Dataset Viewer
Auto-converted to Parquet Duplicate
category
stringlengths
5
21
input
stringlengths
4
67
expected
stringlengths
0
51
keep_like_verb
I like pizza.
I like pizza.
keep_like_verb
Do you like jazz?
Do you like jazz?
keep_like_verb
I'd like a table for two.
I'd like a table for two.
keep_like_verb
She really likes him.
She really likes him.
keep_like_prep
It looks like rain.
It looks like rain.
keep_like_prep
He runs like the wind.
He runs like the wind.
keep_like_prep
What was the concert like?
What was the concert like?
keep_like_prep
It tastes like chicken.
It tastes like chicken.
remove_like_filler
I was, like, gonna call you.
I was gonna call you.
remove_like_filler
It's like the best restaurant in town.
It's the best restaurant in town.
remove_like_filler
He was like an hour late.
He was like an hour late.
mixed_like
I, um, like pizza.
I like pizza.
mixed_like
I hope you will like, you know, it.
I hope you will like it.
keep_you_know
You know the answer.
You know the answer.
keep_you_know
Do you know the way to the station?
Do you know the way to the station?
keep_you_know
You know him better than I do.
You know him better than I do.
remove_you_know
It was, you know, pretty awkward.
It was pretty awkward.
remove_you_know
You know, I never liked that place.
I never liked that place.
keep_i_mean
I mean it.
I mean it.
keep_i_mean
That's not what I mean.
That's not what I mean.
remove_i_mean
I mean, we could always drive there.
We could always drive there.
keep_kind_of
What kind of music do you like?
What kind of music do you like?
keep_kind_of
It's a kind of fruit.
It's a kind of fruit.
keep_kind_of
That's very kind of you.
That's very kind of you.
remove_kind_of
It's kind of cold in here.
It's cold in here.
remove_kind_of
I was sort of hoping you'd come.
I was hoping you'd come.
remove_kind_of
I kind of want to leave.
I want to leave.
um_uh_basic
Um, I think we should go.
I think we should go.
um_uh_basic
The meeting is, uh, at three.
The meeting is at three.
um_uh_basic
Er, could you repeat that?
Could you repeat that?
um_uh_basic
So, um, what do we do now?
So, what do we do now?
repetition
I I think that's right.
I think that's right.
repetition
We should we should leave early.
We should leave early.
repetition
Thanks for, for the invite.
Thanks for the invite.
false_start
I was— we were already there.
We were already there.
false_start
Can you... I mean, could you help me?
Could you help me?
stutter
I took the b- the bus.
I took the bus.
stutter
That's a gr- great idea.
That's a great idea.
numbers_preserved
Um, the API returned a 404 error.
The API returned a 404 error.
numbers_preserved
It costs, like, $19.99 with tax.
It costs $19.99 with tax.
numbers_preserved
Uh, call me at 555-0123.
Call me at 555-0123.
numbers_preserved
The flight is, um, UA 1187 out of Denver.
The flight is UA 1187 out of Denver.
clean_passthrough
The quarterly report is due on Friday.
The quarterly report is due on Friday.
clean_passthrough
She's kind to everyone she meets.
She's kind to everyone she meets.
clean_passthrough
He took it literally.
He took it literally.
clean_passthrough
Could you say that again more slowly?
Could you say that again more slowly?
remove_adverb_filler
I literally just walked in the door.
I just walked in the door.
remove_adverb_filler
Basically, everyone agreed with the plan.
Everyone agreed with the plan.
all_filler
Um, uh...
all_filler
Hmm.
lowercase_stt
um i think we should uh go now
i think we should go now
lowercase_stt
i like pizza
i like pizza
lowercase_stt
it was you know pretty far
it was pretty far
multi
Um, so I was, like, thinking maybe we could, uh, try the new place?
So I was thinking maybe we could try the new place?
repair_replacement
We should we should take the b- the bus, I mean, the train.
We should take the train.
repair_replacement
Meet me on Tuesday, I mean, Wednesday.
Meet me on Wednesday.
q_substitution
Take a left, I mean, a right at the light.
Take a right at the light.
q_substitution
Paint it red, uh, blue.
Paint it blue.
q_substitution
Ask my brother, I mean, my sister.
Ask my sister.
q_substitution
It's on the third, um, the fourth floor.
It's on the fourth floor.
q_substitution_cutoff
Turn le- turn right at the light.
Turn right at the light.
q_doubled
She was, um, um, really tired.
She was really tired.
q_doubled
I was like, like, so tired.
I was so tired.
q_collision
Do you, like, like jazz?
Do you like jazz?
q_collision
I know, you know, that it's late.
I know that it's late.

fdt-disfluency-synthetic

Synthetic disfluency-removal data: clean conversational text (DailyDialog + Synthetic-Persona-Chat) with disfluencies injected programmatically, giving exact word-level labels (KEEP / DELETE / KEEP_STRIP_COMMA / KEEP_CAPITALIZE). Invariant: applying the tags reconstructs the clean text exactly. Rows: {{"words": [...], "tags": [...], "clean": "..."}}.

  • v1/ — fillers (um, uh, er, ah, like, you know, I mean, sort of, kind of, basically, literally), repetitions, false starts, stutters/cut-offs, lowercase-STT variants, adversarial keep-templates.
  • v2/ — v1 plus substitution repairs ("the bus, I mean, the train"), filler-homograph collisions ("Do you, like, like jazz?"), doubled fillers, denser disfluency stacks.
  • eval/adversarial.jsonl — 65 handwritten cases incl. out-of-pool quarantine classes for measuring schema transfer.

License note: DailyDialog is CC BY-NC-SA, so this dataset inherits non-commercial terms. Generator code lives in the FluencyAI repo (ml/disfluency/disfl/injector.py).

Downloads last month
68