Update README.md
Browse files
README.md
CHANGED
@@ -6,7 +6,8 @@ records on each checkpoint saving.
|
|
6 |
The training had 168000 iterations. Therefore multiply the reported data by 67. This would be quite approximate since we were using 16 nodes when doing
|
7 |
the ramp up, then 64 and only the last 3 weeks 128 nodes.
|
8 |
|
9 |
-
Caveat emptor: I'm not sure whether CC-reports overlap since each report is per gpu and I think they may be measuring the same thing, other than the gpu itself.
|
|
|
10 |
|
11 |
Each csv file contains a report for a single gpu.
|
12 |
|
|
|
6 |
The training had 168000 iterations. Therefore multiply the reported data by 67. This would be quite approximate since we were using 16 nodes when doing
|
7 |
the ramp up, then 64 and only the last 3 weeks 128 nodes.
|
8 |
|
9 |
+
Caveat emptor: I'm not sure whether CC-reports overlap since each report is per gpu and I think they may be measuring the same thing, other than the gpu itself.
|
10 |
+
So this requires research.
|
11 |
|
12 |
Each csv file contains a report for a single gpu.
|
13 |
|