Matches in Nanopublications for { ?s ?p "Similar to metadata, benchmark submissions are given a number unique across MLCommons benchmarks. However, they are not resolvable by machines. The convention used is the benchmark round hyphenated with a unique submission number, e.g., 3.1-2000 for benchmark 3.1, first submission. Different benchmarks begin with different number spans, such as training begins at 2000 and HPC begins at 8000. It would be optimal if each benchmark and round had a PID." ?g. }
Showing items 1 to 2 of
2
with 100 items per page.
- declaration considerations "Similar to metadata, benchmark submissions are given a number unique across MLCommons benchmarks. However, they are not resolvable by machines. The convention used is the benchmark round hyphenated with a unique submission number, e.g., 3.1-2000 for benchmark 3.1, first submission. Different benchmarks begin with different number spans, such as training begins at 2000 and HPC begins at 8000. It would be optimal if each benchmark and round had a PID." assertion.
- declaration considerations "Similar to metadata, benchmark submissions are given a number unique across MLCommons benchmarks. However, they are not resolvable by machines. The convention used is the benchmark round hyphenated with a unique submission number, e.g., 3.1-2000 for benchmark 3.1, first submission. Different benchmarks begin with different number spans, such as training begins at 2000 and HPC begins at 8000. It would be optimal if each benchmark and round had a PID." assertion.