I noticed that tensorboard logging causes the code to break when the value happens to be of type string. For example, if throughput/device_kind happens to be a "string", the following code breaks,
def log_summary(self, metrics: dict[str, Any]):
for k, v in metrics.items():
self.writer.add_scalar(k, v, global_step=None)
I noticed that tensorboard logging causes the code to break when the value happens to be of type string. For example, if throughput/device_kind happens to be a "string", the following code breaks,