Hugging Face's logo

Datasets:
Vi0509
/
longmemeval-com-dom-per-question

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:    TypeError
Message:      Couldn't cast array of type string to null
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                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 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 478, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2818, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                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 2303, in cast_table_to_schema
                  cast_array_to_feature(
                  ~~~~~~~~~~~~~~~~~~~~~^
                      table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                      feature,
                      ^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1852, in wrapper
                  return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
                                           ~~~~^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2143, in cast_array_to_feature
                  return array_cast(
                      array,
                  ...<2 lines>...
                      allow_decimal_to_str=allow_decimal_to_str,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1854, in wrapper
                  return func(array, *args, **kwargs)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2005, in array_cast
                  raise TypeError(f"Couldn't cast array of type {_short_str(array.type)} to {_short_str(pa_type)}")
              TypeError: Couldn't cast array of type string to null

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.

longmemeval-com-dom-per-question

Private archive of per_question.jsonl rows from the LME-v2 (LongMemEval COM/DOM) eval runs on Viraj's Windows box (C:/Users/12066/Desktop/longmemeval-com-dom-memory/runs/). These are the last single-copy artifacts of that eval arc -- gitignored under runs/ per the repo's heavy-data-goes-to-HF rule, and this box is OOM-prone, so they live here as the backup copy.

Status

BLOCKED as of 2026-07-04: the Vi0509 HF account's private storage quota (free tier) is already exceeded -- 158.85 GB private used, dominated by savant-training-data (107.4 GB) and savant-v8-h100-checkpoints (30.3 GB) -- so every per_question.jsonl upload attempt 403s at the LFS-batch step ("Private repository storage limit reached, please upgrade your plan"). Zero data files have been uploaded; only this README and the default .gitattributes exist in the repo. Resolving this needs either (a) a paid-plan upgrade (the account currently has canPay: false -- billing isn't set up), or (b) freeing private-storage headroom by pruning/moving some of the older private repos, or (c) an explicit decision from Viraj to make this specific dataset public instead of private. None of those are this agent's call to make unilaterally.

block_id mapping coverage

Every run directory below was searched for a block_id (grep across all files for block_id, blockId, "block", unblock_remember, "memory_id", remember(, banked, brain_block). Zero matches -- no run in this corpus recorded which org-brain block (if any) it was banked as. All 55 rows are therefore block_id: unknown below, per instruction not to invent one.

Mapping: run dir -> description -> block_id

dir description domain questions overall_full_set size (MB) block_id
_argparse_smoke/per_question.jsonl Argparse/harness smoke test (1 question) -- pipeline plumbing check, not a scored eval web 1 0.0000 0.0 unknown
chunked_m2_ent_r1/chunk_0/per_question.jsonl M2-composite-ent lever probe on enterprise paired30, executed in parallel chunks (chunk_0/2/3 survived; chunk_1 produced no per_question.jsonl) enterprise 53 0.4717 12.2 unknown
chunked_m2_ent_r1/chunk_2/per_question.jsonl M2-composite-ent lever probe on enterprise paired30, executed in parallel chunks (chunk_0/2/3 survived; chunk_1 produced no per_question.jsonl) enterprise 53 0.3962 12.4 unknown
chunked_m2_ent_r1/chunk_3/per_question.jsonl M2-composite-ent lever probe on enterprise paired30, executed in parallel chunks (chunk_0/2/3 survived; chunk_1 produced no per_question.jsonl) enterprise 52 0.3269 12.1 unknown
enterprise_arl_a1/per_question.jsonl Enterprise ARL (autonomous-response-loop) arm-1 baseline run 1 enterprise 30 0.2667 6.8 unknown
enterprise_arl_a1_r2/per_question.jsonl Enterprise ARL arm-1 baseline, replicate 2 enterprise 30 0.3000 6.8 unknown
enterprise_arl_a1_r3/per_question.jsonl Enterprise ARL arm-1 baseline, replicate 3 enterprise 30 0.2333 6.7 unknown
enterprise_arl_a2/per_question.jsonl Enterprise ARL arm-2 baseline run 1 enterprise 30 0.3000 7.8 unknown
enterprise_arl_a2_r2/per_question.jsonl Enterprise ARL arm-2 baseline, replicate 2 enterprise 30 0.3000 7.9 unknown
enterprise_arl_a2_r3/per_question.jsonl Enterprise ARL arm-2 baseline, replicate 3 enterprise 30 0.3667 7.9 unknown
enterprise_organc_fab56/per_question.jsonl Enterprise org-context run, fab56-tagged question set (56 questions) enterprise 56 0.1964 13.5 unknown
episodic_smoke30_armA1/per_question.jsonl Episodic-architecture smoke test, arm A1, web domain (240 q -- full web set, not just 30) web 240 0.5000 38.3 unknown
episodic_smoke30_armA1_clean_reref/per_question.jsonl Episodic smoke arm A1, cleaned/re-dereferenced re-run (30 q) web 30 0.5333 4.7 unknown
episodic_smoke30_armA2/per_question.jsonl Episodic-architecture smoke test, arm A2, web domain (30 q) web 30 0.5333 5.0 unknown
full_com_dom_web_small_scoped/per_question.jsonl Full COM+DOM pipeline, web-small domain, scoped-memory variant (prompt-only artifacts, no scored metrics on disk) web n/a n/a 0.5 unknown
full_com_dom_web_small_scoped_retry2_conc4/per_question.jsonl Full COM+DOM pipeline, web-small domain, scoped-memory, retry-2 at concurrency 4 (240 q) web 240 0.3667 72.9 unknown
fulltier_ent_semantic_r1_20260703T172025Z/per_question.jsonl Full-tier enterprise eval, semantic-rank channel (Qwen3-Embed-8B, blend 0.5) -- the ent 0.356->0.44 lift run, replicate 1 (211 q) enterprise 211 0.4408 44.3 unknown
fulltier_ent_semantic_r2_20260703T184920Z/per_question.jsonl Full-tier enterprise eval, semantic-rank channel -- replicate 2 (211 q) enterprise 211 0.4408 44.3 unknown
fulltier_ent_semantic_r3b_20260703T210258Z/per_question.jsonl Full-tier enterprise eval, semantic-rank channel -- replicate 3b (211 q) enterprise 211 0.4171 44.2 unknown
iter_L1a-arc-serve-web_l1a_web_arc30_20260703T081750Z/per_question.jsonl L1a arc-serve-web lever probe, web arc30 set web n/a n/a 4.9 unknown
iter_L1a-arc-serve-web_l1a_web_arc30_20260703T085140Z/per_question.jsonl L1a arc-serve-web lever probe, web arc30 set web 30 0.1667 5.2 unknown
iter_L1b-fullpage-web_web_paired30_20260703T081658Z/per_question.jsonl L1b fullpage-web lever probe, web paired30 set web 30 0.4667 4.9 unknown
iter_L1b-fullpage-web_web_paired30_20260703T085140Z/per_question.jsonl L1b fullpage-web lever probe, web paired30 set web 30 0.5000 4.9 unknown
iter_L1d-multiform-ent_l1d_ent_multiform30_20260703T085140Z/per_question.jsonl L1d multiform-ent lever probe, ent multiform30 set (killed lever -- L1d) enterprise 30 0.1333 7.4 unknown
iter_L1d-multiform-ent_l1d_ent_multiform30_20260703T093752Z/per_question.jsonl L1d multiform-ent lever probe, ent multiform30 set (killed lever -- L1d) enterprise 30 0.0667 7.2 unknown
iter_L2-specific-rejection-ent_ent_fab56_20260703T081707Z/per_question.jsonl L2 specific-rejection-ent lever probe, fab56 question set (56 q, 0.732 -- highest score in corpus) enterprise 56 0.7321 13.5 unknown
iter_L3-vision-web_l3_vision_set_20260703T081914Z/per_question.jsonl L3 vision-web lever probe, dedicated vision question set web 13 0.1538 2.1 unknown
iter_M2-composite-ent_ent_paired30_20260703T235906Z/per_question.jsonl M2-composite-ent lever probe, ent paired30 stratified set enterprise 30 0.4000 6.9 unknown
iter_M2-composite-ent_ent_paired30_smoke5_20260703T211542Z/per_question.jsonl M2-composite-ent lever probe, smoke5 subset enterprise 5 0.2000 1.1 unknown
iter_ent-rerank-iso_ent_paired30_20260704T082644ZC/per_question.jsonl Lever-iteration probe: ent-rerank-iso (Qwen3-Reranker-4B isolated) on ent paired30 stratified set enterprise 30 0.4000 6.0 unknown
iter_ent-semantic-reader_ent_paired30_20260703T110025Z/per_question.jsonl Lever-iteration probe: ent-semantic-reader on ent paired30 stratified set enterprise 30 0.3667 6.2 unknown
iter_ent-semantic-reader_ent_paired30_20260703T113054Z/per_question.jsonl Lever-iteration probe: ent-semantic-reader on ent paired30 stratified set enterprise 30 0.4333 6.2 unknown
iter_ent-semantic-reader_ent_paired30_smoke5_20260703T105512Z/per_question.jsonl Lever-iteration probe: ent-semantic-reader, smoke5 subset of ent paired30 enterprise 5 0.4000 1.0 unknown
iter_ent-topk20_ent_paired30_20260703T093307Z/per_question.jsonl Lever-iteration probe: ent-topk20 (L1 recall lever) on ent paired30 -- topk20 net-0 finding enterprise 30 0.1667 18.7 unknown
iter_ent-topk20_ent_paired30_smoke5_20260703T093000Z/per_question.jsonl Lever-iteration probe: ent-topk20 (L1 recall lever), smoke5 subset -- topk20 net-0 finding enterprise 5 0.4000 3.1 unknown
iter_has-ent_ent_paired30_20260704T053923Z/per_question.jsonl Lever-iteration probe: has-ent gate on ent paired30 stratified set enterprise 30 0.4000 6.3 unknown
iter_has-ent_ent_paired30_20260704T092720ZC/per_question.jsonl Lever-iteration probe: has-ent gate on ent paired30 stratified set enterprise 30 0.4667 6.2 unknown
iter_l1a_arc_l1a_web_arc30_smoke5_20260703T080851Z/per_question.jsonl L1a (arc-serve) lever probe, web arc30 smoke5 subset web 5 0.2000 0.8 unknown
iter_l1b_fullpage_web_paired30_smoke5_20260703T080855Z/per_question.jsonl L1b (fullpage-serve) lever probe, web paired30 smoke5 subset web 5 0.4000 0.9 unknown
iter_l1d_multiform_l1d_ent_multiform30_smoke5_20260703T080856Z/per_question.jsonl L1d (multiform-ent) lever probe, smoke5 subset enterprise 5 0.2000 1.2 unknown
iter_l2_specrej_ent_paired30_smoke5_20260703T080918Z/per_question.jsonl L2 (specific-rejection-ent) lever probe, smoke5 subset enterprise 5 0.2000 1.2 unknown
iter_l3_vision_web_paired30_smoke5_20260703T080905Z/per_question.jsonl L3 (vision) lever probe, web paired30 smoke5 subset web 5 0.6000 0.8 unknown
iter_uncapped-content_web_paired30_20260703T201833Z/per_question.jsonl Uncapped-content lever probe (uncapped token budget), web paired30 set web 30 0.4000 22.8 unknown
rebaseline_enterprise_small_t0_r1_20260701_200638/per_question.jsonl Enterprise rebaseline triad, temperature 0, replicate 1 (211 q) enterprise 211 0.3460 48.3 unknown
rebaseline_enterprise_small_t0_r2_20260701_200638/per_question.jsonl Enterprise rebaseline triad, temperature 0, replicate 2 (211 q) enterprise 211 0.3507 48.1 unknown
rebaseline_enterprise_small_t0_r3_20260701_200638/per_question.jsonl Enterprise rebaseline triad, temperature 0, replicate 3 (211 q) enterprise 211 0.3649 48.4 unknown
rebaseline_web_small_t0_r1_20260701_200638/per_question.jsonl Web rebaseline triad, temperature 0, replicate 1 (240 q) web 240 0.4750 39.4 unknown
rebaseline_web_small_t0_r2_20260701_200638/per_question.jsonl Web rebaseline triad, temperature 0, replicate 2 (240 q) web 240 0.4750 38.9 unknown
rebaseline_web_small_t0_r3_20260701_200638/per_question.jsonl Web rebaseline triad, temperature 0, replicate 3 (240 q) web 240 0.5000 38.9 unknown
smoke_com_dom_web_small_limit2_cached/per_question.jsonl COM+DOM smoke test, web-small, limit-2 questions, cached-memory variant web 2 0.5000 0.5 unknown
smoke_com_dom_web_small_limit2_scoped/per_question.jsonl COM+DOM smoke test, web-small, limit-2 questions, scoped-memory variant web 2 1.0000 0.5 unknown
smoke_com_dom_web_small_patched/per_question.jsonl COM+DOM smoke test, web-small, patched pipeline (1 q) web 1 1.0000 0.3 unknown
smoke_episodic_arm_a_20260702_013013/per_question.jsonl Episodic architecture smoke test, arm A (3 q) web 3 1.0000 0.4 unknown
uncapped_strat35/per_question.jsonl Uncapped-token-budget run, stratified 35-question set web 35 0.5714 13.8 unknown
uncapped_web_small_full/per_question.jsonl Uncapped-token-budget run, full web-small set (240 q) web 240 0.5167 93.6 unknown

Named evidence sets called out in the archive request

  • ent 0.4408 triad: NOT enterprise_arl_a1/a1_r2/a1_r3 (those score 0.267/0.30/0.233, n=30). The actual 0.4408-scoring triad is fulltier_ent_semantic_r1_20260703T172025Z (0.4408), fulltier_ent_semantic_r2_20260703T184920Z (0.4408), and fulltier_ent_semantic_r3b_20260703T210258Z (0.4171, close but not identical -- flagging rather than fudging it), all n=211.
  • web rebaseline triad: rebaseline_web_small_t0_r{1,2,3}_20260701_200638 -- present, n=240 each.
  • chunked_m2_ent_r1: present as chunk_0, chunk_2, chunk_3 (matches expectation; chunk_1 never produced a per_question.jsonl, only logs/qids -- likely failed or was still running).
  • *paired*30*: all completed iter_*_paired30_* runs are present. Three additional chunkedp_*_ent_paired30 directories (chunkedp_ent-rerank-iso, chunkedp_has-ent, chunkedp_rr-iso) exist on disk but never produced a per_question.jsonl in any chunk (only prompt_rows.jsonl / run_args.json) -- nothing to archive from those three, not a bug in this archival job.

Reproducing / verifying 

hf auth login --token <HF_TOKEN>
hf upload Vi0509/longmemeval-com-dom-per-question <local runs/ dir> --repo-type=dataset \
  --include='**/per_question.jsonl'
Downloads last month
84
Total file size:
901 MB