root@pynq:/home/xilinx# cat tvm/apps/vta_rpc/start_rpc_server_to_tracker.sh
#!/bin/bash
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# re
Hi @thkim
I use `su` to start the tracker and modify the corresponding environment
variables.
```
TVM_HOME="/home/xilinx/tvm"
VTA_HW_PATH="/home/xilinx/tvm/3rdparty/vta-hw"
PYTHONPATH="/home/xilinx/tvm/python:/home/xilinx/tvm/topi/python:/home/xilinx/tvm/vta/python"
---
[Visit
Topic](htt
Make sure you start remote device server with root because writing bitstream
need root authority.
---
[Visit
Topic](https://discuss.tvm.apache.org/t/vta-workaround-for-autotuning-with-one-pynq-z1-board/8091/13)
to respond.
You are receiving this because you enabled mailing list mode.
To
Check `dmesg` from your remote device. Sometimes 0.00 GFLOPS means no
bitstream writtten to FPGA.
---
[Visit
Topic](https://discuss.tvm.apache.org/t/vta-workaround-for-autotuning-with-one-pynq-z1-board/8091/12)
to respond.
You are receiving this because you enabled mailing list mode.
T
You can move the 'remote' block behind tuning process to get inference time
cost. Your result shows 'Upload', so I guess you use the inference time cost
block. In my code, I just uncomment these blocks.
---
[Visit
Topic](https://discuss.tvm.apache.org/t/vta-workaround-for-autotuning-with
0.7 release 2020-10-02
---
[Visit
Topic](https://discuss.tvm.apache.org/t/vta-workaround-for-autotuning-with-one-pynq-z1-board/8091/4)
to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click
here](https://discuss.tvm.apache.org/
@hjiang, thanks for your reply. I am glad the bug has been fixed.
---
[Visit Topic](https://discuss.tvm.apache.org/t/rfc-vta-a-hls-c-vta-bug/6743/5)
to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click
here](https://discuss.tv
https://discuss.tvm.apache.org/t/vta-autotuning-from-tutorial-fails-with-one-pynq-but-succeeds-with-two-pynqs/4265/3?u=hht
I find the workaround for autotuning with one PYNQ and locate the problem.
In the VTA autotuning tutorial, there is a handle named `remote`.
The `remote` does two things. On
I found the problem had been reported a year before.
https://discuss.tvm.apache.org/t/autotvm-cannot-get-remote-devices-from-tracker/3453?u=hht
cc @youn123 @diamantopoulos @zhanghaohit @c
---
[Visit
Topic](https://discuss.tvm.apache.org/t/vta-workaround-for-autotuning-with-one-pynq-z1-b
