Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code:   DatasetGenerationError
Exception:    CastError
Message:      Couldn't cast
id: string
doc_id: string
segment_id: string
task: string
instruction: string
input: string
output: string
crop_kind: string
crop_char_start: int64
crop_char_end: int64
preset_index: int64
span_position_ratio: double
to
{'instruction': Value('string'), 'input': Value('string'), 'output': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1816, in _prepare_split_single
                  for key, table in generator:
                                    ^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
                  for item in generator(*args, **kwargs):
                              ~~~~~~~~~^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              id: string
              doc_id: string
              segment_id: string
              task: string
              instruction: string
              input: string
              output: string
              crop_kind: string
              crop_char_start: int64
              crop_char_end: int64
              preset_index: int64
              span_position_ratio: double
              to
              {'instruction': Value('string'), 'input': Value('string'), 'output': Value('string')}
              because column names don't match
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1683, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1869, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

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.

instruction
string
input
string
output
string
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
ས་ གི་རྣས་༼ཤེང་ཀར་་ སོ་ན་བ པས་རེག་ སྨྱའི་ཤེས་ས་ཤར། ཡ་ཤོང་ན་༼སྤྱོད་སའི་སྤྲུལ་ལྔ་འར། གོ་་འརས་ལ་ འཛིན་སས་ན་མཁའི་ཡས་ མས་མིག་ག་ ཡུན་དུ་གཟུ་ རྩི་བྱེ་ལྗག་ཡོངས་༼བོད་དུ་ཤར། འོདོ་་ཀར་སོ་འཛིག །ཡོད་དུ་སྟོན་བས་ དེའི་གཟུགས་རྣམ་སར་དག་པ་༼སྤྱང་༼ ༧ཁྱས་དུ་ཤར། འདོ་ལྗང་གུ་ས་འཛིན་ ཞེས་བྱུང་གི་ཡས་མས་ ཁཱའྀ་ཀུམ་དུ་ཊིའི་བྱེ་ཟག་མོངས རོ་ཐོ་▪ རྨ་ཤ...
{"spans": [{"text": "༧ ཁྱུང་སྟེང་ནོར་གླིང་བོན་གཏེར་སྒྲོམ་བུའི་ཁར།\n", "start": 7168, "end": 7212}]}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
༈ དཔལ་ཧེ་རུ་ཀ་ལ་ཕྱག་འཚལ་ལོ། །ལན་⁽1763⁾ཐ་མལ་གྱི་རྣམ་རྟོག་ཆོས་ཉིད་ཀྱི་དབྱིངས་སུ་ཐིམ་པས་ཚ་ཅན་གྱི་རྒྱ་མཚོ་རུ། །རི་བོ་⁽1763⁾གཉིས་མེད་རྟོགས་པའི་གདེང་དང་ལྡན་པ་གཅིག་ལ་རྒྱབ་གཏད་དེ། །ནམ་མཁའ⁽1763⁾ལྷན་ཅིག་སྐྱེས་པའི་ཅི་ཡང་མ་ཡིན་པའི་དོན་གོས་སུ་གོན་བྱས་ནས། །ངར་⁽1763⁾སོ་སོར་རྟོགས་པའི་ཡེ་ཤེས་ཀྱི་བདེ་སྡུག་བླང་དོར་ཐམས་ཅད་གཉིས་མེད་དུ་ཅན་མ...
{"spans": [{"text": "སློབ་དཔོན་ཆེན་པོ་ཏིལླི་པས", "start": 1939, "end": 1963}]}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
བལྟའོ། །དེ་ནས་དེའི་སྟེང་དུ་ཨཱ་ལི་ཀཱ་ལི་ཡོངས་སུ་གྱུར་པ་རྣམས་ཨོཾ་ཨཱཿཧཱུཾ་གི་ཡི་གེ་རྣམས་ལ་གོ་རིམ་གྱིས་སྟེང་ནས་སྟེང་དུ་ཡོངས་སུ་གནས་པའོ། །འཁོར་ལོ་གསུམ་གྱི་ལྷ་རྣམས་སྤྲོས་པས་འགྲོ་བའི་དོན་བྱས་ཏེ། སྙོམས་པར་འཇུག་པ་སྔོན་དུ་སོང་བས་ཞུ་བར་གྱུར་ཏོ། །ཇི་ལྟར་དེ་རྣམས་ཀྱི་ས་བོན་རྣམས་ལ་རབ་ཏུ་ཞུགས་པའོ། །ཨོཾ་ཡིག་ལ་སོགས་པ་ཡང་རིམ་གྱིས་ཞུ་བར་བ...
{"spans": [{"text": "དཔལ་པྲཛྙཱ་རཀྵི་ཏས", "start": 9655, "end": 9671}]}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
ཐུན་མོང་ཡིད་བཞིན་ནོར་བུའི་བསྲེ་བ་སྐོར་གསུམ་བཞུགས་སོ།། ན་མོ་གུ་རུ། ཐུན་མོང་ཡིད་བཞིན་ནོར་བུའི་སྐབས་སུ་བསྲེ་བ་སྐོར་གསུམ་ནི། རྨི་ལམ་དང་སྦྱར་ཏེ་བསྲེ་བ་དང་། གྲོལ་ལམ་བསྐྱེད་རྫོགས་དང་སྦྱར་ཏེ་བསྲེ་བ་དང་། ཐོབ་བྱ་འབྲས་བུ་སྐུ་གསུམ་དང་སྦྱར་ཏེ་བསྲེ་བའོ། །དེ་ལ་བསྐྱེད་རིམ་ལ་དབང་བཞིའི་དོན་ཚང་སྟེ། ལྷའི་སྐུ་རྣམ་པ་གསལ་བ་ནི་ལུས་ཀྱི་ཉེས་པ་ད...
{"spans": [{"text": "རྗེ་བཙུན་ཆེན་པོ་ནཱ་རོ་པའ", "start": 1292, "end": 1315}]}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
གསུང་ཐུགས་ཀྱི་དམ་ཚིག་བསྲུངས་པས་འབྲས་བུ་སྐུ་བཞི་ཐོབ་པར་འགྱུར་ལ། དབང་ཐོབ་ནས་དེ་ལྟར་མ་བསྒོམ་ན་དེས་བཟློག་པར་ཤེས་པར་བྱའོ། །གཉིས་པ་ཁྱད་པར་དུ་བསྒོམ་ཞིང་ངོ་སྤྲད་པ་ནི། བླ་མའི་སྐུས་ཡི་དམ་གྱི་སྐུ་དྲངས་སློབ་མའི་ལུས་ལ་བཅུག བུམ་དབང་བསྐུར་ལུས་ཀྱི་སྒྲིབ་པ་སྦྱངས། ལམ་བསྐྱེད་རིམ་བསྒོམ། སྐུའི་དམ་ཚིག་བསྲུང་། འབྲས་བུ་གཟུགས་སྐུ་ཐོབ་པ་ལ་རང་གི...
{"spans": []}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
རྡོ་རྗེ་རྣམ་པར་འཇོམས་པའི་གཟུངས་ཀྱི་འགྲེལ་པ། སློབ་དཔོན་ཆེན་པོ་རྡོ་རྗེ་གོ་ཆ། རྒྱ་གར་སྐད་དུ། བཛྲ་བི་དཱ་ར་ཎ་ནཱ་མ་དྷཱ་ར་ཎི་བྲྀཏྟི། བོད་སྐད་དུ། རྡོ་རྗེ་རྣམ་པར་ འཇོམས་པའི་གཟུངས་ཀྱི་འགྲེལ་པ། བཅོམ་ལྡན་འདས་ཕྱག་ན་རྡོ་རྗེ་ལ་ཕྱག་འཚལ་ལོ། །དེ་ ཉིད་རྟོགས་པས་རྡོ་རྗེ་དང་། །འཁོར་བའི་ཉེས་པ་འཇོམས་མཛད་པའི། །རྫོགས་སངས་རྒྱས་ ལ་ཕྱག་འཚལ་ཏེ། །སྤ...
{"spans": [{"text": "སློབ་དཔོན་ཆེན་པོ་རྡོ་རྗེ་གོ་ཆ།\n", "start": 44, "end": 74}]}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
། སློབ་དཔོན་ཆེན་པོ་རྡོ་རྗེ་གོ་ཆ། རྒྱ་གར་སྐད་དུ། བཛྲ་བི་དཱ་ར་ཎ་ནཱ་མ་དྷཱ་ར་ཎི་བྲྀཏྟི། བོད་སྐད་དུ། རྡོ་རྗེ་རྣམ་པར་ འཇོམས་པའི་གཟུངས་ཀྱི་འགྲེལ་པ། བཅོམ་ལྡན་འདས་ཕྱག་ན་རྡོ་རྗེ་ལ་ཕྱག་འཚལ་ལོ། །དེ་ ཉིད་རྟོགས་པས་རྡོ་རྗེ་དང་། །འཁོར་བའི་ཉེས་པ་འཇོམས་མཛད་པའི། །རྫོགས་སངས་རྒྱས་ ལ་ཕྱག་འཚལ་ཏེ། །སྤྱི་ཡི་རྒྱུད་ལ་སོགས་པ་དང་། །བླ་མའི་གདམས་ངག་...
{"spans": [{"text": "སློབ་དཔོན་ཆེན་པོ་རྡོ་རྗེ་གོ་ཆ།\n", "start": 2, "end": 32}]}
Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key "spans". Each span must include "text", "start", and "end" (0-based character offsets relative to the segment text, inclusive at end). Every span text must be an exact substring of the input. If there is no author, reply...
་པར་འཇོམས་པ་ཞེས་བྱ་བའི་གཟུངས་ཀྱི་ འགྲེལ་པ་རིན་ཆེན་ཕྲེང་བ་ཞེས་བྱ་བ། སློབ་དཔོན་ཆེན་པོ་ས་ཝ་རི་པ། རྒྱ་གར་སྐར་དུ། བཛྲ་བི་དཱ་ར་ཎ་ནཱ་མ་དྷཱ་ར་ཎི་རཏྣ་མཱ་ལ་བྲྀཏྟི་ནཱ་མ། བོད་སྐད་དུ། རྡོ་རྗེ་ རྣམ་པར་འཇོམས་པ་ཞེས་བྱ་བའི་གཟུངས་ཀྱི་འགྲེལ་པ་རིན་ཆེན་ཕྲེང་བ་ཞེས་བྱ་བ། བཅོམ་ལྡན་འདས་ཕྱག་ན་རྡོ་རྗེ་ལ་ཕྱག་འཚལ་ལོ། །མ་རུངས་འཇོམས་མཛད་མི་ཤིགས་རི་ ...
{"spans": [{"text": "སློབ་དཔོན་ཆེན་པོ་ས་ཝ་རི་པ།\n", "start": 67, "end": 93}]}
"Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key(...TRUNCATED)
"མན་ངག་བཞུགས་སོ། །\nནུབ་པ་རིག་འཛིན་ག(...TRUNCATED)
"{\"spans\": [{\"text\": \"ནུབ་པ་རིག་འཛིན་གྲགས།\\n\", \"star(...TRUNCATED)
"Extract the bibliographic AUTHOR from the Tibetan segment below. Reply with JSON only using the key(...TRUNCATED)
"༄༅། །རྡོ་རྗེ་རྣམ་པར་འཇོམས་པའི་གཟུ(...TRUNCATED)
"{\"spans\": [{\"text\": \"འབྲུག་པ་པདྨ་དཀར་པོ།\\n\", \"start\"(...TRUNCATED)
End of preview.

Tibetan metadata LLM SFT dataset (10% pilot)

10% stratified random subsample of the full TiLamb SFT JSONL, for pilot LoRA training and smoke tests.

Supervised fine-tuning data for title and author span extraction from BDRC outliner segments, built for TiLamb-7B.

Full dataset

See ganga4364/tibetan-metadata-llm-sft-full for the complete train/val/test JSONL.

Layout

title/{train,val,test}.jsonl       # Alpaca format for LLaMA-Factory
title/{train,val,test}_meta.jsonl
author/{train,val,test}.jsonl
author/{train,val,test}_meta.jsonl
dataset_info.json                  # LLaMA-Factory registry snippet
reports/crop_stats.json            # if present

Each training row:

  • instruction — fixed task prompt (title or author)
  • input — cropped segment text (token-budget ≤3584 via TiLamb tokenizer)
  • output — JSON {"spans":[{"text","start","end"}]} (crop-relative offsets)

Subsampling

  • Source: full data/llm_sft built from extracted annotations
  • Method: 10% random row sample per split file (seed 42)
  • Script: scripts/subsample_llm_sft.py

Cropping

Kind When
full Whole segment fits token budget
positive Random window containing gold span (anti position-bias)
negative Random window, empty spans for that task

Source: ganga4364/tibetan-metadata-extracted.

Built with OpenPecha/tibetan-text-meta-detection llm_sft package.

Downloads last month
104