fix formatting

configs/neox_arguments.md

@@ -926,7 +926,7 @@ Text Generation arguments
 
 - **prompt_end**: str
 
- Default = 
+ Default =
 
 
  a single prompt's end. Defaults to newline
@@ -968,7 +968,7 @@ Text Generation arguments
 
 - **eval_results_prefix**: str
 
- Default = 
+ Default =
 
  prefix to which to save evaluation results - final fp will be {eval_results_prefix}_eval_results_yy-mm-dd-HH-MM.json
 
@@ -1686,7 +1686,7 @@ Args for deepspeed config
 
  Default = None
 
- 
+
 
 
 
@@ -1980,4 +1980,3 @@ Args for deepspeed runner (deepspeed.launcher.runner).
  Default = None
 
  Adds a `--comment` to the DeepSpeed launch command. In DeeperSpeed this is passed on to the SlurmLauncher as well. Sometime necessary for cluster rules, or so I've heard.
-

megatron/data/helpers.cpp

@@ -173,7 +173,6 @@ py::array build_sample_idx_int32(const py::array_t<int32_t>& sizes_,
  free_when_done); // numpy array references
 }
 
-
 py::array build_sample_idx_int64(const py::array_t<int32_t>& sizes_,
  const py::array_t<int32_t>& doc_idx_,
  const int32_t seq_length,

megatron/fused_kernels/__init__.py

@@ -38,7 +38,7 @@ def load_fused_kernels():
  print(e)
  print("=" * 100)
  print(
- f'ERROR: Fused kernels configured but not properly installed. Please run `pip install {str(srcpath)}` to install them'
+ f"ERROR: Fused kernels configured but not properly installed. Please run `pip install {str(srcpath)}` to install them"
  )
  print("=" * 100)
  exit()

megatron/neox_arguments/arguments.py

@@ -388,7 +388,7 @@ def consume_deepy_args(cls):
  raise e
 
  neox_args.wandb_group += "_" + wandb.util.generate_id()
- 
+
  neox_args.print()
 
  return neox_args

tools/merge_datasets.py

@@ -2,8 +2,10 @@
 import sys
 import json
 import argparse
-sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__),
- os.path.pardir)))
+
+sys.path.append(
+ os.path.abspath(os.path.join(os.path.dirname(__file__), os.path.pardir))
+)
 
 from megatron.data import indexed_dataset
 
@@ -20,47 +22,63 @@ def main(args):
  if not os.path.isfile(os.path.join(args.input, basename)):
  continue
 
- ext_pair = '.bin' if ext == '.idx' else '.idx'
- assert os.path.isfile(os.path.join(args.input, prefix) + ext_pair), \
- f'ERROR: {ext_pair} file not provided for {os.path.join(args.input, prefix)}'
+ ext_pair = ".bin" if ext == ".idx" else ".idx"
+ assert os.path.isfile(
+ os.path.join(args.input, prefix) + ext_pair
+ ), f"ERROR: {ext_pair} file not provided for {os.path.join(args.input, prefix)}"
 
  prefixes.add(prefix)
 
  builder = None
  for prefix in sorted(prefixes):
  if builder is None:
- dataset = indexed_dataset.make_dataset(os.path.join(args.input, prefix), 'infer')
+ dataset = indexed_dataset.make_dataset(
+ os.path.join(args.input, prefix), "infer"
+ )
 
  if isinstance(dataset, indexed_dataset.MMapIndexedDataset):
- builder = indexed_dataset.MMapIndexedDatasetBuilder(args.output_prefix + '.bin', dtype=dataset._index.dtype)
+ builder = indexed_dataset.MMapIndexedDatasetBuilder(
+ args.output_prefix + ".bin", dtype=dataset._index.dtype
+ )
  else:
- builder = indexed_dataset.IndexedDatasetBuilder(args.output_prefix + '.bin')
+ builder = indexed_dataset.IndexedDatasetBuilder(
+ args.output_prefix + ".bin"
+ )
 
  del dataset
 
  builder.merge_file_(os.path.join(args.input, prefix))
 
- builder.finalize(args.output_prefix + '.idx')
+ builder.finalize(args.output_prefix + ".idx")
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
  parser = argparse.ArgumentParser()
 
- group = parser.add_argument_group(title='input data')
- group.add_argument('--input', type=str, required=True,
- help='Path to directory containing all document files to merge')
-
- group = parser.add_argument_group(title='output data')
- group.add_argument('--output-prefix', type=str, required=True,
- help='Path to binary output file without suffix')
+ group = parser.add_argument_group(title="input data")
+ group.add_argument(
+ "--input",
+ type=str,
+ required=True,
+ help="Path to directory containing all document files to merge",
+ )
+
+ group = parser.add_argument_group(title="output data")
+ group.add_argument(
+ "--output-prefix",
+ type=str,
+ required=True,
+ help="Path to binary output file without suffix",
+ )
 
  args = parser.parse_args()
 
- assert os.path.isdir(args.input), \
- f'ERROR: {args.input} is not a directory or does not exist'
+ assert os.path.isdir(
+ args.input
+ ), f"ERROR: {args.input} is not a directory or does not exist"
 
- assert os.path.isdir(os.path.dirname(args.output_prefix)), \
- f'ERROR: {os.path.dirname(args.output_prefix)} is not a directory or does not exist'
+ assert os.path.isdir(
+ os.path.dirname(args.output_prefix)
+ ), f"ERROR: {os.path.dirname(args.output_prefix)} is not a directory or does not exist"
 
  main(args)
-