areusch commented on a change in pull request #5417:
URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r415949272
##########
File path: python/tvm/autotvm/tuner/tuner.py
##########
@@ -150,7 +150,15 @@ def tune(self, n_trial, measure_option,
early_stopping=None, callbacks=(), si_pr
i + k + 1, si_prefix, format_si_prefix(flops,
si_prefix),
format_si_prefix(self.best_flops, si_prefix),
res, config)
- i += len(results)
+ num_successes = 0
+ for result in results:
+ if isinstance(result.costs[0], float):
+ num_successes += 1
+ if num_successes != len(results):
+ logger.debug('not counting %d failures towards trial count',
+ len(results) - num_successes)
+ i += num_successes
+
Review comment:
hmm okay. we do need to improve AutoTVM error handling. @tqchen what is
best to do here?
----------------------------------------------------------------
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:
us...@infra.apache.org
