DeepGS_GPU #18
Description
1、When I use the CPU version, due to the large size of my files, it occupies too much memory, causing a machine with 156G of memory to be unable to run the DeepGS software normally. Is there any way to reduce memory usage?
2、When running the GPU version, using the sample data results in an error: Error in mx.model.train(symbol, ctx, input.shape, output.shape, params$arg.params, : object 'TRUE...' not found