[VTA] Instruction debug info

#1

Hi,

I’ve tried to print out debug infor using

 with relay.build_config(opt_level=3, disabled_pass={"AlterOpLayout"}):
     if target.device_name != "vta":
         graph, lib, params = relay.build(
             relay_prog, target=target,
             params=params, target_host=env.target_host)
     else:
         # debugging
         with vta.build_config(debug_flag=0x6):
             graph, lib, params = relay.build(
                 relay_prog, target=target,
                 params=params, target_host=env.target_host)

However, I’m getting an error as following

Traceback (most recent call last):

  File "debug_imagenet_inference.py", line 241, in <module>
    params=params, target_host=env.target_host)

  File "/home/jwlee/tvm-zcu104-v2/tvm/python/tvm/relay/build_module.py", line 244, in build
    graph_json, mod, params = bld_mod.build(func, target, target_host, params)

  File "/home/jwlee/tvm-zcu104-v2/tvm/python/tvm/relay/build_module.py", line 109, in build
    self._build(func, target, target_host)

  File "/home/jwlee/tvm-zcu104-v2/tvm/python/tvm/_ffi/_ctypes/function.py", line 207, in __call__
    raise get_last_ffi_error()
...

AttributeError: module 'tvm.make' has no attribute 'stmt_seq'

During handling of the above exception, another exception occurred:

AttributeError: module 'tvm.make' has no attribute 'stmt_seq'

Error during compile function

Have anyone gone through this problem?

It there a parameter that must be enabeled in config.cmake?

Thanks,

Jake

#2

I think in https://github.com/apache/incubator-tvm/blob/master/vta/python/vta/build_module.py#L71, the function tvm.make.stmt_seq(debug, stmt) should be replaced with tvm.stmt.stmt_seq(debug, stmt)

2 Likes
#3

Thank you for responding.

I’ll give it a shot.

Regards,

Jake