny
init_from_pretrained_model: None
is_local: 1
learning_rate: 1e-05
max_duration: 27.0
mean_std_path: data/tiny/mean_std.npz
min_duration: 0.0
num_conv_layers: 2
num_epoch: 20
num_iter_print: 1
num_rnn_layers: 3
num_samples: 64
output_model_dir: ./checkpoints/tiny
rnn_layer_size: 2048
save_epoch: 1
share_rnn_weights: 1
shuffle_method: batch_shuffle_clipped
specgram_type: linear
test_off: 0
train_manifest: data/tiny/manifest.tiny
use_gpu: 1
use_gru: 0
use_sortagrad: 1
vocab_path: data/tiny/vocab.txt
------------------------------------------------
E0121 09:57:47.763219    62 pybind.cc:1190] Cannot use GPU because there is no GPU detected on your machine.
Failed in training!

It seems starting from this docker image that i can't properly see the GPU, I am on a server with CUDA 10.2 but inside the docker there is CUDA 9 installed.

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Sep__1_21:08:03_CDT_2017
Cuda compilation tools, release 9.0, V9.0.176

What's another docker image I can pull from that works on CUDA 10? Thanks!