ingest2parquet Readme file needs to be cleaned up

[shahrokhDaijavad]

1. There are still references to IBM bluepile in the Readme (command line options)
   AST string containing input/output paths. input_folder: Path to input folder of files to
   be processed output_folder: Path to output folder of processed files Example: {
   'input_folder': '/cos-optimal-llm-pile/bluepile-
   processing/rel0_8/cc15_30_preproc_ededup', 'output_folder': '/cos-optimal-llm-
   pile/bluepile-processing/rel0_8/cc15_30_preproc_ededup/processed' }

2. In the section: Run the script via command-line, shouldn't it be:
   python ingest2parquet_local.py , instead of python ingest2parque.py ?

[shahrokhDaijavad]

This is more than a Readme file issue, because for S3 runs of this tool, we do it differently from other transforms, in which we use "make minio-start" (which starts the minio server, creates the input directory and puts a sample input file there), so I re-classify this a bug, because of the inconsistency of S3 runs of this tool with S3 runs of all other transforms.