Datasets:
Dataset Preview
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
slice: string
cik: int64
accession: string
accession_url: string
exhibit_sha256: string
gold_guidance: null
gold_non_gaap_eps_usd: double
stratum: string
full_exhibit_verified: bool
claim_supported: string
machine_scan: struct<ok: bool, sha256: string, guidance_hint: bool, n_guidance_sentences: int64, guidance_sentence (... 96 chars omitted)
child 0, ok: bool
child 1, sha256: string
child 2, guidance_hint: bool
child 3, n_guidance_sentences: int64
child 4, guidance_sentences: list<item: string>
child 0, item: string
child 5, non_gaap_marker: bool
child 6, non_gaap_value: double
child 7, paragraphs_scanned: int64
hand_audit_status: string
hand_audit_by: null
category: string
row_id: int64
period_end: timestamp[s]
traits: list<item: string>
child 0, item: string
source: string
rendered: bool
n_table_cols: int64
prior_table_col: int64
shard: string
period_months: int64
filing_date: timestamp[s]
table_col: int64
license: string
reference: string
entity_name: string
completion: string
split: string
scale: int64
prompt: string
n_tokens: int64
column_order: list<item: int64>
child 0, item: int64
to
{'id': Value('string'), 'row_id': Value('int64'), 'split': Value('string'), 'category': Value('string'), 'shard': Value('string'), 'prompt': Value('string'), 'completion': Value('string'), 'reference': Value('string'), 'cik': Value('int64'), 'entity_name': Value('string'), 'accession': Value('string'), 'accession_url': Value('string'), 'exhibit_sha256': Value('string'), 'filing_date': Value('timestamp[s]'), 'period_end': Value('timestamp[s]'), 'period_months': Value('int64'), 'scale': Value('int64'), 'table_col': Value('int64'), 'prior_table_col': Value('int64'), 'n_table_cols': Value('int64'), 'column_order': List(Value('int64')), 'rendered': Value('bool'), 'traits': List(Value('string')), 'full_exhibit_verified': Value('bool'), 'license': Value('string'), 'source': Value('string'), 'n_tokens': Value('int64')}
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
slice: string
cik: int64
accession: string
accession_url: string
exhibit_sha256: string
gold_guidance: null
gold_non_gaap_eps_usd: double
stratum: string
full_exhibit_verified: bool
claim_supported: string
machine_scan: struct<ok: bool, sha256: string, guidance_hint: bool, n_guidance_sentences: int64, guidance_sentence (... 96 chars omitted)
child 0, ok: bool
child 1, sha256: string
child 2, guidance_hint: bool
child 3, n_guidance_sentences: int64
child 4, guidance_sentences: list<item: string>
child 0, item: string
child 5, non_gaap_marker: bool
child 6, non_gaap_value: double
child 7, paragraphs_scanned: int64
hand_audit_status: string
hand_audit_by: null
category: string
row_id: int64
period_end: timestamp[s]
traits: list<item: string>
child 0, item: string
source: string
rendered: bool
n_table_cols: int64
prior_table_col: int64
shard: string
period_months: int64
filing_date: timestamp[s]
table_col: int64
license: string
reference: string
entity_name: string
completion: string
split: string
scale: int64
prompt: string
n_tokens: int64
column_order: list<item: int64>
child 0, item: int64
to
{'id': Value('string'), 'row_id': Value('int64'), 'split': Value('string'), 'category': Value('string'), 'shard': Value('string'), 'prompt': Value('string'), 'completion': Value('string'), 'reference': Value('string'), 'cik': Value('int64'), 'entity_name': Value('string'), 'accession': Value('string'), 'accession_url': Value('string'), 'exhibit_sha256': Value('string'), 'filing_date': Value('timestamp[s]'), 'period_end': Value('timestamp[s]'), 'period_months': Value('int64'), 'scale': Value('int64'), 'table_col': Value('int64'), 'prior_table_col': Value('int64'), 'n_table_cols': Value('int64'), 'column_order': List(Value('int64')), 'rendered': Value('bool'), 'traits': List(Value('string')), 'full_exhibit_verified': Value('bool'), 'license': Value('string'), 'source': Value('string'), 'n_tokens': Value('int64')}
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 datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
id string | row_id int64 | split string | category string | shard string | prompt string | completion string | reference string | cik int64 | entity_name string | accession string | accession_url string | exhibit_sha256 string | filing_date timestamp[s] | period_end timestamp[s] | period_months int64 | scale int64 | table_col int64 | prior_table_col int64 | n_table_cols int64 | column_order list | rendered bool | traits list | full_exhibit_verified bool | license string | source string | n_tokens int64 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ff-30000000 | 30,000,000 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 473,438 becomes 473438000; yoy = (473438000-415684000)/415684000 = 13.9%
{"company": "Natus Medical Incorporated", "period_end": "2021-12-31", "period_months": 12, "revenue_usd": 473438000, "net_income_usd": 13177000, "prior_year_revenue_usd": 415684000, "eps_basic_usd": 0.39, "eps_diluted_... | {"company": "Natus Medical Incorporated", "period_end": "2021-12-31", "period_months": 12, "revenue_usd": 473438000, "net_income_usd": 13177000, "prior_year_revenue_usd": 415684000, "eps_basic_usd": 0.39, "eps_diluted_usd": 0.39, "revenue_yoy_pct": 13.9, "guidance": null, "non_gaap_eps_usd": null} | 878,526 | NATUS MEDICAL INC | 0000878526-22-000024 | https://www.sec.gov/Archives/edgar/data/878526/000087852622000024/0000878526-22-000024-index.htm | fc2ebf1163d1f4f39d962108b45f8837cfeda1de3c16f727e1c00a667df4b9ae | 2022-02-24T00:00:00 | 2021-12-31T00:00:00 | 12 | 1,000 | 3 | 2 | 4 | [
0,
1,
3,
2
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/878526/000087852622000024/0000878526-22-000024-index.htm | 1,787 |
ff-30000001 | 30,000,001 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 1,003 becomes 1003000; yoy = (1003000-1216000)/1216000 = -17.5%
{"company": "Neonode Inc", "period_end": "2023-09-30", "period_months": 3, "revenue_usd": 1003000, "net_income_usd": -1266000, "prior_year_revenue_usd": 1216000, "eps_basic_usd": -0.08, "eps_diluted_usd": -0.08, "revenue_yoy_pc... | {"company": "Neonode Inc", "period_end": "2023-09-30", "period_months": 3, "revenue_usd": 1003000, "net_income_usd": -1266000, "prior_year_revenue_usd": 1216000, "eps_basic_usd": -0.08, "eps_diluted_usd": -0.08, "revenue_yoy_pct": -17.5, "guidance": null, "non_gaap_eps_usd": null} | 87,050 | Neonode Inc. | 0001213900-23-084971 | https://www.sec.gov/Archives/edgar/data/87050/000121390023084971/0001213900-23-084971-index.htm | ed12e282a43059fd75febef119f7a26ebffbc9ded0e5a09504f842c67d685b8f | 2023-11-09T00:00:00 | 2023-09-30T00:00:00 | 3 | 1,000 | 3 | 0 | 4 | [
1,
2,
3,
0
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/87050/000121390023084971/0001213900-23-084971-index.htm | 1,710 |
ff-30000002 | 30,000,002 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 6,212,878 becomes 6212878000; yoy = (6212878000-4465308000)/4465308000 = 39.1%
{"company": "Steel Dynamics, Inc", "period_end": "2022-06-30", "period_months": 3, "revenue_usd": 6212878000, "net_income_usd": 1214652000, "prior_year_revenue_usd": 4465308000, "eps_basic_usd": 6.49, "eps_dilute... | {"company": "Steel Dynamics, Inc", "period_end": "2022-06-30", "period_months": 3, "revenue_usd": 6212878000, "net_income_usd": 1214652000, "prior_year_revenue_usd": 4465308000, "eps_basic_usd": 6.49, "eps_diluted_usd": 6.44, "revenue_yoy_pct": 39.1, "guidance": null, "non_gaap_eps_usd": null} | 1,022,671 | STEEL DYNAMICS INC | 0001104659-22-081503 | https://www.sec.gov/Archives/edgar/data/1022671/000110465922081503/0001104659-22-081503-index.htm | e47acb7d68118fda2765f620e7cf7ed8261ac93ab0baa4f77812fe86bd0f219e | 2022-07-21T00:00:00 | 2022-06-30T00:00:00 | 3 | 1,000 | 2 | 0 | 5 | [
1,
2,
0,
3,
4
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1022671/000110465922081503/0001104659-22-081503-index.htm | 1,947 |
ff-30000003 | 30,000,003 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 5,689,233 becomes 5689233000; yoy = (5689233000-4829897000)/4829897000 = 17.8%
{"company": "Academy Sports", "period_end": "2021-01-30", "period_months": 12, "revenue_usd": 5689233000, "net_income_usd": 308764000, "prior_year_revenue_usd": 4829897000, "eps_basic_usd": 3.96, "eps_diluted_usd... | {"company": "Academy Sports", "period_end": "2021-01-30", "period_months": 12, "revenue_usd": 5689233000, "net_income_usd": 308764000, "prior_year_revenue_usd": 4829897000, "eps_basic_usd": 3.96, "eps_diluted_usd": 3.79, "revenue_yoy_pct": 17.8, "guidance": null, "non_gaap_eps_usd": 3.83} | 1,817,358 | Academy Sports & Outdoors, Inc. | 0001817358-21-000030 | https://www.sec.gov/Archives/edgar/data/1817358/000181735821000030/0001817358-21-000030-index.htm | 22e44e9d6cfd2e6723cfdaf74d594cea3b5817e5d67398674e70dffba2697106 | 2021-03-30T00:00:00 | 2021-01-30T00:00:00 | 12 | 1,000 | 1 | 2 | 4 | [
1,
0,
2,
3
] | false | [
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1817358/000181735821000030/0001817358-21-000030-index.htm | 1,721 |
ff-30000004 | 30,000,004 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 7,085 becomes 7085000; yoy = (7085000-12422000)/12422000 = -43.0%
{"company": "Byrna Technologies Inc", "period_end": "2023-08-31", "period_months": 3, "revenue_usd": 7085000, "net_income_usd": -4094000, "prior_year_revenue_usd": 12422000, "eps_basic_usd": -0.19, "eps_diluted_usd": -0.19, "... | {"company": "Byrna Technologies Inc", "period_end": "2023-08-31", "period_months": 3, "revenue_usd": 7085000, "net_income_usd": -4094000, "prior_year_revenue_usd": 12422000, "eps_basic_usd": -0.19, "eps_diluted_usd": -0.19, "revenue_yoy_pct": -43.0, "guidance": null, "non_gaap_eps_usd": null} | 1,354,866 | Byrna Technologies Inc. | 0001437749-23-028089 | https://www.sec.gov/Archives/edgar/data/1354866/000143774923028089/0001437749-23-028089-index.htm | 41bd68b77f8f02741038d47778fc5e5c35dc795fd15102e3c956b044926a9bcc | 2023-10-12T00:00:00 | 2023-08-31T00:00:00 | 3 | 1,000 | 0 | 1 | 4 | [
0,
1,
2,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1354866/000143774923028089/0001437749-23-028089-index.htm | 1,667 |
ff-30000005 | 30,000,005 | train | period_hard | period_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table already in dollars so 8890984 stands; yoy = (8890984-9247710)/9247710 = -3.9%
{"company": "Hall of Fame Resort & Entertainment Company", "period_end": "2024-06-30", "period_months": 6, "revenue_usd": 8890984, "net_income_usd": -30119158, "prior_year_revenue_usd": 9247710, "eps_basic_usd": -4.71, "eps_dilut... | {"company": "Hall of Fame Resort & Entertainment Company", "period_end": "2024-06-30", "period_months": 6, "revenue_usd": 8890984, "net_income_usd": -30119158, "prior_year_revenue_usd": 9247710, "eps_basic_usd": -4.71, "eps_diluted_usd": -4.71, "revenue_yoy_pct": -3.9, "guidance": null, "non_gaap_eps_usd": null} | 1,708,176 | Hall of Fame Resort & Entertainment Co | 0001213900-24-069619 | https://www.sec.gov/Archives/edgar/data/1708176/000121390024069619/0001213900-24-069619-index.htm | eb3c08b2fae872bf4e1ddc5a6858a62fe50ecaaa9246d97bd3da16cb61dbebcc | 2024-08-15T00:00:00 | 2024-06-30T00:00:00 | 6 | 1 | 2 | 3 | 4 | [
0,
1,
2,
3
] | false | [
"null_discipline",
"period_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1708176/000121390024069619/0001213900-24-069619-index.htm | 1,980 |
ff-30000006 | 30,000,006 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 1,602,202 becomes 1602202000; yoy = (1602202000-1474510000)/1474510000 = 8.7%
{"company": "Civitas Solutions, Inc", "period_end": "2018-09-30", "period_months": 12, "revenue_usd": 1602202000, "net_income_usd": 14886000, "prior_year_revenue_usd": 1474510000, "eps_basic_usd": 0.4, "eps_dilute... | {"company": "Civitas Solutions, Inc", "period_end": "2018-09-30", "period_months": 12, "revenue_usd": 1602202000, "net_income_usd": 14886000, "prior_year_revenue_usd": 1474510000, "eps_basic_usd": 0.4, "eps_diluted_usd": 0.4, "revenue_yoy_pct": 8.7, "guidance": null, "non_gaap_eps_usd": null} | 1,608,638 | Civitas Solutions, Inc. | 0001608638-18-000026 | https://www.sec.gov/Archives/edgar/data/1608638/000160863818000026/0001608638-18-000026-index.htm | 2d6c5583df95c2245d142347e3dc11e35b30a3c0d580878c9235e0d4d817a4c5 | 2018-12-13T00:00:00 | 2018-09-30T00:00:00 | 12 | 1,000 | 0 | 3 | 4 | [
2,
0,
1,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1608638/000160863818000026/0001608638-18-000026-index.htm | 1,717 |
ff-30000007 | 30,000,007 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 169,432 becomes 169432000; yoy = (169432000-165755000)/165755000 = 2.2%
{"company": "American Assets Trust, Inc", "period_end": "2019-06-30", "period_months": 6, "revenue_usd": 169432000, "net_income_usd": 27184000, "prior_year_revenue_usd": 165755000, "eps_basic_usd": 0.41, "eps_diluted_us... | {"company": "American Assets Trust, Inc", "period_end": "2019-06-30", "period_months": 6, "revenue_usd": 169432000, "net_income_usd": 27184000, "prior_year_revenue_usd": 165755000, "eps_basic_usd": 0.41, "eps_diluted_usd": 0.41, "revenue_yoy_pct": 2.2, "guidance": null, "non_gaap_eps_usd": null} | 1,500,217 | American Assets Trust, Inc. | 0001500217-19-000069 | https://www.sec.gov/Archives/edgar/data/1500217/000150021719000069/0001500217-19-000069-index.htm | 06dc0390e1511186dd1e04b0ada43c41a1594cb500e7d26d866a02ba979b6ab7 | 2019-07-30T00:00:00 | 2019-06-30T00:00:00 | 6 | 1,000 | 1 | 3 | 4 | [
0,
2,
1,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1500217/000150021719000069/0001500217-19-000069-index.htm | 1,778 |
ff-30000008 | 30,000,008 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in millions so 265.1 becomes 265100000; yoy = (265100000-371900000)/371900000 = -28.7%
{"company": "Innospec Inc", "period_end": "2020-09-30", "period_months": 3, "revenue_usd": 265100000, "net_income_usd": 12700000, "prior_year_revenue_usd": 371900000, "eps_basic_usd": 0.52, "eps_diluted_usd": 0.51, "reve... | {"company": "Innospec Inc", "period_end": "2020-09-30", "period_months": 3, "revenue_usd": 265100000, "net_income_usd": 12700000, "prior_year_revenue_usd": 371900000, "eps_basic_usd": 0.52, "eps_diluted_usd": 0.51, "revenue_yoy_pct": -28.7, "guidance": null, "non_gaap_eps_usd": 0.71} | 1,054,905 | INNOSPEC INC. | 0001193125-20-285089 | https://www.sec.gov/Archives/edgar/data/1054905/000119312520285089/0001193125-20-285089-index.htm | 88fae91ecbbc516b17d8a8a2fa119ce7e466cf804ee769fa032f952380d10ec5 | 2020-11-04T00:00:00 | 2020-09-30T00:00:00 | 3 | 1,000,000 | 0 | 1 | 4 | [
0,
1,
2,
3
] | false | [
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1054905/000119312520285089/0001193125-20-285089-index.htm | 1,693 |
ff-30000009 | 30,000,009 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 15,098,258 becomes 15098258000; yoy = (15098258000-12657285000)/12657285000 = 19.3%
{"company": "QUANTA SERVICES", "period_end": "2023-09-30", "period_months": 9, "revenue_usd": 15098258000, "net_income_usd": 537079000, "prior_year_revenue_usd": 12657285000, "eps_basic_usd": 3.68, "eps_dilu... | {"company": "QUANTA SERVICES", "period_end": "2023-09-30", "period_months": 9, "revenue_usd": 15098258000, "net_income_usd": 537079000, "prior_year_revenue_usd": 12657285000, "eps_basic_usd": 3.68, "eps_diluted_usd": 3.59, "revenue_yoy_pct": 19.3, "guidance": null, "non_gaap_eps_usd": null} | 1,050,915 | QUANTA SERVICES, INC. | 0001193125-23-268784 | https://www.sec.gov/Archives/edgar/data/1050915/000119312523268784/0001193125-23-268784-index.htm | c6bcdb3e35a150dff247cf66919670a87eb9dc8af565f5150d706ed0e471084c | 2023-11-02T00:00:00 | 2023-09-30T00:00:00 | 9 | 1,000 | 0 | 3 | 4 | [
2,
0,
1,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1050915/000119312523268784/0001193125-23-268784-index.htm | 1,804 |
ff-30000010 | 30,000,010 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 162,974 becomes 162974000; yoy = (162974000-140507000)/140507000 = 16.0%
{"company": "Box, Inc", "period_end": "2019-04-30", "period_months": 3, "revenue_usd": 162974000, "net_income_usd": -36828000, "prior_year_revenue_usd": 140507000, "eps_basic_usd": -0.25, "eps_diluted_usd": -0.25, "rev... | {"company": "Box, Inc", "period_end": "2019-04-30", "period_months": 3, "revenue_usd": 162974000, "net_income_usd": -36828000, "prior_year_revenue_usd": 140507000, "eps_basic_usd": -0.25, "eps_diluted_usd": -0.25, "revenue_yoy_pct": 16.0, "guidance": null, "non_gaap_eps_usd": null} | 1,372,612 | BOX INC | 0001564590-19-021446 | https://www.sec.gov/Archives/edgar/data/1372612/000156459019021446/0001564590-19-021446-index.htm | df1c29dc7d95ad7c35f0444bb91c81bd427554d4b9a7ca7204437cc75a3cbeb4 | 2019-06-03T00:00:00 | 2019-04-30T00:00:00 | 3 | 1,000 | 1 | 0 | 2 | [
1,
0
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1372612/000156459019021446/0001564590-19-021446-index.htm | 1,598 |
ff-30000011 | 30,000,011 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in millions so 988.5 becomes 988500000; yoy = (988500000-976700000)/976700000 = 1.2%
{"company": "Aerojet Rocketdyne Holdings, Inc", "period_end": "2020-06-30", "period_months": 6, "revenue_usd": 988500000, "net_income_usd": 70600000, "prior_year_revenue_usd": 976700000, "eps_basic_usd": 0.9, "eps_diluted_... | {"company": "Aerojet Rocketdyne Holdings, Inc", "period_end": "2020-06-30", "period_months": 6, "revenue_usd": 988500000, "net_income_usd": 70600000, "prior_year_revenue_usd": 976700000, "eps_basic_usd": 0.9, "eps_diluted_usd": 0.84, "revenue_yoy_pct": 1.2, "guidance": null, "non_gaap_eps_usd": null} | 40,888 | AEROJET ROCKETDYNE HOLDINGS, INC. | 0001171843-20-005267 | https://www.sec.gov/Archives/edgar/data/40888/000117184320005267/0001171843-20-005267-index.htm | cfb3d52968a53c410acf8be34242a1b63a634072ded5c6cb8a12ea27384f3288 | 2020-07-27T00:00:00 | 2020-06-30T00:00:00 | 6 | 1,000,000 | 1 | 3 | 4 | [
0,
2,
1,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/40888/000117184320005267/0001171843-20-005267-index.htm | 1,652 |
ff-30000012 | 30,000,012 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 2,443 becomes 2443000; yoy = (2443000-7191000)/7191000 = -66.0%
{"company": "Gritstone bio, Inc", "period_end": "2023-03-31", "period_months": 3, "revenue_usd": 2443000, "net_income_usd": -33982000, "prior_year_revenue_usd": 7191000, "eps_basic_usd": -0.3, "eps_diluted_usd": -0.3, "revenue_... | {"company": "Gritstone bio, Inc", "period_end": "2023-03-31", "period_months": 3, "revenue_usd": 2443000, "net_income_usd": -33982000, "prior_year_revenue_usd": 7191000, "eps_basic_usd": -0.3, "eps_diluted_usd": -0.3, "revenue_yoy_pct": -66.0, "guidance": null, "non_gaap_eps_usd": null} | 1,656,634 | Gritstone bio, Inc. | 0000950170-23-021226 | https://www.sec.gov/Archives/edgar/data/1656634/000095017023021226/0000950170-23-021226-index.htm | c40a9f5c6c4ca81209da6f94912e04eecd0c469b84a7525d0ec1847f9c1eee76 | 2023-05-11T00:00:00 | 2023-03-31T00:00:00 | 3 | 1,000 | 1 | 0 | 2 | [
1,
0
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1656634/000095017023021226/0000950170-23-021226-index.htm | 1,472 |
ff-30000013 | 30,000,013 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 1,445 becomes 1445000; yoy = (1445000-5301000)/5301000 = -72.7%
{"company": "HOOKIPA Pharma Inc", "period_end": "2022-03-31", "period_months": 3, "revenue_usd": 1445000, "net_income_usd": -17968000, "prior_year_revenue_usd": 5301000, "eps_basic_usd": -0.4, "eps_diluted_usd": -0.4, "revenue_... | {"company": "HOOKIPA Pharma Inc", "period_end": "2022-03-31", "period_months": 3, "revenue_usd": 1445000, "net_income_usd": -17968000, "prior_year_revenue_usd": 5301000, "eps_basic_usd": -0.4, "eps_diluted_usd": -0.4, "revenue_yoy_pct": -72.7, "guidance": null, "non_gaap_eps_usd": null} | 1,760,542 | HOOKIPA Pharma Inc. | 0001558370-22-008849 | https://www.sec.gov/Archives/edgar/data/1760542/000155837022008849/0001558370-22-008849-index.htm | ccd9610001cb6e2616bb7eb0c9fcefbf16bb68270c400b17ffaf0c77a8257597 | 2022-05-16T00:00:00 | 2022-03-31T00:00:00 | 3 | 1,000 | 0 | 1 | 2 | [
0,
1
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1760542/000155837022008849/0001558370-22-008849-index.htm | 1,418 |
ff-30000014 | 30,000,014 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 64,457 becomes 64457000; yoy = (64457000-68302000)/68302000 = -5.6%
{"company": "Nobilis Health Corp", "period_end": "2018-03-31", "period_months": 3, "revenue_usd": 64457000, "net_income_usd": -3141000, "prior_year_revenue_usd": 68302000, "eps_basic_usd": -0.05, "eps_diluted_usd": -0.05, "... | {"company": "Nobilis Health Corp", "period_end": "2018-03-31", "period_months": 3, "revenue_usd": 64457000, "net_income_usd": -3141000, "prior_year_revenue_usd": 68302000, "eps_basic_usd": -0.05, "eps_diluted_usd": -0.05, "revenue_yoy_pct": -5.6, "guidance": null, "non_gaap_eps_usd": null} | 1,409,916 | Nobilis Health Corp. | 0001628280-18-006232 | https://www.sec.gov/Archives/edgar/data/1409916/000162828018006232/0001628280-18-006232-index.htm | 0d91c7dc9290041edc47eb7700b0d1b6b199ce287dba2423fcc620456eee0b80 | 2018-05-08T00:00:00 | 2018-03-31T00:00:00 | 3 | 1,000 | 1 | 0 | 2 | [
1,
0
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1409916/000162828018006232/0001628280-18-006232-index.htm | 1,556 |
ff-30000015 | 30,000,015 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in millions so 4,760 becomes 4760000000; yoy = (4760000000-4812000000)/4812000000 = -1.1%
{"company": "COGNIZANT", "period_end": "2024-03-31", "period_months": 3, "revenue_usd": 4760000000, "net_income_usd": 546000000, "prior_year_revenue_usd": 4812000000, "eps_basic_usd": 1.1, "eps_diluted_usd": 1.1, "rev... | {"company": "COGNIZANT", "period_end": "2024-03-31", "period_months": 3, "revenue_usd": 4760000000, "net_income_usd": 546000000, "prior_year_revenue_usd": 4812000000, "eps_basic_usd": 1.1, "eps_diluted_usd": 1.1, "revenue_yoy_pct": -1.1, "guidance": null, "non_gaap_eps_usd": null} | 1,058,290 | COGNIZANT TECHNOLOGY SOLUTIONS CORP | 0001058290-24-000115 | https://www.sec.gov/Archives/edgar/data/1058290/000105829024000115/0001058290-24-000115-index.htm | 3302a71049675912ddcdc60bfc76a5b68d95c249857c6dad4a3f1a2395e60275 | 2024-05-01T00:00:00 | 2024-03-31T00:00:00 | 3 | 1,000,000 | 0 | 1 | 2 | [
0,
1
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1058290/000105829024000115/0001058290-24-000115-index.htm | 1,468 |
ff-30000016 | 30,000,016 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 12,724 becomes 12724000; yoy = (12724000-9447000)/9447000 = 34.7%
{"company": "POSTAL REALTY TRUST, INC", "period_end": "2022-06-30", "period_months": 3, "revenue_usd": 12724000, "net_income_usd": 1165000, "prior_year_revenue_usd": 9447000, "eps_basic_usd": 0.04, "eps_diluted_usd": 0.04, "r... | {"company": "POSTAL REALTY TRUST, INC", "period_end": "2022-06-30", "period_months": 3, "revenue_usd": 12724000, "net_income_usd": 1165000, "prior_year_revenue_usd": 9447000, "eps_basic_usd": 0.04, "eps_diluted_usd": 0.04, "revenue_yoy_pct": 34.7, "guidance": null, "non_gaap_eps_usd": null} | 1,759,774 | Postal Realty Trust, Inc. | 0001759774-22-000024 | https://www.sec.gov/Archives/edgar/data/1759774/000175977422000024/0001759774-22-000024-index.htm | e87fe786fd7bebd941c2890a1f9d903bb7d8e2cd28559ea9a44d2f738f67e38c | 2022-08-02T00:00:00 | 2022-06-30T00:00:00 | 3 | 1,000 | 2 | 0 | 4 | [
1,
2,
0,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1759774/000175977422000024/0001759774-22-000024-index.htm | 1,740 |
ff-30000017 | 30,000,017 | train | period_hard | period_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table already in dollars so 63008100 stands; yoy = (63008100-33879208)/33879208 = 86.0%
{"company": "ADMA Biologics, Inc", "period_end": "2022-06-30", "period_months": 6, "revenue_usd": 63008100, "net_income_usd": -38772763, "prior_year_revenue_usd": 33879208, "eps_basic_usd": -0.2, "eps_diluted_usd": -0.2, "rev... | {"company": "ADMA Biologics, Inc", "period_end": "2022-06-30", "period_months": 6, "revenue_usd": 63008100, "net_income_usd": -38772763, "prior_year_revenue_usd": 33879208, "eps_basic_usd": -0.2, "eps_diluted_usd": -0.2, "revenue_yoy_pct": 86.0, "guidance": null, "non_gaap_eps_usd": null} | 1,368,514 | ADMA BIOLOGICS, INC. | 0001140361-22-029079 | https://www.sec.gov/Archives/edgar/data/1368514/000114036122029079/0001140361-22-029079-index.htm | 3e5aaebdd245057ba676a6514a2a4ec48d8ad894c2bed0847c4d3beeb7cba319 | 2022-08-10T00:00:00 | 2022-06-30T00:00:00 | 6 | 1 | 0 | 3 | 4 | [
2,
0,
1,
3
] | false | [
"null_discipline",
"period_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1368514/000114036122029079/0001140361-22-029079-index.htm | 1,957 |
ff-30000018 | 30,000,018 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 22,505 becomes 22505000; yoy = (22505000-16106000)/16106000 = 39.7%
{"company": "Strongbridge Biopharma plc", "period_end": "2020-09-30", "period_months": 9, "revenue_usd": 22505000, "net_income_usd": -33129000, "prior_year_revenue_usd": 16106000, "eps_basic_usd": -0.6, "eps_diluted_usd": -... | {"company": "Strongbridge Biopharma plc", "period_end": "2020-09-30", "period_months": 9, "revenue_usd": 22505000, "net_income_usd": -33129000, "prior_year_revenue_usd": 16106000, "eps_basic_usd": -0.6, "eps_diluted_usd": -0.6, "revenue_yoy_pct": 39.7, "guidance": null, "non_gaap_eps_usd": null} | 1,634,432 | Strongbridge Biopharma plc | 0001634432-20-000039 | https://www.sec.gov/Archives/edgar/data/1634432/000163443220000039/0001634432-20-000039-index.htm | 3838346eb8ca03de920d524bcaf33ad754cf492d0aebfaffad3b950e209c85c6 | 2020-10-29T00:00:00 | 2020-09-30T00:00:00 | 9 | 1,000 | 2 | 3 | 4 | [
0,
1,
2,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1634432/000163443220000039/0001634432-20-000039-index.htm | 1,891 |
ff-30000019 | 30,000,019 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 1,049,040 becomes 1049040000; yoy = (1049040000-918834000)/918834000 = 14.2%
{"company": "National Instruments Corporation", "period_end": "2021-09-30", "period_months": 9, "revenue_usd": 1049040000, "net_income_usd": 48979000, "prior_year_revenue_usd": 918834000, "eps_basic_usd": 0.37, "ep... | {"company": "National Instruments Corporation", "period_end": "2021-09-30", "period_months": 9, "revenue_usd": 1049040000, "net_income_usd": 48979000, "prior_year_revenue_usd": 918834000, "eps_basic_usd": 0.37, "eps_diluted_usd": 0.37, "revenue_yoy_pct": 14.2, "guidance": null, "non_gaap_eps_usd": null} | 935,494 | NATIONAL INSTRUMENTS CORP | 0001140361-21-035748 | https://www.sec.gov/Archives/edgar/data/935494/000114036121035748/0001140361-21-035748-index.htm | 766c7d1462c3b11b703b3fc11738872a44a72730bd735b5c65bd2b163e04e8f2 | 2021-10-28T00:00:00 | 2021-09-30T00:00:00 | 9 | 1,000 | 1 | 3 | 4 | [
0,
2,
1,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/935494/000114036121035748/0001140361-21-035748-index.htm | 1,888 |
ff-30000020 | 30,000,020 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 1,548,980 becomes 1548980000; yoy = (1548980000-1583077000)/1583077000 = -2.2%
{"company": "Academy Sports", "period_end": "2024-08-03", "period_months": 3, "revenue_usd": 1548980000, "net_income_usd": 142588000, "prior_year_revenue_usd": 1583077000, "eps_basic_usd": 1.99, "eps_diluted_usd"... | {"company": "Academy Sports", "period_end": "2024-08-03", "period_months": 3, "revenue_usd": 1548980000, "net_income_usd": 142588000, "prior_year_revenue_usd": 1583077000, "eps_basic_usd": 1.99, "eps_diluted_usd": 1.95, "revenue_yoy_pct": -2.2, "guidance": null, "non_gaap_eps_usd": null} | 1,817,358 | Academy Sports & Outdoors, Inc. | 0001817358-24-000170 | https://www.sec.gov/Archives/edgar/data/1817358/000181735824000170/0001817358-24-000170-index.htm | 21e32046c0fa43cee01e2a0a6c3ffd4cfec4b8c8a732c740cde9f8ad6ea54e49 | 2024-09-10T00:00:00 | 2024-08-03T00:00:00 | 3 | 1,000 | 3 | 1 | 4 | [
1,
2,
3,
0
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1817358/000181735824000170/0001817358-24-000170-index.htm | 1,589 |
ff-30000021 | 30,000,021 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 311,189 becomes 311189000; yoy = (311189000-314448000)/314448000 = -1.0%
{"company": "Alamo Group Inc", "period_end": "2021-03-31", "period_months": 3, "revenue_usd": 311189000, "net_income_usd": 17462000, "prior_year_revenue_usd": 314448000, "eps_basic_usd": 1.48, "eps_diluted_usd": 1.47, ... | {"company": "Alamo Group Inc", "period_end": "2021-03-31", "period_months": 3, "revenue_usd": 311189000, "net_income_usd": 17462000, "prior_year_revenue_usd": 314448000, "eps_basic_usd": 1.48, "eps_diluted_usd": 1.47, "revenue_yoy_pct": -1.0, "guidance": null, "non_gaap_eps_usd": null} | 897,077 | ALAMO GROUP INC | 0000897077-21-000041 | https://www.sec.gov/Archives/edgar/data/897077/000089707721000041/0000897077-21-000041-index.htm | b19ecfdef4d1f21d06e1aa396f9ac34f83f2f9c45db1804fec2b04eeae3a5084 | 2021-05-06T00:00:00 | 2021-03-31T00:00:00 | 3 | 1,000 | 1 | 0 | 2 | [
1,
0
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/897077/000089707721000041/0000897077-21-000041-index.htm | 1,675 |
ff-30000022 | 30,000,022 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 107,891 becomes 107891000; yoy = (107891000-78088000)/78088000 = 38.2%
{"company": "LivePerson, Inc", "period_end": "2021-03-31", "period_months": 3, "revenue_usd": 107891000, "net_income_usd": -21195000, "prior_year_revenue_usd": 78088000, "eps_basic_usd": -0.31, "eps_diluted_usd": -0.31, ... | {"company": "LivePerson, Inc", "period_end": "2021-03-31", "period_months": 3, "revenue_usd": 107891000, "net_income_usd": -21195000, "prior_year_revenue_usd": 78088000, "eps_basic_usd": -0.31, "eps_diluted_usd": -0.31, "revenue_yoy_pct": 38.2, "guidance": null, "non_gaap_eps_usd": null} | 1,102,993 | LIVEPERSON INC | 0001102993-21-000079 | https://www.sec.gov/Archives/edgar/data/1102993/000110299321000079/0001102993-21-000079-index.htm | b46d25844f70ba82b71e79c2ead15bb08d3084ff1daba3763ec657fd129f9b7b | 2021-05-04T00:00:00 | 2021-03-31T00:00:00 | 3 | 1,000 | 0 | 1 | 2 | [
0,
1
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1102993/000110299321000079/0001102993-21-000079-index.htm | 1,553 |
ff-30000023 | 30,000,023 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 73,834 becomes 73834000; yoy = (73834000-25786000)/25786000 = 186.3%
{"company": "Esperion", "period_end": "2024-06-30", "period_months": 3, "revenue_usd": 73834000, "net_income_usd": -61925000, "prior_year_revenue_usd": 25786000, "eps_basic_usd": -0.33, "eps_diluted_usd": -0.33, "revenue_y... | {"company": "Esperion", "period_end": "2024-06-30", "period_months": 3, "revenue_usd": 73834000, "net_income_usd": -61925000, "prior_year_revenue_usd": 25786000, "eps_basic_usd": -0.33, "eps_diluted_usd": -0.33, "revenue_yoy_pct": 186.3, "guidance": null, "non_gaap_eps_usd": null} | 1,434,868 | Esperion Therapeutics, Inc. | 0001628280-24-036597 | https://www.sec.gov/Archives/edgar/data/1434868/000162828024036597/0001628280-24-036597-index.htm | 7418c6323e0926f6c288777bf10c44d3878e2246f9edc10f7466e74f6c3c6b0b | 2024-08-12T00:00:00 | 2024-06-30T00:00:00 | 3 | 1,000 | 2 | 0 | 4 | [
1,
2,
0,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1434868/000162828024036597/0001628280-24-036597-index.htm | 1,630 |
ff-30000024 | 30,000,024 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 291,356 becomes 291356000; yoy = (291356000-295137000)/295137000 = -1.3%
{"company": "Endurance International Group Holdings, Inc", "period_end": "2018-03-31", "period_months": 3, "revenue_usd": 291356000, "net_income_usd": -7088000, "prior_year_revenue_usd": 295137000, "eps_basic_usd": -0.... | {"company": "Endurance International Group Holdings, Inc", "period_end": "2018-03-31", "period_months": 3, "revenue_usd": 291356000, "net_income_usd": -7088000, "prior_year_revenue_usd": 295137000, "eps_basic_usd": -0.05, "eps_diluted_usd": -0.05, "revenue_yoy_pct": -1.3, "guidance": null, "non_gaap_eps_usd": null} | 1,237,746 | Endurance International Group Holdings, Inc. | 0001193125-18-144970 | https://www.sec.gov/Archives/edgar/data/1237746/000119312518144970/0001193125-18-144970-index.htm | 277d628d82aa83ee948bc65295cc048f5de1206cc411d9d960fb9413aae64461 | 2018-05-01T00:00:00 | 2018-03-31T00:00:00 | 3 | 1,000 | 0 | 1 | 2 | [
1,
0
] | false | [
"null_discipline",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1237746/000119312518144970/0001193125-18-144970-index.htm | 1,587 |
ff-30000025 | 30,000,025 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in millions so 641.7 becomes 641700000; yoy = (641700000-472700000)/472700000 = 35.8%
{"company": "Trinity Industries, Inc", "period_end": "2023-03-31", "period_months": 3, "revenue_usd": 641700000, "net_income_usd": 13700000, "prior_year_revenue_usd": 472700000, "eps_basic_usd": 0.05, "eps_diluted_usd": 0... | {"company": "Trinity Industries, Inc", "period_end": "2023-03-31", "period_months": 3, "revenue_usd": 641700000, "net_income_usd": 13700000, "prior_year_revenue_usd": 472700000, "eps_basic_usd": 0.05, "eps_diluted_usd": 0.05, "revenue_yoy_pct": 35.8, "guidance": null, "non_gaap_eps_usd": 0.07} | 99,780 | TRINITY INDUSTRIES INC | 0000099780-23-000043 | https://www.sec.gov/Archives/edgar/data/99780/000009978023000043/0000099780-23-000043-index.htm | 7aa900d8b6707e9c04b81a74e42b71783ef44c15971cbb086223e2398e66c06f | 2023-05-02T00:00:00 | 2023-03-31T00:00:00 | 3 | 1,000,000 | 1 | 0 | 2 | [
1,
0
] | false | [
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/99780/000009978023000043/0000099780-23-000043-index.htm | 1,574 |
ff-30000026 | 30,000,026 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 653,472 becomes 653472000; yoy = (653472000-645699000)/645699000 = 1.2%
{"company": "Quanex Building Products Corporation", "period_end": "2019-07-31", "period_months": 9, "revenue_usd": 653472000, "net_income_usd": -15782000, "prior_year_revenue_usd": 645699000, "eps_basic_usd": -0.48, "ep... | {"company": "Quanex Building Products Corporation", "period_end": "2019-07-31", "period_months": 9, "revenue_usd": 653472000, "net_income_usd": -15782000, "prior_year_revenue_usd": 645699000, "eps_basic_usd": -0.48, "eps_diluted_usd": -0.48, "revenue_yoy_pct": 1.2, "guidance": null, "non_gaap_eps_usd": null} | 1,423,221 | Quanex Building Products CORP | 0001171843-19-005830 | https://www.sec.gov/Archives/edgar/data/1423221/000117184319005830/0001171843-19-005830-index.htm | ee8214cffb39d5f4259240726e6c6d0c3375e98b2812a5c7fbdb86ad2cc6adc3 | 2019-09-05T00:00:00 | 2019-07-31T00:00:00 | 9 | 1,000 | 3 | 2 | 4 | [
0,
1,
3,
2
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1423221/000117184319005830/0001171843-19-005830-index.htm | 1,687 |
ff-30000027 | 30,000,027 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 343,112 becomes 343112000; yoy = (343112000-293179000)/293179000 = 17.0%
{"company": "Via Renewables, Inc", "period_end": "2022-09-30", "period_months": 9, "revenue_usd": 343112000, "net_income_usd": 38691000, "prior_year_revenue_usd": 293179000, "eps_basic_usd": 0.7, "eps_diluted_usd": 0.7... | {"company": "Via Renewables, Inc", "period_end": "2022-09-30", "period_months": 9, "revenue_usd": 343112000, "net_income_usd": 38691000, "prior_year_revenue_usd": 293179000, "eps_basic_usd": 0.7, "eps_diluted_usd": 0.7, "revenue_yoy_pct": 17.0, "guidance": null, "non_gaap_eps_usd": null} | 1,606,268 | Via Renewables, Inc. | 0001606268-22-000089 | https://www.sec.gov/Archives/edgar/data/1606268/000160626822000089/0001606268-22-000089-index.htm | 3ed9bd59392ec716afa65b5ab1829d663cd4c16fbee5b122a604497ccf0f0364 | 2022-11-03T00:00:00 | 2022-09-30T00:00:00 | 9 | 1,000 | 0 | 3 | 4 | [
2,
0,
1,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1606268/000160626822000089/0001606268-22-000089-index.htm | 1,863 |
ff-30000028 | 30,000,028 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 86,423 becomes 86423000; yoy = (86423000-83004000)/83004000 = 4.1%
{"company": "Synchronoss Technologies Inc", "period_end": "2024-06-30", "period_months": 6, "revenue_usd": 86423000, "net_income_usd": 3981000, "prior_year_revenue_usd": 83004000, "eps_basic_usd": 0.24, "eps_diluted_usd": 0.... | {"company": "Synchronoss Technologies Inc", "period_end": "2024-06-30", "period_months": 6, "revenue_usd": 86423000, "net_income_usd": 3981000, "prior_year_revenue_usd": 83004000, "eps_basic_usd": 0.24, "eps_diluted_usd": 0.24, "revenue_yoy_pct": 4.1, "guidance": null, "non_gaap_eps_usd": null} | 1,131,554 | SYNCHRONOSS TECHNOLOGIES INC | 0001131554-24-000068 | https://www.sec.gov/Archives/edgar/data/1131554/000113155424000068/0001131554-24-000068-index.htm | 127082df1429f9180149f2066066486b21e3827946c4aa878630bf13a9dcf032 | 2024-08-06T00:00:00 | 2024-06-30T00:00:00 | 6 | 1,000 | 3 | 2 | 4 | [
0,
1,
3,
2
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1131554/000113155424000068/0001131554-24-000068-index.htm | 1,896 |
ff-30000029 | 30,000,029 | train | scale_hard | scale_hard | You are an earnings desk assistant at a news wire. Read the press release excerpt below and emit the earnings flash for the requested period.
First write one line beginning "notes:" that states the table's scale normalization and the year over year arithmetic in one short sentence. Then write one JSON object with exac... | notes: table in thousands so 105,132 becomes 105132000; yoy = (105132000-91154000)/91154000 = 15.3%
{"company": "AvidXchange Holdings, Inc", "period_end": "2024-06-30", "period_months": 3, "revenue_usd": 105132000, "net_income_usd": 436000, "prior_year_revenue_usd": 91154000, "eps_basic_usd": 0.0, "eps_diluted_usd": 0.... | {"company": "AvidXchange Holdings, Inc", "period_end": "2024-06-30", "period_months": 3, "revenue_usd": 105132000, "net_income_usd": 436000, "prior_year_revenue_usd": 91154000, "eps_basic_usd": 0.0, "eps_diluted_usd": 0.0, "revenue_yoy_pct": 15.3, "guidance": null, "non_gaap_eps_usd": null} | 1,858,257 | AvidXchange Holdings, Inc. | 0000950170-24-088223 | https://www.sec.gov/Archives/edgar/data/1858257/000095017024088223/0000950170-24-088223-index.htm | e4286e46e7c292ceddb00a7165cdfa5f9b80b5e575cfe8c51e6aba38871b2904 | 2024-07-31T00:00:00 | 2024-06-30T00:00:00 | 3 | 1,000 | 2 | 0 | 4 | [
1,
2,
0,
3
] | false | [
"null_discipline",
"period_hard",
"scale_hard"
] | false | SEC EDGAR public records, no formal license text | https://www.sec.gov/Archives/edgar/data/1858257/000095017024088223/0000950170-24-088223-index.htm | 1,736 |
End of preview.
YAML Metadata Warning:The task_categories "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other
FlashFacts: 13,976 rows of earnings extraction, every one from a real SEC filing
A training corpus for pulling structured financial facts out of an earnings exhibit: the revenue
figure, the period it covers, the units it is reported in, and the year-over-year change. The model
must return exact JSON, and it must return null when the filing does not contain the fact.
What this dataset proves, and how you check it
| rows | 13,976, and every one is a real SEC EX-99 filing. Zero synthetic rows |
| traceability | each row carries its SEC accession number and a SHA256 of the source exhibit |
| labels authored or judged by a language model | zero |
| what it does to a model | gemma-3-4b-it: 0.0% to 90.8% exact match, held out |
| rows the base model wins, across 650 held out | 0 |
| significance | p = 1.1e-109, exact McNemar |
| can the score be faked by position | no: shortcut gain measured at 0.00 against the majority baseline |
| decontamination | screened against FinQA, ConvFinQA and TatQA, 3,250 questions, max 4-gram Jaccard 0.0151 |
| rows dropped rather than truncated | 920, because a truncated completion is a poisoned label |
| what ships beside the rows | scorer, verifier, every eval slice including the OOD anchor, per-row verdicts |
Every number above is recomputable on a laptop with no GPU, from files in this repository.
Built for the Adaption AutoScientist Challenge, Part 2 (Market Analysis and News), and co-optimized with Adaptive Data (Adaption Labs).
Not one synthetic row
| rows | 13,976 |
| real filings | 13,976 (100%) |
| rendered or synthetic | 0 |
| distinct companies (CIK) | 1,607 |
| distinct exhibits | 9,070 |
| filing window | 2018-01-04 to 2024-12-20 |
Every prompt is an excerpt from an actual EX-99 earnings exhibit filed with the SEC, and every label was read out of that exhibit's own table and checked against it. Nothing was written by a language model and nothing was judged by one.
The build could have hit its row target faster by rendering synthetic tables. It did not, because the headline evaluation is on real filings either way, and a corpus that is 70% synthetic cannot support a claim about real ones.
The shortcut test, and why the number is zero
The obvious way to fake this task is positional: always read the first numeric column. So the build measures it directly.
| accuracy from column position alone | 57.58% |
| majority-class baseline | 57.58% |
| shortcut gain | 0.00 |
A classifier that sees only which column a number sits in does exactly as well as always guessing the most common answer, and no better. The positional shortcut was present in an early build at 0.189 and was engineered out. A model that scores on this corpus has to read the table.
What makes a row hard
| trait | rows | what it requires |
|---|---|---|
scale_hard |
13,045 | the value is in thousands or millions and must be normalized |
period_hard |
9,560 | quarter against year-to-date must be disambiguated |
null_discipline |
12,888 | at least one requested field is genuinely absent |
plain |
8 | none of the above |
Traits are multi-label, so a single row is usually scale-hard and period-hard and null-bearing at once.
Every row in the corpus has at least one null field. This is the part that most extraction corpora omit and the part that decides whether a model is usable: a system that invents a plausible revenue figure when the filing does not state one is worse than useless in this domain. Abstention is graded here, not assumed.
Period coverage is deliberately unbalanced toward quarters, matching how companies actually file:
| period | rows |
|---|---|
| 3 months | 8,047 |
| 9 months | 2,398 |
| 6 months | 2,296 |
| 12 months | 1,235 |
Decontamination
Checked against three public financial QA benchmarks by 8-gram overlap and 4-gram Jaccard at a 0.6 threshold, over the excerpt and completion only:
| reference set | questions | contaminated rows |
|---|---|---|
| FinQA | 1,147 | 0 |
| ConvFinQA | 434 | 0 |
| TatQA | 1,669 | 0 |
Maximum 4-gram Jaccard against any reference question was 0.0151, far below the 0.6 threshold.
The shared instruction block is excluded from the comparison on purpose: every row carries it, so including it would measure our own boilerplate rather than contamination.
Not covered: FinanceBench is named in the build spec but was not present locally, so this run does not cover it. Stated here rather than omitted, because a decontamination claim that quietly skips a benchmark is worse than one with a declared gap.
Held-out slices
Disjoint from training on filing, company and period:
| slice | rows | what it isolates |
|---|---|---|
| held-out | 400-row held-out set from the same distribution | the competition's own yardstick |
| hard | 150-row hard shard | scale and period disambiguation together |
| external | 100-row out-of-distribution anchor from 10-Q filings | a different document type entirely |
| Day-0 | 60 Day-0 probes | the pre-registered base-model gate |
Token budget, measured
max_len is measured with the real gemma-3-4b-it tokenizer, not assumed:
| tokens | |
|---|---|
| median (prompt plus completion) | 1,716 |
| p95 | 1,944 |
| maximum | 2,000 |
Training therefore requires a 2048 window. At 1536 80.2% of rows would truncate and at 1024 100% would. Under completion-only masking a truncated completion is a silently poisoned label, so 920 rows over budget were dropped at build time rather than trained on in mutilated form.
Licensing
SEC filings are US government works in the public domain, and the excerpts are reproduced from EDGAR. Each row carries its accession number and a SHA256 of the source exhibit, so any label can be traced back to the exact document it came from and rechecked.
Try it, and everything that backs it
Live side by side demo: https://huggingface.co/spaces/Jainamshahhh/flashfacts-demo Enter your own input and watch the base model and FlashFacts-4B answer it under identical greedy decoding. The GPU backend scales to zero, so a cold first request takes about a minute.
Released on both platforms, with the scorer, every eval slice, the per row verdicts and
significance.py alongside, so every number on this page can be recomputed rather than
trusted.
Limitations
US registrants filing in English with the SEC, 2018 to 2024, EX-99 earnings exhibits. Revenue and period extraction specifically; not a general financial reasoning corpus, not guidance or sentiment, and no forward-looking statements. Not investment advice.
- Downloads last month
- 9