See below for full command line and output log: See below for full command line and output log:\n\n%s\n\n%s' % (cmd, output)) Raise RuntimeError('NVCC returned an error. _run_cmd(nvcc_cmd + ' "%s" -shared -o "%s" -keep -keep-dir "%s"' % (cuda_file, tmp_file, tmp_dir))įile "/var/www/dnnlib/tflib/custom_ops.py", line 69, in _run_cmd Return custom_ops.get_plugin(os.path.splitext( file) + '.cu')įile "/var/www/dnnlib/tflib/custom_ops.py", line 159, in get_plugin Return impl_dict(x=x, b=b, axis=axis, act=act, alpha=alpha, gain=gain, clamp=clamp)įile "/var/line 18, in _get_plugin Return fused_bias_act(x, b=tf.cast(b, x.dtype), act=act, gain=gain, clamp=clamp)įile "/var/www/dnnlib/tflib/ops/fused_bias_act.py", line 72, in fused_bias_act S = apply_bias_act(s, bias_var='mod_bias', trainable=trainable) + 1 # Add bias (initially 1).įile "/var/www/training/networks.py", line 50, in apply_bias_act X = modulated_conv2d_layer(x, dlatents_in, fmaps=fmaps, kernel=kernel, up=up, resample_kernel=resample_kernel, fused_modconv=fused_modconv)įile "/var/www/training/networks.py", line 105, in modulated_conv2d_layer X = layer(x, layer_idx=0, fmaps=nf(1), kernel=3)įile "/var/www/training/networks.py", line 392, in layer Self._input_shapes = įile "/var/line 439, in G_synthesis Num_layers = _shapeįile "/var/line 219, in input_shapes Out_expr = self._build_func(*self._input_templates, **build_kwargs)įile "/var/www/training/networks.py", line 231, in G_main Self._vars = OrderedDict(self._get_own_vars())įile "/var/line 151, in _init_graph Training_aining_loop(**training_options)įile "/var/line 457, in cloneįile "/var/line 297, in _get_vars Failed!įile "train.py", line 473, in run_training Setting up TensorFlow plugin "fused_bias_act.cu": Compiling. It didn't work initially, then I realized I have a few more steps to do, so I installed nvidia-docker2 ( nvidia-container-toolkit ) thinking that it should certainly work. It was working on my titan rtx though, on a few different computer rigs.įinally I thought that maintainers claimed it is working on their end for rtx 3000, maybe I can try their docker container. I tried different approaches by reinstalling things and wasted more than 10 hours, it never worked for me. It always said it couldn't find a gpu when trying to start training (or other errors like attempting to import cublas.10 files with a failure, while I had cuda 11 installed instead ). I tried to install nvidia driver ( 455 ) by myself on my ubuntu 18.04 with python 3.7 and tensorflow 1.14 (also tried 1.15).
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |