Hugging Face's logo

Datasets:
KPrashanth
/
rachana-training-dataset

Dataset card Files
xet
 Community
Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
text_path: string
model_path: string
n_lines: int64
n_tokens: int64
dtype: string
vocab_size: int64
eos_id: int64
bos_id: int64
pad_id: int64
unk_id: int64
flush_tokens: int64
created_at: string
policy: struct<min_chars: int64, max_chars: int64, sp_model_path: string, batch_lines: int64, workers: int64 (... 87 chars omitted)
  child 0, min_chars: int64
  child 1, max_chars: int64
  child 2, sp_model_path: string
  child 3, batch_lines: int64
  child 4, workers: int64
  child 5, dedup_db_path: string
  child 6, dedup_commit_every: int64
  child 7, sangraha_scores: list<item: int64>
      child 0, item: int64
out_path: string
dedup_samples: struct<indiccorpv2: list<item: null>, pralekha: list<item: null>, bpcc: list<item: null>, sangraha:  (... 19 chars omitted)
  child 0, indiccorpv2: list<item: null>
      child 0, item: null
  child 1, pralekha: list<item: null>
      child 0, item: null
  child 2, bpcc: list<item: null>
      child 0, item: null
  child 3, sangraha: list<item: string>
      child 0, item: string
sources: struct<indiccorpv2: struct<path: string, rows_out: int64, tokens_out: int64>, pralekha: struct<path: (... 228 chars omitted)
  child 0, indiccorpv2: struct<path: string, rows_out: int64, tokens_out: int64>
      child 0, path: string
      child 1, rows_out: int64
      child 2, tokens_out: int64
  child 1, pralekha: struct<path: string, rows_out: int64, tokens_out: int64>
      child 0, path: string
      child 1, rows_out: int64
      child 2, tokens_out: int64
  child 2, bpcc: struct<path: string, rows_out: int64, tokens_out: int64>
      child 0, path: string
      child 1, rows_out: int64
      child 2, tokens_out: int64
  child 3, sangraha: struct<path: string, tags_path: string, rows_out: int64, tokens_out: int64, scores_used: list<item:  (... 7 chars omitted)
      child 0, path: string
      child 1, tags_path: string
      child 2, rows_out: int64
      child 3, tokens_out: int64
      child 4, scores_used: list<item: int64>
          child 0, item: int64
stats: struct<rows_seen_indiccorpv2: int64, rows_out_total: int64, tokens_out_total: int64, chars_out_indic (... 268 chars omitted)
  child 0, rows_seen_indiccorpv2: int64
  child 1, rows_out_total: int64
  child 2, tokens_out_total: int64
  child 3, chars_out_indiccorpv2: int64
  child 4, rows_seen_pralekha: int64
  child 5, chars_out_pralekha: int64
  child 6, rows_seen_bpcc: int64
  child 7, dropped_too_short_bpcc: int64
  child 8, chars_out_bpcc: int64
  child 9, rows_seen_sangraha: int64
  child 10, chars_out_sangraha: int64
  child 11, dropped_exact_duplicate_sangraha: int64
  child 12, dedup_db_size_bytes: int64
to
{'created_at': Value('string'), 'out_path': Value('string'), 'policy': {'min_chars': Value('int64'), 'max_chars': Value('int64'), 'sp_model_path': Value('string'), 'batch_lines': Value('int64'), 'workers': Value('int64'), 'dedup_db_path': Value('string'), 'dedup_commit_every': Value('int64'), 'sangraha_scores': List(Value('int64'))}, 'sources': {'indiccorpv2': {'path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64')}, 'pralekha': {'path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64')}, 'bpcc': {'path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64')}, 'sangraha': {'path': Value('string'), 'tags_path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64'), 'scores_used': List(Value('int64'))}}, 'stats': {'rows_seen_indiccorpv2': Value('int64'), 'rows_out_total': Value('int64'), 'tokens_out_total': Value('int64'), 'chars_out_indiccorpv2': Value('int64'), 'rows_seen_pralekha': Value('int64'), 'chars_out_pralekha': Value('int64'), 'rows_seen_bpcc': Value('int64'), 'dropped_too_short_bpcc': Value('int64'), 'chars_out_bpcc': Value('int64'), 'rows_seen_sangraha': Value('int64'), 'chars_out_sangraha': Value('int64'), 'dropped_exact_duplicate_sangraha': Value('int64'), 'dedup_db_size_bytes': Value('int64')}, 'dedup_samples': {'indiccorpv2': List(Value('null')), 'pralekha': List(Value('null')), 'bpcc': List(Value('null')), 'sangraha': List(Value('string'))}}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                         ^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/utils.py", line 77, in get_rows
                  rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 299, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2321, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2249, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              text_path: string
              model_path: string
              n_lines: int64
              n_tokens: int64
              dtype: string
              vocab_size: int64
              eos_id: int64
              bos_id: int64
              pad_id: int64
              unk_id: int64
              flush_tokens: int64
              created_at: string
              policy: struct<min_chars: int64, max_chars: int64, sp_model_path: string, batch_lines: int64, workers: int64 (... 87 chars omitted)
                child 0, min_chars: int64
                child 1, max_chars: int64
                child 2, sp_model_path: string
                child 3, batch_lines: int64
                child 4, workers: int64
                child 5, dedup_db_path: string
                child 6, dedup_commit_every: int64
                child 7, sangraha_scores: list<item: int64>
                    child 0, item: int64
              out_path: string
              dedup_samples: struct<indiccorpv2: list<item: null>, pralekha: list<item: null>, bpcc: list<item: null>, sangraha:  (... 19 chars omitted)
                child 0, indiccorpv2: list<item: null>
                    child 0, item: null
                child 1, pralekha: list<item: null>
                    child 0, item: null
                child 2, bpcc: list<item: null>
                    child 0, item: null
                child 3, sangraha: list<item: string>
                    child 0, item: string
              sources: struct<indiccorpv2: struct<path: string, rows_out: int64, tokens_out: int64>, pralekha: struct<path: (... 228 chars omitted)
                child 0, indiccorpv2: struct<path: string, rows_out: int64, tokens_out: int64>
                    child 0, path: string
                    child 1, rows_out: int64
                    child 2, tokens_out: int64
                child 1, pralekha: struct<path: string, rows_out: int64, tokens_out: int64>
                    child 0, path: string
                    child 1, rows_out: int64
                    child 2, tokens_out: int64
                child 2, bpcc: struct<path: string, rows_out: int64, tokens_out: int64>
                    child 0, path: string
                    child 1, rows_out: int64
                    child 2, tokens_out: int64
                child 3, sangraha: struct<path: string, tags_path: string, rows_out: int64, tokens_out: int64, scores_used: list<item:  (... 7 chars omitted)
                    child 0, path: string
                    child 1, tags_path: string
                    child 2, rows_out: int64
                    child 3, tokens_out: int64
                    child 4, scores_used: list<item: int64>
                        child 0, item: int64
              stats: struct<rows_seen_indiccorpv2: int64, rows_out_total: int64, tokens_out_total: int64, chars_out_indic (... 268 chars omitted)
                child 0, rows_seen_indiccorpv2: int64
                child 1, rows_out_total: int64
                child 2, tokens_out_total: int64
                child 3, chars_out_indiccorpv2: int64
                child 4, rows_seen_pralekha: int64
                child 5, chars_out_pralekha: int64
                child 6, rows_seen_bpcc: int64
                child 7, dropped_too_short_bpcc: int64
                child 8, chars_out_bpcc: int64
                child 9, rows_seen_sangraha: int64
                child 10, chars_out_sangraha: int64
                child 11, dropped_exact_duplicate_sangraha: int64
                child 12, dedup_db_size_bytes: int64
              to
              {'created_at': Value('string'), 'out_path': Value('string'), 'policy': {'min_chars': Value('int64'), 'max_chars': Value('int64'), 'sp_model_path': Value('string'), 'batch_lines': Value('int64'), 'workers': Value('int64'), 'dedup_db_path': Value('string'), 'dedup_commit_every': Value('int64'), 'sangraha_scores': List(Value('int64'))}, 'sources': {'indiccorpv2': {'path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64')}, 'pralekha': {'path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64')}, 'bpcc': {'path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64')}, 'sangraha': {'path': Value('string'), 'tags_path': Value('string'), 'rows_out': Value('int64'), 'tokens_out': Value('int64'), 'scores_used': List(Value('int64'))}}, 'stats': {'rows_seen_indiccorpv2': Value('int64'), 'rows_out_total': Value('int64'), 'tokens_out_total': Value('int64'), 'chars_out_indiccorpv2': Value('int64'), 'rows_seen_pralekha': Value('int64'), 'chars_out_pralekha': Value('int64'), 'rows_seen_bpcc': Value('int64'), 'dropped_too_short_bpcc': Value('int64'), 'chars_out_bpcc': Value('int64'), 'rows_seen_sangraha': Value('int64'), 'chars_out_sangraha': Value('int64'), 'dropped_exact_duplicate_sangraha': Value('int64'), 'dedup_db_size_bytes': Value('int64')}, 'dedup_samples': {'indiccorpv2': List(Value('null')), 'pralekha': List(Value('null')), 'bpcc': List(Value('null')), 'sangraha': List(Value('string'))}}
              because column names don't match

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Rachana Training Dataset

This package contains the tokenized Telugu pretraining dataset for the Rachana GPT project.

Included Files

  • tokens.bin
  • meta.json
  • rachana_bpe32k.model
  • rachana_bpe32k.vocab
  • final_pretrain_corpus_v2_sangraha76.meta.json

Summary

  • corpus language: Telugu-first
  • tokenizer: SentencePiece BPE
  • vocab size: 32000
  • token count: 3,556,233,011
  • EOS id: 3
  • token dtype: uint32

Intended Use

This package is intended for:

  • causal language model pretraining
  • continuing existing Rachana GPT family runs
  • architecture comparison across GPT / LLaMA / Mistral / Hybrid variants

Notes

  • tokens.bin is the primary training artifact
  • meta.json is required by the training scripts
  • the tokenizer files are required for decoding and generation evaluation
  • the corpus metadata file documents the upstream text-corpus build
Downloads last month
49
Total file size:
14.2 GB