Redundancy in the log files

[rayrayraykk]

A Fedavg on 5% of FEMNIST trail will produce a 500 kb log each round:
with 80% eval logs like 2022-04-13 16:33:24,901 (client:264) INFO: Client #1: (Evaluation (test set) at Round #26) test_loss is 79.352451. And 10% is server results and 10% is train informations.

If the round is 500, 1000, or much larger, the log files will take up too much space with a lot of redundancy. @yxdyc

[yxdyc]

The output logging files are currently redundant, leading to disk space wastage and difficulties to find key results from the logging files. We may save the raw_metrics and training_metrics in other two log files with compact format.

[rayrayraykk]

The output logging files are currently redundant, leading to disk space wastage and difficulties to find key results from the logging files. We may save the raw_metrics and training_metrics in other two log files with compact format.

And we should delete the logs below, which have already been reported on the server-side.

FederatedScope/federatedscope/core/worker/client.py

Line 260 in 24deff5

for key in eval_metrics:

[rayrayraykk]

Another issue is the eval results of the global evaluation mode are empty, we should fix this.

[yxdyc]

The output logging files are currently redundant, leading to disk space wastage and difficulties to find key results from the logging files. We may save the raw_metrics and training_metrics in other two log files with compact format.

And we should delete the logs below, which have already been reported on the server-side.

FederatedScope/federatedscope/core/worker/client.py

Line 260 in 24deff5

for key in eval_metrics:

We still need this log when we are in distributed mode. Do you agree we mask these logs only in the ``simulation'' mode? @rayrayraykk

[rayrayraykk]

The output logging files are currently redundant, leading to disk space wastage and difficulties to find key results from the logging files. We may save the raw_metrics and training_metrics in other two log files with compact format.

And we should delete the logs below, which have already been reported on the server-side.

FederatedScope/federatedscope/core/worker/client.py

Line 260 in 24deff5

for key in eval_metrics:

We still need this log when we are in distributed mode. Do you agree we mask these logs only in the ``simulation'' mode? @rayrayraykk

Agree.

[yxdyc]

Solved in #29