search

mlco2 / codecarbon

Track emissions from Compute and recommend ways to reduce their impact on the environment.
https://mlco2.github.io/codecarbon
MIT License
1k stars 157 forks source link

Update cpu_power.csv #534

Closed Lucas-Otavio closed 1 month ago

Lucas-Otavio commented 2 months ago

As described in issue https://github.com/mlco2/codecarbon/issues/532, I merged the TDP values with the data in the eco2ai project. For contrasting values, a search was made to find the right one or the value from code carbon was kept; except when the difference was inferior to 1 (due to int-float difference), motivating the adoption of eco2ai's float value. The list of contrasting values can be found in the issue.

Reading the diff file, it also seems my code deleted repeated and contrasting lines within the original CSV file. Let me know if there is a reason why they should be maintained or if a given criteria should have been used to select among them.

benoit-cty commented 2 months ago

Thank you !

It seems that the value is now in string instead of integer : https://github.com/mlco2/codecarbon/actions/runs/8800751288/job/24152593103?pr=534#step:5:819

Lucas-Otavio commented 1 month ago

Hey, sorry for taking some time to come back. I thought that maybe it happened because of the inclusion of float values, so I accounted for that when reading values from the data frame on core/cpu.py