gsaltintas commited on
Commit
bb04787
·
verified ·
1 Parent(s): 7a94779

Uploading tokenizer_robustness_completion_english_spaced_styling subset

Browse files
README.md CHANGED
@@ -2984,6 +2984,130 @@ dataset_info:
2984
  num_examples: 40
2985
  download_size: 39744
2986
  dataset_size: 20441
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2987
  configs:
2988
  - config_name: tokenizer_robustness_completion_english_abbreviations
2989
  data_files:
@@ -3081,6 +3205,10 @@ configs:
3081
  data_files:
3082
  - split: test
3083
  path: tokenizer_robustness_completion_english_space_removal/test-*
 
 
 
 
3084
  ---
3085
 
3086
  # Dataset Card for Tokenization Robustness
 
2984
  num_examples: 40
2985
  download_size: 39744
2986
  dataset_size: 20441
2987
+ - config_name: tokenizer_robustness_completion_english_spaced_styling
2988
+ features:
2989
+ - name: question
2990
+ dtype: string
2991
+ - name: choices
2992
+ list: string
2993
+ - name: answer
2994
+ dtype: int64
2995
+ - name: answer_label
2996
+ dtype: string
2997
+ - name: split
2998
+ dtype: string
2999
+ - name: subcategories
3000
+ dtype: string
3001
+ - name: category
3002
+ dtype: string
3003
+ - name: lang
3004
+ dtype: string
3005
+ - name: second_lang
3006
+ dtype: string
3007
+ - name: notes
3008
+ dtype: string
3009
+ - name: id
3010
+ dtype: string
3011
+ - name: set_id
3012
+ dtype: string
3013
+ - name: variation_id
3014
+ dtype: string
3015
+ - name: vanilla_cos_sim_to_canonical
3016
+ struct:
3017
+ - name: CohereLabs/aya-expanse-8b
3018
+ dtype: float64
3019
+ - name: Qwen/Qwen3-8B
3020
+ dtype: float64
3021
+ - name: bigscience/bloom
3022
+ dtype: float64
3023
+ - name: common-pile/comma-v0.1-1t
3024
+ dtype: float64
3025
+ - name: facebook/xglm-564M
3026
+ dtype: float64
3027
+ - name: google-bert/bert-base-multilingual-cased
3028
+ dtype: float64
3029
+ - name: google/byt5-small
3030
+ dtype: float64
3031
+ - name: google/gemma-2-2b
3032
+ dtype: float64
3033
+ - name: gpt2
3034
+ dtype: float64
3035
+ - name: meta-llama/Llama-3.2-1B
3036
+ dtype: float64
3037
+ - name: microsoft/Phi-3-mini-4k-instruct
3038
+ dtype: float64
3039
+ - name: mistralai/tekken
3040
+ dtype: float64
3041
+ - name: tiktoken/gpt-4o
3042
+ dtype: float64
3043
+ - name: tokenmonster/englishcode-32000-consistent-v1
3044
+ dtype: float64
3045
+ - name: trimmed_cos_sim_to_canonical
3046
+ struct:
3047
+ - name: CohereLabs/aya-expanse-8b
3048
+ dtype: float64
3049
+ - name: Qwen/Qwen3-8B
3050
+ dtype: float64
3051
+ - name: bigscience/bloom
3052
+ dtype: float64
3053
+ - name: common-pile/comma-v0.1-1t
3054
+ dtype: float64
3055
+ - name: facebook/xglm-564M
3056
+ dtype: float64
3057
+ - name: google-bert/bert-base-multilingual-cased
3058
+ dtype: float64
3059
+ - name: google/byt5-small
3060
+ dtype: float64
3061
+ - name: google/gemma-2-2b
3062
+ dtype: float64
3063
+ - name: gpt2
3064
+ dtype: float64
3065
+ - name: meta-llama/Llama-3.2-1B
3066
+ dtype: float64
3067
+ - name: microsoft/Phi-3-mini-4k-instruct
3068
+ dtype: float64
3069
+ - name: mistralai/tekken
3070
+ dtype: float64
3071
+ - name: tiktoken/gpt-4o
3072
+ dtype: float64
3073
+ - name: tokenmonster/englishcode-32000-consistent-v1
3074
+ dtype: float64
3075
+ - name: token_counts
3076
+ struct:
3077
+ - name: CohereLabs/aya-expanse-8b
3078
+ dtype: int64
3079
+ - name: Qwen/Qwen3-8B
3080
+ dtype: int64
3081
+ - name: bigscience/bloom
3082
+ dtype: int64
3083
+ - name: common-pile/comma-v0.1-1t
3084
+ dtype: int64
3085
+ - name: facebook/xglm-564M
3086
+ dtype: int64
3087
+ - name: google-bert/bert-base-multilingual-cased
3088
+ dtype: int64
3089
+ - name: google/byt5-small
3090
+ dtype: int64
3091
+ - name: google/gemma-2-2b
3092
+ dtype: int64
3093
+ - name: gpt2
3094
+ dtype: int64
3095
+ - name: meta-llama/Llama-3.2-1B
3096
+ dtype: int64
3097
+ - name: microsoft/Phi-3-mini-4k-instruct
3098
+ dtype: int64
3099
+ - name: mistralai/tekken
3100
+ dtype: int64
3101
+ - name: tiktoken/gpt-4o
3102
+ dtype: int64
3103
+ - name: tokenmonster/englishcode-32000-consistent-v1
3104
+ dtype: int64
3105
+ splits:
3106
+ - name: test
3107
+ num_bytes: 23738
3108
+ num_examples: 40
3109
+ download_size: 41512
3110
+ dataset_size: 23738
3111
  configs:
3112
  - config_name: tokenizer_robustness_completion_english_abbreviations
3113
  data_files:
 
3205
  data_files:
3206
  - split: test
3207
  path: tokenizer_robustness_completion_english_space_removal/test-*
3208
+ - config_name: tokenizer_robustness_completion_english_spaced_styling
3209
+ data_files:
3210
+ - split: test
3211
+ path: tokenizer_robustness_completion_english_spaced_styling/test-*
3212
  ---
3213
 
3214
  # Dataset Card for Tokenization Robustness
tokenizer_robustness_completion_english_spaced_styling/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7bc68fb09e0366a911363712bd31eddb752aff8d152818a95029044be167f8c
3
+ size 41512