Add support for tensor parallelism and add OLMo2-26B model config / train script #839
Job | Run time |
---|---|
1m 24s | |
24s | |
1m 52s | |
1m 23s | |
51s | |
1m 9s | |
26s | |
24s | |
20s | |
42s | |
25s | |
0s | |
9m 20s |
Job | Run time |
---|---|
1m 24s | |
24s | |
1m 52s | |
1m 23s | |
51s | |
1m 9s | |
26s | |
24s | |
20s | |
42s | |
25s | |
0s | |
9m 20s |