I am trying to implement the code provided in the following link: https://tvm.d2l.ai/chapter_cpu_schedules/depthwise_conv.html
Everything works, except when I get to the following code block:
mod = tvm.build(sch, args, target) ctx = tvm.context(target, 0) data, weight, out = d2ltvm.get_conv_data( c, c, n, k, p, s, lambda x: tvm.nd.array(x, ctx=ctx), conv_type='depthwise') mod(data, weight, out) data, weight, bias, out_mx = d2ltvm.get_conv_data_mxnet(c, c, n, k, p, s, conv_type='depthwise') d2ltvm.depthwise_conv_mxnet(data, weight, bias, out_mx, k, p, s) np.testing.assert_allclose(out_mx.asnumpy(), out.asnumpy(), atol=1e-5)
Specifically, the line ctx = tvm.context(target, 0) seems to be throwing an error. The error is shown below:
I tried searching online, and in this forum, but was unable to find a fix. Please let me know what I am doing wrong! I apologize if this has been asked before/is a known issue!