gsaltintas commited on
Commit
56efb74
·
verified ·
1 Parent(s): a255a70

Uploading tokenizer_robustness_completion_turkish_word_reordering subset

Browse files
README.md CHANGED
@@ -64,6 +64,10 @@ configs:
64
  data_files:
65
  - split: test
66
  path: tokenizer_robustness_completion_turkish_web_search_query/test-*
 
 
 
 
67
  dataset_info:
68
  - config_name: tokenizer_robustness_completion_turkish_cannonical
69
  features:
@@ -1801,6 +1805,130 @@ dataset_info:
1801
  num_examples: 36
1802
  download_size: 38921
1803
  dataset_size: 18772
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1804
  ---
1805
 
1806
  # Dataset Card for Tokenization Robustness
 
64
  data_files:
65
  - split: test
66
  path: tokenizer_robustness_completion_turkish_web_search_query/test-*
67
+ - config_name: tokenizer_robustness_completion_turkish_word_reordering
68
+ data_files:
69
+ - split: test
70
+ path: tokenizer_robustness_completion_turkish_word_reordering/test-*
71
  dataset_info:
72
  - config_name: tokenizer_robustness_completion_turkish_cannonical
73
  features:
 
1805
  num_examples: 36
1806
  download_size: 38921
1807
  dataset_size: 18772
1808
+ - config_name: tokenizer_robustness_completion_turkish_word_reordering
1809
+ features:
1810
+ - name: question
1811
+ dtype: string
1812
+ - name: choices
1813
+ list: string
1814
+ - name: answer
1815
+ dtype: int64
1816
+ - name: answer_label
1817
+ dtype: string
1818
+ - name: split
1819
+ dtype: string
1820
+ - name: subcategories
1821
+ dtype: string
1822
+ - name: categories
1823
+ dtype: string
1824
+ - name: lang
1825
+ dtype: string
1826
+ - name: second_lang
1827
+ dtype: string
1828
+ - name: notes
1829
+ dtype: string
1830
+ - name: id
1831
+ dtype: string
1832
+ - name: set_id
1833
+ dtype: float64
1834
+ - name: variation_id
1835
+ dtype: float64
1836
+ - name: vanilla_cos_sim_to_canonical
1837
+ struct:
1838
+ - name: CohereLabs/aya-expanse-8b
1839
+ dtype: float64
1840
+ - name: Qwen/Qwen3-8B
1841
+ dtype: float64
1842
+ - name: bigscience/bloom
1843
+ dtype: float64
1844
+ - name: common-pile/comma-v0.1-1t
1845
+ dtype: float64
1846
+ - name: facebook/xglm-564M
1847
+ dtype: float64
1848
+ - name: google-bert/bert-base-multilingual-cased
1849
+ dtype: float64
1850
+ - name: google/byt5-small
1851
+ dtype: float64
1852
+ - name: google/gemma-2-2b
1853
+ dtype: float64
1854
+ - name: gpt2
1855
+ dtype: float64
1856
+ - name: meta-llama/Llama-3.2-1B
1857
+ dtype: float64
1858
+ - name: microsoft/Phi-3-mini-4k-instruct
1859
+ dtype: float64
1860
+ - name: mistralai/tekken
1861
+ dtype: float64
1862
+ - name: tiktoken/gpt-4o
1863
+ dtype: float64
1864
+ - name: tokenmonster/englishcode-32000-consistent-v1
1865
+ dtype: float64
1866
+ - name: trimmed_cos_sim_to_canonical
1867
+ struct:
1868
+ - name: CohereLabs/aya-expanse-8b
1869
+ dtype: float64
1870
+ - name: Qwen/Qwen3-8B
1871
+ dtype: float64
1872
+ - name: bigscience/bloom
1873
+ dtype: float64
1874
+ - name: common-pile/comma-v0.1-1t
1875
+ dtype: float64
1876
+ - name: facebook/xglm-564M
1877
+ dtype: float64
1878
+ - name: google-bert/bert-base-multilingual-cased
1879
+ dtype: float64
1880
+ - name: google/byt5-small
1881
+ dtype: float64
1882
+ - name: google/gemma-2-2b
1883
+ dtype: float64
1884
+ - name: gpt2
1885
+ dtype: float64
1886
+ - name: meta-llama/Llama-3.2-1B
1887
+ dtype: float64
1888
+ - name: microsoft/Phi-3-mini-4k-instruct
1889
+ dtype: float64
1890
+ - name: mistralai/tekken
1891
+ dtype: float64
1892
+ - name: tiktoken/gpt-4o
1893
+ dtype: float64
1894
+ - name: tokenmonster/englishcode-32000-consistent-v1
1895
+ dtype: float64
1896
+ - name: token_counts
1897
+ struct:
1898
+ - name: CohereLabs/aya-expanse-8b
1899
+ dtype: int64
1900
+ - name: Qwen/Qwen3-8B
1901
+ dtype: int64
1902
+ - name: bigscience/bloom
1903
+ dtype: int64
1904
+ - name: common-pile/comma-v0.1-1t
1905
+ dtype: int64
1906
+ - name: facebook/xglm-564M
1907
+ dtype: int64
1908
+ - name: google-bert/bert-base-multilingual-cased
1909
+ dtype: int64
1910
+ - name: google/byt5-small
1911
+ dtype: int64
1912
+ - name: google/gemma-2-2b
1913
+ dtype: int64
1914
+ - name: gpt2
1915
+ dtype: int64
1916
+ - name: meta-llama/Llama-3.2-1B
1917
+ dtype: int64
1918
+ - name: microsoft/Phi-3-mini-4k-instruct
1919
+ dtype: int64
1920
+ - name: mistralai/tekken
1921
+ dtype: int64
1922
+ - name: tiktoken/gpt-4o
1923
+ dtype: int64
1924
+ - name: tokenmonster/englishcode-32000-consistent-v1
1925
+ dtype: int64
1926
+ splits:
1927
+ - name: test
1928
+ num_bytes: 23335
1929
+ num_examples: 44
1930
+ download_size: 41116
1931
+ dataset_size: 23335
1932
  ---
1933
 
1934
  # Dataset Card for Tokenization Robustness
tokenizer_robustness_completion_turkish_word_reordering/test-00000-of-00001.parquet CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0709a08cdf122f83e9726111e165a3eb4410c16e98e159fa765993ad866ffb24
3
- size 9232
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f4874bcd443fac79427928f15da968399bd41079e605205b0d87fe787e48a9a
3
+ size 41116