gsaltintas commited on
Commit
d9e9232
·
verified ·
1 Parent(s): 2d2d1c8

Uploading tokenizer_robustness_completion_english_web_search_query subset

Browse files
README.md CHANGED
@@ -3356,6 +3356,130 @@ dataset_info:
3356
  num_examples: 40
3357
  download_size: 41912
3358
  dataset_size: 25705
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3359
  configs:
3360
  - config_name: tokenizer_robustness_completion_english_abbreviations
3361
  data_files:
@@ -3465,6 +3589,10 @@ configs:
3465
  data_files:
3466
  - split: test
3467
  path: tokenizer_robustness_completion_english_superscript_subscript_styling/test-*
 
 
 
 
3468
  ---
3469
 
3470
  # Dataset Card for Tokenization Robustness
 
3356
  num_examples: 40
3357
  download_size: 41912
3358
  dataset_size: 25705
3359
+ - config_name: tokenizer_robustness_completion_english_web_search_query
3360
+ features:
3361
+ - name: question
3362
+ dtype: string
3363
+ - name: choices
3364
+ list: string
3365
+ - name: answer
3366
+ dtype: int64
3367
+ - name: answer_label
3368
+ dtype: string
3369
+ - name: split
3370
+ dtype: string
3371
+ - name: subcategories
3372
+ dtype: string
3373
+ - name: category
3374
+ dtype: string
3375
+ - name: lang
3376
+ dtype: string
3377
+ - name: second_lang
3378
+ dtype: string
3379
+ - name: notes
3380
+ dtype: string
3381
+ - name: id
3382
+ dtype: string
3383
+ - name: set_id
3384
+ dtype: string
3385
+ - name: variation_id
3386
+ dtype: string
3387
+ - name: vanilla_cos_sim_to_canonical
3388
+ struct:
3389
+ - name: CohereLabs/aya-expanse-8b
3390
+ dtype: float64
3391
+ - name: Qwen/Qwen3-8B
3392
+ dtype: float64
3393
+ - name: bigscience/bloom
3394
+ dtype: float64
3395
+ - name: common-pile/comma-v0.1-1t
3396
+ dtype: float64
3397
+ - name: facebook/xglm-564M
3398
+ dtype: float64
3399
+ - name: google-bert/bert-base-multilingual-cased
3400
+ dtype: float64
3401
+ - name: google/byt5-small
3402
+ dtype: float64
3403
+ - name: google/gemma-2-2b
3404
+ dtype: float64
3405
+ - name: gpt2
3406
+ dtype: float64
3407
+ - name: meta-llama/Llama-3.2-1B
3408
+ dtype: float64
3409
+ - name: microsoft/Phi-3-mini-4k-instruct
3410
+ dtype: float64
3411
+ - name: mistralai/tekken
3412
+ dtype: float64
3413
+ - name: tiktoken/gpt-4o
3414
+ dtype: float64
3415
+ - name: tokenmonster/englishcode-32000-consistent-v1
3416
+ dtype: float64
3417
+ - name: trimmed_cos_sim_to_canonical
3418
+ struct:
3419
+ - name: CohereLabs/aya-expanse-8b
3420
+ dtype: float64
3421
+ - name: Qwen/Qwen3-8B
3422
+ dtype: float64
3423
+ - name: bigscience/bloom
3424
+ dtype: float64
3425
+ - name: common-pile/comma-v0.1-1t
3426
+ dtype: float64
3427
+ - name: facebook/xglm-564M
3428
+ dtype: float64
3429
+ - name: google-bert/bert-base-multilingual-cased
3430
+ dtype: float64
3431
+ - name: google/byt5-small
3432
+ dtype: float64
3433
+ - name: google/gemma-2-2b
3434
+ dtype: float64
3435
+ - name: gpt2
3436
+ dtype: float64
3437
+ - name: meta-llama/Llama-3.2-1B
3438
+ dtype: float64
3439
+ - name: microsoft/Phi-3-mini-4k-instruct
3440
+ dtype: float64
3441
+ - name: mistralai/tekken
3442
+ dtype: float64
3443
+ - name: tiktoken/gpt-4o
3444
+ dtype: float64
3445
+ - name: tokenmonster/englishcode-32000-consistent-v1
3446
+ dtype: float64
3447
+ - name: token_counts
3448
+ struct:
3449
+ - name: CohereLabs/aya-expanse-8b
3450
+ dtype: int64
3451
+ - name: Qwen/Qwen3-8B
3452
+ dtype: int64
3453
+ - name: bigscience/bloom
3454
+ dtype: int64
3455
+ - name: common-pile/comma-v0.1-1t
3456
+ dtype: int64
3457
+ - name: facebook/xglm-564M
3458
+ dtype: int64
3459
+ - name: google-bert/bert-base-multilingual-cased
3460
+ dtype: int64
3461
+ - name: google/byt5-small
3462
+ dtype: int64
3463
+ - name: google/gemma-2-2b
3464
+ dtype: int64
3465
+ - name: gpt2
3466
+ dtype: int64
3467
+ - name: meta-llama/Llama-3.2-1B
3468
+ dtype: int64
3469
+ - name: microsoft/Phi-3-mini-4k-instruct
3470
+ dtype: int64
3471
+ - name: mistralai/tekken
3472
+ dtype: int64
3473
+ - name: tiktoken/gpt-4o
3474
+ dtype: int64
3475
+ - name: tokenmonster/englishcode-32000-consistent-v1
3476
+ dtype: int64
3477
+ splits:
3478
+ - name: test
3479
+ num_bytes: 18991
3480
+ num_examples: 37
3481
+ download_size: 38294
3482
+ dataset_size: 18991
3483
  configs:
3484
  - config_name: tokenizer_robustness_completion_english_abbreviations
3485
  data_files:
 
3589
  data_files:
3590
  - split: test
3591
  path: tokenizer_robustness_completion_english_superscript_subscript_styling/test-*
3592
+ - config_name: tokenizer_robustness_completion_english_web_search_query
3593
+ data_files:
3594
+ - split: test
3595
+ path: tokenizer_robustness_completion_english_web_search_query/test-*
3596
  ---
3597
 
3598
  # Dataset Card for Tokenization Robustness
tokenizer_robustness_completion_english_web_search_query/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a90ccc6edbe620c262dc86aa28571fc6a1ca29179e6162138510db63d9ff48b9
3
+ size 38294