huochaitiantang opened a new issue #7790: URL: https://github.com/apache/tvm/issues/7790
After PR7534(https://github.com/apache/tvm/pull/7534), while tuning an autotvm task, self.target = None, self.target_host = None (default) in class tvm.autotvm.task.Task, will introduce the error : @zxybazh ``` Traceback (most recent call last): File "test_quantize.py", line 66, in <module> tuner.tune(20, logfile=logfile) File "~/Desktop/tvms/tvm-oplib/tvmoplib/tvm_tuner/onnx_tuner.py", line 215, in tune tuner.load_history(tvm.autotvm.record.load_from_file(tmp_log_file)) File "~/Desktop/tvms/tvm/python/tvm/autotvm/tuner/model_based_tuner.py", line 294, in load_history success = base_model.fit_log(data_set, self.plan_size) File "~/Desktop/tvms/tvm/python/tvm/autotvm/tuner/xgboost_cost_model.py", line 250, in fit_log res = pool.map(feature_extract_func, data) File "~/.pyenv/versions/3.6.4/lib/python3.6/multiprocessing/pool.py", line 266, in map return self._map_async(func, iterable, mapstar, chunksize).get() File "~/.pyenv/versions/3.6.4/lib/python3.6/multiprocessing/pool.py", line 644, in get raise self._value File "~/.pyenv/versions/3.6.4/lib/python3.6/multiprocessing/pool.py", line 424, in _handle_tasks put(task) File "~/.pyenv/versions/3.6.4/lib/python3.6/multiprocessing/connection.py", line 206, in send self._send_bytes(_ForkingPickler.dumps(obj)) File "~/.pyenv/versions/3.6.4/lib/python3.6/multiprocessing/reduction.py", line 51, in dumps cls(buf, protocol).dump(obj) File "~/Desktop/tvms/tvm/python/tvm/autotvm/task/task.py", line 179, in __getstate__ self.target, self.target_host File "~/Desktop/tvms/tvm/python/tvm/target/target.py", line 197, in check_and_update_host_consist target = Target(target, host) File "~/Desktop/tvms/tvm/python/tvm/target/target.py", line 98, in __init__ raise ValueError("target has to be a string or dictionary.") ValueError: target has to be a string or dictionary. ``` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
