Skip models column when summarizing mean

[yifanmai]

Fixes an error when trying to use aggregation_strategies: ["mean"] in schema.yaml with helm-summarize:

Traceback (most recent call last):
  File "/path/to/crfm-helm/bin/helm-summarize", line 8, in <module>
    sys.exit(main())
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/common/hierarchical_logger.py", line 104, in wrapper
    return fn(*args, **kwargs)
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/benchmark/presentation/summarize.py", line 1315, in main
    summarizer.run_pipeline(skip_completed=args.skip_completed_run_display_json)
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/benchmark/presentation/summarize.py", line 1223, in run_pipeline
    self.write_groups()
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/benchmark/presentation/summarize.py", line 1167, in write_groups
    tables: List[Table] = self.create_group_tables_by_metric_group(group)
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/benchmark/presentation/summarize.py", line 1060, in create_group_tables_by_metric_group
    table = self.create_group_table(
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/benchmark/presentation/summarize.py", line 983, in create_group_table
    means = compute_aggregate_row_means(table)
  File "/path/to/crfm-helm/lib/python3.10/site-packages/helm/benchmark/presentation/summarize.py", line 279, in compute_aggregate_row_means
    total += float(cell.value)
ValueError: could not convert string to float: 'DBRX Instruct'

If the first column contains the names of models, helm-summarize should skip it instead of treating it like a value column when computing the mean.