Hugging Face's logo

Datasets:
Nandan007
/
NepFakeV2

Tasks:
Text Classification
Languages:
Nepali
Size:
1K<n<10K
Tags:
fact-checking
misinformation
fake-news
nepali
low-resource
nlp
License:
Dataset card Data Studio Files
xet
 Community
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
last_updated: string
total_examples: int64
total_including_unknown: int64
label_distribution: struct<REAL: struct<count: int64, pct: string>, FALSE_MISLEADING: struct<count: int64, pct: string>, (... 47 chars omitted)
  child 0, REAL: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 1, FALSE_MISLEADING: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 2, UNVERIFIED: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
source_distribution: struct<nepalfactcheck: struct<count: int64, pct: string>, techpana: struct<count: int64, pct: string (... 2 chars omitted)
  child 0, nepalfactcheck: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 1, techpana: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
topic_distribution: struct<politics: struct<count: int64, pct: string>, other: struct<count: int64, pct: string>, enviro (... 219 chars omitted)
  child 0, politics: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 1, other: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 2, environment: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 3, technology: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 4, health: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 5, society: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
  child 6, economy: struct<count: int64, pct: string>
      child 0, count: int64
      child 1, pct: string
date_range: struct<oldest: timestamp[s], newest: timestamp[s], total_with_date: int64, total_without_date: int64 (... 1 chars omitted)
  child 0, oldest: timestamp[s]
  child 1, newest: timestamp[s]
  child 2, total_with_date: int64
  child 3, total_without_date: int64
language: struct<nepali_script: int64, nepali_pct: string>
  child 0, nepali_script: int64
  child 1, nepali_pct: string
quality: struct<flagged_for_review: int64, flagged_pct: string>
  child 0, flagged_for_review: int64
  child 1, flagged_pct: string
schema_version: string
to
{'example_id': Value('string'), 'claim_text': Value('string'), 'verdict_label': Value('int64'), 'verdict_label_text': Value('string'), 'evidence_text': Value('string'), 'evidence_sentences': Value('string'), 'external_evidence_urls': Value('string'), 'source_name': Value('string'), 'source_type': Value('string'), 'source_url': Value('string'), 'source_outlet': Value('string'), 'date_published': Value('timestamp[s]'), 'is_native_nepali': Value('bool'), 'label_basis': Value('string'), 'topic_category': Value('string'), 'annotator_notes': Value('string'), 'split': Value('string'), 'schema_version': 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 1869, in _prepare_split_single
                  num_examples, num_bytes = writer.finalize()
                                            ~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 781, in finalize
                  self.write_rows_on_file()
                  ~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 663, in write_rows_on_file
                  self._write_table(table)
                  ~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              last_updated: string
              total_examples: int64
              total_including_unknown: int64
              label_distribution: struct<REAL: struct<count: int64, pct: string>, FALSE_MISLEADING: struct<count: int64, pct: string>, (... 47 chars omitted)
                child 0, REAL: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 1, FALSE_MISLEADING: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 2, UNVERIFIED: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
              source_distribution: struct<nepalfactcheck: struct<count: int64, pct: string>, techpana: struct<count: int64, pct: string (... 2 chars omitted)
                child 0, nepalfactcheck: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 1, techpana: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
              topic_distribution: struct<politics: struct<count: int64, pct: string>, other: struct<count: int64, pct: string>, enviro (... 219 chars omitted)
                child 0, politics: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 1, other: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 2, environment: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 3, technology: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 4, health: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 5, society: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
                child 6, economy: struct<count: int64, pct: string>
                    child 0, count: int64
                    child 1, pct: string
              date_range: struct<oldest: timestamp[s], newest: timestamp[s], total_with_date: int64, total_without_date: int64 (... 1 chars omitted)
                child 0, oldest: timestamp[s]
                child 1, newest: timestamp[s]
                child 2, total_with_date: int64
                child 3, total_without_date: int64
              language: struct<nepali_script: int64, nepali_pct: string>
                child 0, nepali_script: int64
                child 1, nepali_pct: string
              quality: struct<flagged_for_review: int64, flagged_pct: string>
                child 0, flagged_for_review: int64
                child 1, flagged_pct: string
              schema_version: string
              to
              {'example_id': Value('string'), 'claim_text': Value('string'), 'verdict_label': Value('int64'), 'verdict_label_text': Value('string'), 'evidence_text': Value('string'), 'evidence_sentences': Value('string'), 'external_evidence_urls': Value('string'), 'source_name': Value('string'), 'source_type': Value('string'), 'source_url': Value('string'), 'source_outlet': Value('string'), 'date_published': Value('timestamp[s]'), 'is_native_nepali': Value('bool'), 'label_basis': Value('string'), 'topic_category': Value('string'), 'annotator_notes': Value('string'), 'split': Value('string'), 'schema_version': 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 1694, 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 1880, 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.

example_id
string
claim_text
string
verdict_label
int64
verdict_label_text
string
evidence_text
string
evidence_sentences
string
external_evidence_urls
string
source_name
string
source_type
string
source_url
string
source_outlet
string
date_published
timestamp[us]
is_native_nepali
bool
label_basis
string
topic_category
string
annotator_notes
string
split
string
schema_version
string
NF2_20260910_0001
फ्याक्टचेक: जेनजी सहिदको सट्टा अर्थमन्त्रीको शालिक ? भाइरल तस्बिर वास्तविक होइन, एआई हो
1
FALSE_MISLEADING
फ्याक्टचेक: जेनजी सहिदको सट्टा अर्थमन्त्रीको शालिक ? भाइरल तस्बिर वास्तविक होइन, एआई हो रञ्जिता उप्रेती भदौ २५, २०८३ १९:२३ काठमाडौं । अर्थमन्त्री डा. स्वर्णिम वाग्लेको शालिक बनाइएको दाबीसहित यतिबेला सामाजिक सञ्जालमा एउटा तस्बिर भाइरल भइरहेको छ । जेनजी आन्दोलनका सहिदको शालिक बनाउन खोज्दा अर्थमन्त्री डा. वाग्लेको जस्तो द...
https://www.facebook.com/share/p/1EhDuA4Vuz/ | https://www.facebook.com/dipak.joshi.31968?__cft__[0]=AZhDQ8sbuZg9PAt5MBPFmp0gmTeQrNHsopBHwAWzmTjZo-cUR6JBfA83m7MXwwi9G9Q7ymcIyax2B_LpJYUKfpcTxBBY1qNPfKln3-HR0ozGs9nOA960AhHzISQ-Speo0GJlKGiKxCad4LcnM8w9GaVw64jlKr0F&__tn__=-UC%2CP-R | https://www.facebook.com/share/p/1BDEbG...
techpana
fact_checker
https://techpana.com/2026/159019/swornim-wagle-statue-viral-image-ai-generated-fact-check
2026-09-10T00:00:00
true
fact_checker_verdict
politics
1.0
NF2_20260910_0002
फ्याक्टचेकः १२ औं दिनमा हर्क साम्पाङले बेत्रावतीमा एक युवतीको जीवितै उद्धार गरे भन्दै फैलिएको भिडिओ एआई सिर्जित
1
FALSE_MISLEADING
फ्याक्टचेकः १२ औं दिनमा हर्क साम्पाङले बेत्रावतीमा एक युवतीको जीवितै उद्धार गरे भन्दै फैलिएको भिडिओ एआई सिर्जित हिक्मत आचार्य भदौ २५, २०८३ १८:२५ काठमाडौं । बुधबारदेखि सामाजिक सञ्जालका विभिन्न प्लेटफर्ममा श्रम संस्कृति पार्टीका अध्यक्ष एवम् प्रतिनिधि सभा सदस्य हर्कराज राई (हर्क सामपाङ) ले भोटेकोशी बाढीपछि बेपत्ता भएकी य...
https://www.facebook.com/share/r/1JqkZBEpPv/ | https://www.facebook.com/share/r/1DZ7c2fRLV/ | https://www.facebook.com/share/r/1CQKzqpbdc/ | https://www.facebook.com/reel/1427680442575748/ | https://www.facebook.com/share/r/1Dt5FgmAgU/
techpana
fact_checker
https://techpana.com/2026/159017/hark-sampang-rescues-girl-after-12-days-ai-generated-video-fact-check
2026-09-10T00:00:00
true
fact_checker_verdict
environment
1.0
NF2_20260910_0003
फ्याक्टचेक: हर्क साम्पाङले आक्रोश पोख्दै शेयर गरेको प्रविधिमन्त्री महावीर पुनको भनिएको पोस्ट नक्कली
1
FALSE_MISLEADING
फ्याक्टचेक: हर्क साम्पाङले आक्रोश पोख्दै शेयर गरेको प्रविधिमन्त्री महावीर पुनको भनिएको पोस्ट नक्कली टेकपाना भदौ २५, २०८३ ७:४९ काठमाडौं । प्रतिनिधि सभा सदस्य हर्कराज राई (हर्क सामपाङ) ले १४ लाखभन्दा बढी फलोअर भएको आफ्नो फेसबुक पेजमा बिहीबार आक्रोश पोख्दै एक पोस्ट शेयर गरेका छन् । उनको पोस्टमा विज्ञान, प्रविधि तथा नवप्रव...
https://www.facebook.com/share/p/1Dhup9v5dP/ | https://ghostarchive.org/archive/LPB6k | https://www.youtube.com/shorts/NVMxQypOx2Y | https://www.facebook.com/mahabir.pun
techpana
fact_checker
https://techpana.com/2026/159000/hark-sampang-shared-mahabir-pun-facebook-post-fake
2026-09-10T00:00:00
true
fact_checker_verdict
politics
1.0
NF2_20260909_0004
फ्याक्टचेकः चिलिमे जलविद्युत आयोजनाको सुरुङबाट २४ जनाको सकुशल उद्धार भएको दाबी झुटो
1
FALSE_MISLEADING
फ्याक्टचेकः चिलिमे जलविद्युत आयोजनाको सुरुङबाट २४ जनाको सकुशल उद्धार भएको दाबी झुटो टेकपाना भदौ २४, २०८३ १७:३ काठमाडौं । सामाजिक सञ्जालका विभिन्न प्लेटफर्म, विशेषगरी फेसबुकमा ‘चिलिमे जलविद्युत् आयोजनाको सुरुङबाट २४ जनाको सकुशल उद्धार भएको’ दाबी भाइरल भइरहेको छ । १० भदौमा भोटेकोशीमा आएको आकस्मिक बाढीपछि प्रभावित क्षेत्र...
https://www.facebook.com/share/p/1HDJkMSh7v/ | https://www.facebook.com/profile.php?id=100064682620131&__cft__[0]=AZgzdunacj95qWyfpC_rbF4mLd7w5LiVGk0aHKNUlHckiVEcZG-_udSHnKL6lza6F8iP6tIaob93DbbsiXa46mJukF3scS-RxPULedIW6J0hNI9Lin7BJrmyjnr2dR4N47sw6NU9Gz9legsYCwFya7jr5jXlXIoVGh9aoFbxK00hCUxBgtpEC5IhXyg&__tn__=-UC%2CP-R |...
techpana
fact_checker
https://techpana.com/2026/158993/fact-check-claim-24-people-rescued-from-chilime-hydropower-tunnel-false
2026-09-09T00:00:00
true
fact_checker_verdict
environment
1.0
NF2_20260908_0005
फ्याक्टचेकः सरकार विरुद्ध सडक आन्दोलन भन्दै फैलिएको यो भिडिओ भ्रामक
1
FALSE_MISLEADING
फ्याक्टचेकः सरकार विरुद्ध सडक आन्दोलन भन्दै फैलिएको यो भिडिओ भ्रामक हिक्मत आचार्य भदौ २३, २०८३ २०:१९ काठमाडौं । मङ्गलबार सामाजिक सञ्जाल फेसबुकमा ‘Fact Nepal’ नामको फेसबुक पेजले एक भिडिओ सार्वजनिक गर्‍यो । भिडिओमा नेपालको राष्ट्रिय झन्डा बोकेका युवाहरूले सरकार विरुद्ध नाराबाजी गरेको देखिन्छ । टेकपाना फ्याक्टचेक डेस्कले ...
https://www.facebook.com/share/v/186nitty1B/ | https://www.setopati.com/politics/397897 | https://www.indiatoday.in/india-today-insight/story/why-nepal-legitimacy-crisis-begs-shuffling-beyond-power-chairs-2785221-2025-09-10 | https://www.facebook.com/TechPanaNews/ | https://www.instagram.com/techpananews
techpana
fact_checker
https://techpana.com/2026/158977/factcheck-government-protest-viral-video-misleading
2026-09-08T00:00:00
true
fact_checker_verdict
politics
1.0
NF2_20260908_0006
फ्याक्टचेक: भोटेकोशी बाढीपछि नेपालका हिन्दुहरूले इस्लाम धर्म अँगालेको भन्दै फैलिएको दाबी झुटो
1
FALSE_MISLEADING
फ्याक्टचेक: भोटेकोशी बाढीपछि नेपालका हिन्दुहरूले इस्लाम धर्म अँगालेको भन्दै फैलिएको दाबी झुटो हिक्मत आचार्य भदौ २३, २०८३ १६:६ काठमाडौं । गत १० भदौमा भोटेकोशी नदीमा आएको बाढीपछि नेपालका गाउँ-गाउँमा हिन्दुहरूले धर्म परिवर्तन गरी मुस्लिम बनेको दाबीसहितका भिडिओहरू सामाजिक सञ्जालमा भाइरल भइरहेका छन् । ती भिडिओमा केही व्यक्त...
https://www.facebook.com/reel/38041623495483560/ | https://www.facebook.com/share/r/1CAjb6k9Sp/ | https://www.facebook.com/share/r/19UFJfR2WS/ | https://www.facebook.com/share/r/1QLkB5tum7/ | https://www.facebook.com/share/r/19B9MiUnkP/
techpana
fact_checker
https://techpana.com/2026/158973/viral-video-claiming-nepali-converted-to-islam-is-misleading
2026-09-08T00:00:00
true
fact_checker_verdict
environment
1.0
NF2_20260910_0007
Fact Check: Viral Screenshot Claiming Mahabir Pun Criticized Harka Sampang Is Fake
1
FALSE_MISLEADING
Fact Check: Viral Screenshot Claiming Mahabir Pun Criticized Harka Sampang Is Fake Techpana भदौ २५, २०८३ १४:३८ Kathmandu. Member of the House of Representatives Harka Raj Rai (Harka Sampang) shared a post on his Facebook page on Thursday expressing anger over the government's response to rescue and relief efforts follo...
https://www.facebook.com/permalink.php?story_fbid=pfbid02YbHaVGJ3RHZvXJU6zSeRgBc6c8Lep4Mdzsz94Kp9UxBVYxU1BNzHFSC3vsw75r1Ql&id=100057352183960&rdid=wYPXOQjnS1v6sTc9# | https://ghostarchive.org/archive/LPB6k | https://www.youtube.com/shorts/NVMxQypOx2Y | https://www.facebook.com/mahabir.pun | https://www.facebook.com/Tec...
techpana
fact_checker
https://techpana.com/2026/159011/fact-check-viral-screenshot-claiming-mahabir-pun-criticized-harka-sampang-is-fake
2026-09-10T00:00:00
true
fact_checker_verdict
politics
1.0
NF2_20260908_0008
Fact Check: Video Claiming Nepal’s Hindus Converted to Islam After Bhotekoshi Flood Is Misleading
1
FALSE_MISLEADING
Fact Check: Video Claiming Nepal’s Hindus Converted to Islam After Bhotekoshi Flood Is Misleading Hikmat Acharya भदौ २३, २०८३ २०:३६ Kathmandu. A video circulating on social media claims that people in villages across Nepal began converting from Hinduism to Islam following the Bhotekoshi River flood on August 26. The vi...
https://www.facebook.com/reel/38041623495483560 | https://www.facebook.com/reel/2150632099144897/?rdid=Cx3hcDlPVL2IAvJj&share_url=https%3A%2F%2Fwww.facebook.com%2Fshare%2Fr%2F1CAjb6k9Sp%2F# | https://www.facebook.com/reel/1572514087888297/?rdid=rlqdGkOcZO8KtbcT&share_url=https%3A%2F%2Fwww.facebook.com%2Fshare%2Fr%2F19U...
techpana
fact_checker
https://techpana.com/2026/158978/video-claiming-nepal-hindus-converted-to-islam-after-bhotekoshi-flood-is-misleading
2026-09-08T00:00:00
false
fact_checker_verdict
technology
1.0
NF2_20260906_0009
Fact Check: Old Horse Rescue Video From Venezuela Misleadingly Shared as a Rescue From Recent Bhotekoshi Flood
1
FALSE_MISLEADING
Fact Check: Old Horse Rescue Video From Venezuela Misleadingly Shared as a Rescue From Recent Bhotekoshi Flood Hikmat Acharya भदौ २१, २०८३ १८:५२ Kathmandu. Photos and videos showing a horse being rescued from a flood have been circulating on social media with claims that the animal was rescued in Nepal following the re...
https://www.facebook.com/share/p/1LnoSFrc5t/ | https://archive.is/vlFVH | https://www.facebook.com/groups/2143067189308099/permalink/4525174451097349/?rdid=LS5JkT9BMf9I2PAH# | https://www.facebook.com/yogen.yakha.39/posts/pfbid02u3xisH3dtT6Qta7VBuKeGdYsns58TcmjxYZSBtLz6dSjQYFZX5YfAjMdmRwVm7wHl | https://factcheck.afp.c...
techpana
fact_checker
https://techpana.com/2026/158940/old-horse-rescue-video-from-venezuela-misleadingly-shared-as-rescue-from-a-recent-bhotekoshi-flood
2026-09-06T00:00:00
false
fact_checker_verdict
technology
1.0
NF2_20260906_0010
Fact Check: Old Flood Video Misleadingly Shared as Recent Rasuwa Flood Footage
1
FALSE_MISLEADING
Fact Check: Old Flood Video Misleadingly Shared as Recent Rasuwa Flood Footage Ranjita Upreti भदौ २१, २०८३ १५:५५ Kathmandu. A video showing a man filming a fast-moving flood while standing dangerously close to the rising water has been widely shared on social media with claims that it was recorded during the recent flo...
https://x.com/7sw_8 | https://www.instagram.com/bluejournalism/reels/ | https://www.instagram.com/reels/Dc2klSBsHZ3/ | https://web.facebook.com/reel/2522942808135569?_rdc=1&_rdr# | https://web.facebook.com/iveth.hoyes?_rdc=1&_rdr#
techpana
fact_checker
https://techpana.com/2026/158935/old-flood-video-misleadingly-shared-as-recent-rasuwa-flood-footage
2026-09-06T00:00:00
true
fact_checker_verdict
technology
1.0
NF2_20260906_0011
Fact Check: AI-Edited Photo Falsely Shows KP Oli Wearing a Shoe Garland in Rasuwa
1
FALSE_MISLEADING
Fact Check: AI-Edited Photo Falsely Shows KP Oli Wearing a Shoe Garland in Rasuwa Techpana भदौ २१, २०८३ १३:३६ Kathmandu. A photo showing CPN-UML Chairman and former Prime Minister KP Sharma Oli wearing a garland of shoes has recently been circulating on social media, with claims that flood victims in Rasuwa gave him a ...
https://web.facebook.com/budda.magar.77/posts/pfbid02t2SbEpiTvedUXXum6RvNnX4edrdr9fhbdgWSKcKNK8kjkic1T5YcKJZeVdUZGwpjl?rdid=YNekpBXRvnuaxCB7# | https://web.facebook.com/jungabahadur2/posts/pfbid0aymEk4MpzBaYPwfz8zwtfhwV8ioAveKk3oRXcUhEqyq53QkgRGHpDPtZGfnUXBpFl?rdid=gwigJ5rcDHJ2C4s2# | https://facebook.com/share/v/1EEjc...
techpana
fact_checker
https://techpana.com/2026/158931/fact-check-ai-edited-photo-falsely-shows-kp-oli-wearing-a-shoe-garland-in-rasuwa
2026-09-06T00:00:00
true
fact_checker_verdict
politics
1.0
NF2_20260901_0012
Fact Check: Viral Video of Armed Police Rescuing Girl in Rasuwa Is AI-Generated
1
FALSE_MISLEADING
Fact Check: Viral Video of Armed Police Rescuing Girl in Rasuwa Is AI-Generated Hikmat Acharya भदौ १६, २०८३ १७:७ Kathmandu. A video showing armed police personnel rescuing a girl has recently gone viral on social media, with claims that it shows the rescue of a girl who was swept away by floods and landslides in Timure...
https://facebook.com/reel/1827317935102656 | https://www.facebook.com/reel/2027472654620129 | https://www.facebook.com/watch/?ref=saved&v=2124049011515974 | https://www.facebook.com/reel/896551733257320 | https://www.facebook.com/reel/2054718815170683
techpana
fact_checker
https://techpana.com/2026/158846/fact-check-viral-video-of-armed-police-rescuing-girl-in-rasuwa-is-ai-generated
2026-09-01T00:00:00
true
fact_checker_verdict
environment
1.0
NF2_20260911_0013
च्याटजीपीटीमा सकेसम्म अपलोड नगर्नुहोस् यस्ता फाइल तथा डकुमेन्ट
1
FALSE_MISLEADING
16 Shares च्याटजीपीटीमा सकेसम्म अपलोड नगर्नुहोस् यस्ता फाइल तथा डकुमेन्ट टेकपाना भदौ २६, २०८३ ११:४० काठमाडौं । इमेलको ड्राफ्ट तयार पार्न, पीडीएफको सारांश निकाल्न वा सम्झौता पत्रहरूको समीक्षा गर्नका लागि च्याटजीपीटी मानिसहरूको दैनिक सहयोगी बनेको छ । तर यो सुविधासँगै एउटा ठुलो जोखिम पनि जोडिएको छ- तपाईंले च्याट विन्डोमा ...
techpana
fact_checker
https://techpana.com/2026/159023/files-you-should-avoid-uploading-to-chatgpt
2026-09-11T00:00:00
true
fact_checker_verdict
technology
1.0
NF2_20260911_0014
नेपाल भित्रियो १०,००० एमएएचसम्मको ब्याट्री भएको रेड्मी नोट १७ सिरिज, अब चार्जर बोक्नै नपर्ने !
1
FALSE_MISLEADING
नेपाल भित्रियो १०,००० एमएएचसम्मको ब्याट्री भएको रेड्मी नोट १७ सिरिज, अब चार्जर बोक्नै नपर्ने ! टेकपाना भदौ २६, २०८३ १४:२५ काठमाडौं । चिनियाँ स्मार्टफोन निर्माता कम्पनी साओमीले उच्च क्षमताको ब्याट्री र शक्तिशाली कार्यक्षमतासहितको नयाँ रेड्मी नोट १७ सिरिज नेपाली बजारमा सार्वजनिक गरेको छ । कम्पनीले नयाँ सिरिजअन्तर्गत रेड्...
techpana
fact_checker
https://techpana.com/2026/159027/redmi-note-17-series-nepal-10000mah-battery-charger
2026-09-11T00:00:00
true
fact_checker_verdict
other
1.0
NF2_20260911_0015
गुड्दागुड्दैको ईभीमा आगलागी भएर मृत्यु हुनेमा नारायणी विद्या मन्दिरका पूर्व सञ्चालक अमर पराजुली
1
FALSE_MISLEADING
गुड्दागुड्दैको ईभीमा आगलागी भएर मृत्यु हुनेमा नारायणी विद्या मन्दिरका पूर्व सञ्चालक अमर पराजुली टेकपाना भदौ २६, २०८३ १३:२२ काठमाडौं । शुक्रबार चितवनको भरतपुरमा गुड्दै गरेको जीएसी आयोन वाई मोडेलको विद्युतीय गाडी (ईभी) मा आगलागी भई मृत्यु हुने व्यक्तिको पहिचान खुलेको छ । मृतक चितवनको भरतपुर महानगरपालिका-११ का ५५ वर्षीय अ...
techpana
fact_checker
https://techpana.com/2026/159025/ev-fire-death-narayani-vidya-mandir-former-owner-amar-parajuli
2026-09-11T00:00:00
true
fact_checker_verdict
other
1.0
End of preview.

NepFakeV2 🇳🇵

The first real, non-synthetic Nepali fact-checking dataset. Built directly from verified fact-checks published by professional Nepali fact-checkers — not machine-translated, not LLM-generated.

Why This Dataset Exists

All existing Nepali fake news datasets fail in one of two ways:

  • Machine-translated US political content — introduces topic-domain confounds and MT artifacts. Models learn to detect translation, not misinformation.
  • LLM-generated examples — do not reflect real misinformation that actually circulated in Nepal.

NepFakeV2 breaks this cycle by sourcing directly from two active IFCN-adjacent Nepali fact-checkers. It grows automatically — new fact-checks are scraped, normalized, and pushed to this repository every week.

Load the Dataset 

from datasets import load_dataset

ds = load_dataset("Nandan007/NepFakeV2")
print(ds["train"][0])

Or load directly as a DataFrame:

import pandas as pd

df = pd.read_csv(
    "https://huggingface.co/datasets/Nandan007/NepFakeV2/resolve/main/data/nepfakev2.csv"
)

Dataset Summary

Language Nepali (Devanagari script)
Task Fact verification / misinformation detection
Labels 3 (REAL, FALSE_MISLEADING, UNVERIFIED)
Sources 2 (TechPana, NepalFactCheck)
Updates Weekly — auto-synced from GitHub
Schema version 1.0

Labels

Label Int Description
REAL 0 Verified factual — confirmed by fact-checkers
FALSE_MISLEADING 1 Debunked — false or misleading content
UNVERIFIED 2 Cannot be confirmed or denied

On label imbalance: 90%+ of examples are FALSE_MISLEADING. This reflects real-world distribution — fact-checkers publish mostly debunks by design. This is not a collection error.

Data Sources

Source Type Language Coverage
TechPana IFCN-certified fact-checker Nepali Feb 2025 → present
NepalFactCheck Non-profit fact-checker (CMR-Nepal) Nepali Mar 2020 → present

Data Fields

Field Type Description
example_id string Unique ID — NF2_YYYYMMDD_NNNN
claim_text string Article headline (the claim)
verdict_label int 0 / 1 / 2
verdict_label_text string REAL / FALSE_MISLEADING / UNVERIFIED
evidence_text string Full fact-check article body
source_name string techpana / nepalfactcheck
source_url string Original article URL
date_published string ISO date YYYY-MM-DD
is_native_nepali bool True if Devanagari script detected
topic_category string politics / health / technology / society / environment / economy / other
annotator_notes string Edge case flags — empty if clean
schema_version string 1.0

Dataset Statistics

See stats.json for live statistics updated after every pipeline run.

Limitations

  • Label imbalance — fact-checkers publish mostly debunks by design. 90%+ FALSE_MISLEADING reflects real-world distribution, not a collection error.
  • claim_text is the article headline — not an atomically extracted claim. Future versions will improve atomic claim extraction.
  • Cross-source duplicates — the same claim may be fact-checked by both sources independently. Both records are retained. Cross-source deduplication requires Nepali NER and is documented as future work.

Citation 

@dataset{singh2026nepfakev2,
  author    = {Singh, Nandan},
  title     = {NepFakeV2: A Live Benchmark Dataset for Nepali Misinformation Detection},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/Nandan007/NepFakeV2},
  note      = {Schema v1.0. Auto-updated weekly.}
}

License

Links

Downloads last month
88
Total file size:
29.5 MB