image
imagewidth (px)
1
799
label
stringlengths
1
23
incommensurately
Isotropic
NITRATES
Locales
neodymium
proposals
BOUILLON
GASEOUS
Subjective
roominess
Secretariats
LOADERS
Brownsville
COMPORT
Hexagram
restocking
BUNGEES
immortality
Sparks
SHIV
Turbocharged
RETRIEVER
GRISTMILL
credentialing
Summarizes
spatter
Insinuator
scrawl
confederation
BOMBASTIC
resign
grantsmanship
expostulates
spoilsport
MONSOONAL
Continually
semiconscious
BACKSTROKES
OMEGA
COSTA
BLUESIER
Workingwoman
GYPSY
Cries
DIVINER
Busting
Plucky
FUTURISM
barbered
DEFIANT
gloucester
undemanding
energizing
NAN
mouthy
CORINNE
DIBBLE
NOTARIZES
MICHELIN
laminate
Eviscerating
tablecloth
diktat
Coalescing
DEMARCATES
Timeserving
COOTS
annalists
Alphonse
spastics
Quicksands
Conflation
Sizzling
OOZIEST
LIMPED
Cavorting
IMPROBABILITY
sedna
GENE
Goiania
POLONIUM
BECKON
PRESENTMENTS
EVERLASTING
Certifiable
G
SUSTAINED
NYPD
Windsurf
foreseeable
TALKING
BLAGGED
Rewound
twinkle
OCTAL
Decompose
COMPERED
misuse
Ribbons
unenlightened

Dataset Card for "Synth-Text Recognition"

This is the dataset for text recognition on document images, synthetically generated, covering 90K English words. It includes training, validation and test splits.

Downloads last month
17
Edit dataset card

Models trained or fine-tuned on redactable-llm/synth-text-recognition