id stringlengths 14 16 | messages listlengths 3 3 | control dict | applied listlengths 0 6 | clean_text stringlengths 30 192 | target_text stringlengths 32 237 | source_split stringclasses 1
value | source_index int64 0 4.5k |
|---|---|---|---|---|---|---|---|
train-00000-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers"
],
"non_speech_events": false
} | [
"fillers"
] | Yeah, I do. Yes, I don't work, though, I used to work when I had two children. | Yeah, I do. Yes, {F uh, } I don't work, though, I used to work when I had two children. | train | 0 |
train-00000-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | Yeah, I do. Yes, I don't work, though, I used to work when I had two children. | Yeah, I do <laughter>. Yes, I don't work, though, I used to work when I had two children. | train | 0 |
train-00000-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | Yeah, I do. Yes, I don't work, though, I used to work when I had two children. | Yeah, I do <laughter>. Yes, I don't work, though, I used to work when I had two children. | train | 0 |
train-00000-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | Yeah, I do. Yes, I don't work, though, I used to work when I had two children. | Yeah, I do <laughter>. Yes, I don't work, though, I used to work when I had two children. | train | 0 |
train-00001-none | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "none",
"phenomena": [],
"non_speech_events": false
} | [] | I think that's an interesting policy your company has about testing immediately after an on the job accident. | I think that's an interesting policy your company has about testing immediately after an on the job accident. | train | 1 |
train-00002-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers",
"fillers",
"repetitions"
] | Exactly. I see more men, participating in the family things with children. | Exactly. [ I, + I ] see more men, {F uh, } {D like } participating in the family things with children. | train | 2 |
train-00002-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"repairs"
],
"non_speech_events": true
} | [
"fillers"
] | Exactly. I see more men, participating in the family things with children. | Exactly. I see more men, {F uh, } participating in the family things with children. | train | 2 |
train-00002-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers",
"fillers"
] | Exactly. I see more men, participating in the family things with children. | Exactly. I see more men, {F uh, } {D like } participating in {D like } the family things with children. | train | 2 |
train-00002-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"editing terms",
"repairs"
],
"non_speech_events": false
} | [
"fillers"
] | Exactly. I see more men, participating in the family things with children. | Exactly. I see more men, {F uh, } participating in the family things with children. | train | 2 |
train-00003-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers"
] | helping to take care of them more. and, doing the jobs equally, What about politics on it? | {D you know, } helping to take care of them more. and, doing the jobs equally, What about politics on it? | train | 3 |
train-00003-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"repetitions"
],
"non_speech_events": false
} | [
"discourse markers",
"repetitions"
] | helping to take care of them more. and, doing the jobs equally, What about politics on it? | {D Like, } {D you know, } helping to take care of them more. [ And + and, ] doing the jobs equally, {D you know, } What about politics on it? | train | 3 |
train-00003-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers",
"fillers",
"non-speech events"
] | helping to take care of them more. and, doing the jobs equally, What about politics on it? | {D Like, } helping to take care of them more. -- and, {F uh, } {F uh, } doing the jobs equally, {D you know, } What about politics on it? | train | 3 |
train-00003-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"repetitions"
],
"non_speech_events": true
} | [
"fillers",
"non-speech events"
] | helping to take care of them more. and, doing the jobs equally, What about politics on it? | helping to take care of them more. -- and, {F uh, } doing the jobs equally, What about politics on it? | train | 3 |
train-00004-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | I'm not that big on politics. I'm not that educated on it. I'm, I really think we should have more I agree with that. | I'm <laughter> not that big on politics. I'm not <laughter> -- -- that educated on it. I'm, - I really think we should have more -- I agree with that. | train | 4 |
train-00004-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | I'm not that big on politics. I'm not that educated on it. I'm, I really think we should have more I agree with that. | I'm <laughter> not that big on politics. I'm not <laughter> -- -- that educated on it. I'm, - I really think we should have more -- I agree with that. | train | 4 |
train-00004-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | I'm not that big on politics. I'm not that educated on it. I'm, I really think we should have more I agree with that. | I'm <laughter> not that big on politics. I'm not <laughter> -- -- that educated on it. I'm, - I really think we should have more -- I agree with that. | train | 4 |
train-00004-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"discourse markers"
],
"non_speech_events": true
} | [
"non-speech events"
] | I'm not that big on politics. I'm not that educated on it. I'm, I really think we should have more I agree with that. | I'm <laughter> not that big on politics. I'm not <laughter> -- -- that educated on it. I'm, - I really think we should have more -- I agree with that. | train | 4 |
train-00005-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | Yeah. She took such a rough go of it. their tax deal and, At least she was strong to try it and | Yeah. She took such a rough go of it. their tax deal and, - At least she was strong to try it and | train | 5 |
train-00005-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers",
"non-speech events",
"repairs"
] | Yeah. She took such a rough go of it. their tax deal and, At least she was strong to try it and | Yeah. She took such a rough go of it. [ they th-, + their tax ] deal and, - At least she was strong to {D like } try it and {D you know. } | train | 5 |
train-00005-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"discourse markers"
],
"non_speech_events": true
} | [
"discourse markers",
"non-speech events"
] | Yeah. She took such a rough go of it. their tax deal and, At least she was strong to try it and | Yeah. She took such a rough go of it. their tax deal and, - At least she was strong to try it and {D you know. } | train | 5 |
train-00005-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers"
],
"non_speech_events": false
} | [
"discourse markers"
] | Yeah. She took such a rough go of it. their tax deal and, At least she was strong to try it and | Yeah. She took such a rough go of it. their tax deal and, At least she was strong to {D like } try it and {D you know. } | train | 5 |
train-00006-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"editing terms",
"repairs"
],
"non_speech_events": false
} | [
"fillers",
"repairs"
] | right now I don't think we have too many. in the forefront. Schroeder's about the biggest one in Washington. | right now I don't think we have too many. {F Uh, } [ th-, + in the ] forefront. Schroeder's about the biggest one in Washington. | train | 6 |
train-00006-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"fillers",
"repairs"
] | right now I don't think we have too many. in the forefront. Schroeder's about the biggest one in Washington. | right now I don't think we have too many. {F Uh, } [ th-, + in the ] forefront. Schroeder's about the biggest one in Washington. | train | 6 |
train-00006-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"fillers"
] | right now I don't think we have too many. in the forefront. Schroeder's about the biggest one in Washington. | right now I don't think we have too many. {F Uh, } in the forefront. Schroeder's about the biggest one in Washington. | train | 6 |
train-00006-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"repairs"
] | right now I don't think we have too many. in the forefront. Schroeder's about the biggest one in Washington. | right now I don't think we have too many. [ th-, + in the ] forefront. Schroeder's about the biggest one in Washington. | train | 6 |
train-00007-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"repairs"
],
"non_speech_events": false
} | [
"repairs"
] | Okay. I think the first thing they said, I have written this down it would, do you think it's possible to have honesty in government or an honest government? | Okay. I think the first thing they said, I have written this down it would, [ is it po-, + do you think it's possible ] to have honesty in government or an honest government? | train | 7 |
train-00007-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"repairs"
] | Okay. I think the first thing they said, I have written this down it would, do you think it's possible to have honesty in government or an honest government? | Okay. I think the first thing they said, I have written this down it would, [ is it po-, + do you think it's possible ] to have honesty in government or an honest government? | train | 7 |
train-00007-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events",
"repairs"
] | Okay. I think the first thing they said, I have written this down it would, do you think it's possible to have honesty in government or an honest government? | Okay. I think the first thing they said, - I have written this down it would, - [ is it po-, + do you think it's possible ] to have honesty in government or an honest government? | train | 7 |
train-00007-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers"
],
"non_speech_events": true
} | [
"non-speech events"
] | Okay. I think the first thing they said, I have written this down it would, do you think it's possible to have honesty in government or an honest government? | Okay. I think the first thing they said, - I have written this down it would, - do you think it's possible to have honesty in government or an honest government? | train | 7 |
train-00008-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers"
],
"non_speech_events": true
} | [
"fillers"
] | other than that I think maybe it just depends on how you define honesty. | {F uh, } other than that I think maybe it just depends on how you define honesty. | train | 8 |
train-00008-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"fillers"
] | other than that I think maybe it just depends on how you define honesty. | {F uh, } other than that I think maybe it just depends on how you define honesty. | train | 8 |
train-00008-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"fillers"
] | other than that I think maybe it just depends on how you define honesty. | {F uh, } other than that I think maybe it just depends on how you define honesty. | train | 8 |
train-00008-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"editing terms"
],
"non_speech_events": false
} | [
"fillers"
] | other than that I think maybe it just depends on how you define honesty. | {F uh, } other than that I think maybe it just depends on how you define honesty. | train | 8 |
train-00009-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"repetitions"
],
"non_speech_events": true
} | [
"discourse markers"
] | how good a name can I make for myself and so forth. I guess it would be like anything else. | {D you know, } how good a name can I make for myself and so forth. I guess it would be like anything else. | train | 9 |
train-00009-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers"
] | how good a name can I make for myself and so forth. I guess it would be like anything else. | {D you know, } how good a name can I make for myself and so forth. I guess it would be like anything else. | train | 9 |
train-00009-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers"
] | how good a name can I make for myself and so forth. I guess it would be like anything else. | {D you know, } how good a name can I make for myself and so forth. I guess it would be like anything else. | train | 9 |
train-00009-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers"
] | how good a name can I make for myself and so forth. I guess it would be like anything else. | {D you know, } how good a name can I make for myself and so forth. I guess it would be like anything else. | train | 9 |
train-00010-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"fillers",
"non-speech events"
] | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, {F uh, } - | train | 10 |
train-00010-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers"
] | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, {F uh, } | train | 10 |
train-00010-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers"
] | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, {F uh, } | train | 10 |
train-00010-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, | I think ] a little more honesty in the campaigns would be, more to the point. If you knew what you were getting when you voted, - | train | 10 |
train-00011-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers"
] | The other thing they asked about was whether the verdict should be required to be unanimous. | The other thing they asked about was whether {F uh, } the verdict should be required to be unanimous. | train | 11 |
train-00011-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers"
] | The other thing they asked about was whether the verdict should be required to be unanimous. | The other thing they asked about was whether {F uh, } {F uh, } the verdict should be required to be unanimous. | train | 11 |
train-00011-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers"
] | The other thing they asked about was whether the verdict should be required to be unanimous. | The other thing they asked about was whether {F uh, } {F uh, } the verdict should be required to be unanimous. | train | 11 |
train-00011-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"fillers"
] | The other thing they asked about was whether the verdict should be required to be unanimous. | The other thing they asked about was whether {F uh, } the verdict should be required to be unanimous. | train | 11 |
train-00012-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"repetitions"
] | Because I think it's gotten just care-, we're not talking about issues we're talking about somebody's personal life we're getting away from what we ought to be looking at. | Because I think it's gotten just care-, we're not talking about issues we're talking about somebody's personal life [ + ] we're getting away from what we ought to be looking at. | train | 12 |
train-00012-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"editing terms",
"repetitions"
],
"non_speech_events": false
} | [
"repetitions"
] | Because I think it's gotten just care-, we're not talking about issues we're talking about somebody's personal life we're getting away from what we ought to be looking at. | Because I think it's gotten just care-, we're not talking about issues we're talking about somebody's personal life [ + ] we're getting away from what we ought to be looking at. | train | 12 |
train-00012-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"discourse markers",
"non-speech events",
"repetitions"
] | Because I think it's gotten just care-, we're not talking about issues we're talking about somebody's personal life we're getting away from what we ought to be looking at. | Because I think it's gotten just care-, {D well, } - we're not talking about issues we're talking about somebody's personal life [ + ] we're getting away from what we ought to be looking at <throat_clearing>. | train | 12 |
train-00012-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"repairs"
] | Because I think it's gotten just care-, we're not talking about issues we're talking about somebody's personal life we're getting away from what we ought to be looking at. | Because [ it's, + I think it's ] gotten just care-, we're not talking about issues we're talking about somebody's personal life we're getting away from what we ought to be looking at. | train | 12 |
train-00013-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers",
"repairs"
] | Actually, I think other countries may have it worse, the Japanese government is always having some kind of a scandal. | Actually, I think other countries may have it worse, {F uh, } the Japanese government is always having [ a, + some ] kind of a scandal. | train | 13 |
train-00013-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers",
"repairs"
] | Actually, I think other countries may have it worse, the Japanese government is always having some kind of a scandal. | Actually, I think other countries may have it worse, {F uh, } the Japanese government is always having [ a, + some ] kind of a scandal. | train | 13 |
train-00013-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms",
"repairs"
],
"non_speech_events": false
} | [
"repairs"
] | Actually, I think other countries may have it worse, the Japanese government is always having some kind of a scandal. | Actually, I think other countries may have it worse, the Japanese government is always having [ a, + some ] kind of a scandal. | train | 13 |
train-00013-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"fillers",
"non-speech events"
] | Actually, I think other countries may have it worse, the Japanese government is always having some kind of a scandal. | <throat_clearing> Actually, I think other countries may have it worse, {F uh, } the Japanese government is always having some kind of a scandal. | train | 13 |
train-00014-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"repetitions"
],
"non_speech_events": false
} | [
"fillers"
] | I've got a male Siamese about a year old here. I live alone in an apartment, except for the cat. | I've got a male Siamese about a year old here. I live alone in an apartment, {F uh, } except for the cat. | train | 14 |
train-00014-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"editing terms",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events",
"repairs"
] | I've got a male Siamese about a year old here. I live alone in an apartment, except for the cat. | I've got a male Siamese about a year old here. I live alone in an apartment, [ and, + ] -- -- except for the cat. | train | 14 |
train-00014-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"fillers",
"repairs"
] | I've got a male Siamese about a year old here. I live alone in an apartment, except for the cat. | I've got a male Siamese about a year old here. I live alone in an apartment, [ and, + ] {F uh, } except for the cat. | train | 14 |
train-00014-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repairs"
],
"non_speech_events": true
} | [
"fillers",
"non-speech events",
"repairs"
] | I've got a male Siamese about a year old here. I live alone in an apartment, except for the cat. | I've got a male Siamese about a year old here. I live alone in an apartment, [ and, + ] {F uh, } -- -- except for the cat. | train | 14 |
train-00015-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers",
"fillers"
] | we have a cat, he's probably, a good two years old, big, old, fat and sassy tabby. | {D Well, } we have a cat, {F um, } he's probably, {F oh, } a good two years old, big, old, fat and sassy tabby. | train | 15 |
train-00015-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers"
] | we have a cat, he's probably, a good two years old, big, old, fat and sassy tabby. | {D Well, } we have a cat, he's probably, a good two years old, big, old, fat and sassy tabby. | train | 15 |
train-00015-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers",
"fillers"
] | we have a cat, he's probably, a good two years old, big, old, fat and sassy tabby. | {D Well, } we have a cat, {F um, } he's probably, {F oh, } a good two years old, big, old, fat and sassy tabby. | train | 15 |
train-00015-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers"
],
"non_speech_events": false
} | [
"fillers"
] | we have a cat, he's probably, a good two years old, big, old, fat and sassy tabby. | we have a cat, {F um, } he's probably, a good two years old, big, old, fat and sassy tabby. | train | 15 |
train-00016-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": false
} | [
"discourse markers"
] | yeah, he has been really aggressive, especially towards my little girl, for some reason. He's, been so mean to her. | {D Well, } yeah, he has been really aggressive, especially towards my little girl, for some reason. He's, been so mean to her. | train | 16 |
train-00016-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"fillers",
"non-speech events",
"repairs"
] | yeah, he has been really aggressive, especially towards my little girl, for some reason. He's, been so mean to her. | yeah, [ he's, + he has ] been really aggressive, especially towards my little girl, for -- -- some reason. He's, {F uh, } been so mean to her <laughter>. | train | 16 |
train-00016-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": false
} | [
"discourse markers",
"fillers"
] | yeah, he has been really aggressive, especially towards my little girl, for some reason. He's, been so mean to her. | {D Well, } yeah, {F uh, } he has been really aggressive, {F uh, } especially towards my little girl, for some reason. He's, {F uh, } been so mean to her. | train | 16 |
train-00016-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"discourse markers",
"fillers",
"non-speech events"
] | yeah, he has been really aggressive, especially towards my little girl, for some reason. He's, been so mean to her. | {D Well, } yeah, {F uh, } he has been really aggressive, {F uh, } especially towards my little girl, for -- -- some reason. He's, {F uh, } been so mean to her <laughter>. | train | 16 |
train-00017-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal. | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal <laughter>. | train | 17 |
train-00017-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms"
],
"non_speech_events": true
} | [
"non-speech events"
] | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal. | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal <laughter>. | train | 17 |
train-00017-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal. | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal <laughter>. | train | 17 |
train-00017-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms"
],
"non_speech_events": true
} | [
"non-speech events"
] | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal. | He meows and wants to get in the bathroom while he showers, it's so hilarious. he's quite a animal <laughter>. | train | 17 |
train-00018-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers"
] | Yeah, I guess. as long as he doesn't get too aggressive. I guess with the daughter he's all right. | Yeah, I guess. {D Well } as long as he doesn't get too aggressive. I guess with the daughter he's all right. | train | 18 |
train-00018-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers"
],
"non_speech_events": false
} | [
"discourse markers"
] | Yeah, I guess. as long as he doesn't get too aggressive. I guess with the daughter he's all right. | Yeah, I guess. {D Well } as long as he doesn't get too aggressive. I guess with the daughter he's all right. | train | 18 |
train-00018-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms"
],
"non_speech_events": false
} | [
"discourse markers"
] | Yeah, I guess. as long as he doesn't get too aggressive. I guess with the daughter he's all right. | Yeah, I guess. {D Well } as long as he doesn't get too aggressive. I guess with the daughter he's all right. | train | 18 |
train-00018-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": false
} | [
"discourse markers"
] | Yeah, I guess. as long as he doesn't get too aggressive. I guess with the daughter he's all right. | Yeah, I guess. {D Well } as long as he doesn't get too aggressive. I guess with the daughter he's all right. | train | 18 |
train-00019-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | Were it not for my son, the cat would not be in this house. | <laughter> Were it not for my son, the cat would not be in this house <laughter>. | train | 19 |
train-00019-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers"
],
"non_speech_events": true
} | [
"non-speech events"
] | Were it not for my son, the cat would not be in this house. | <laughter> Were it not for my son, the cat would not be in this house <laughter>. | train | 19 |
train-00019-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | Were it not for my son, the cat would not be in this house. | <laughter> Were it not for my son, the cat would not be in this house <laughter>. | train | 19 |
train-00019-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"non-speech events"
] | Were it not for my son, the cat would not be in this house. | <laughter> Were it not for my son, the cat would not be in this house <laughter>. | train | 19 |
train-00020-none | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "none",
"phenomena": [],
"non_speech_events": false
} | [] | I guess, I would imagine that they don't have many more than one to start with, either. | I guess, I would imagine that they don't have many more than one to start with, either. | train | 20 |
train-00021-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"editing terms"
],
"non_speech_events": true
} | [
"non-speech events"
] | in capital. I think in lesser cases it's ten out of twelve or five out of six, whatever. | -- in capital. I think in lesser cases it's ten out of twelve or five out of six, whatever. | train | 21 |
train-00021-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers"
] | in capital. I think in lesser cases it's ten out of twelve or five out of six, whatever. | in capital. I think in {D like, } lesser cases it's {D like } ten out of twelve or five out of six, whatever. | train | 21 |
train-00021-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": true
} | [
"discourse markers",
"fillers",
"non-speech events"
] | in capital. I think in lesser cases it's ten out of twelve or five out of six, whatever. | -- in capital. I think in {D like, } {F uh, } lesser cases it's {D like } ten out of twelve or five out of six, whatever. | train | 21 |
train-00021-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions"
],
"non_speech_events": false
} | [
"discourse markers"
] | in capital. I think in lesser cases it's ten out of twelve or five out of six, whatever. | in capital. I think in lesser cases it's {D like } ten out of twelve or five out of six, whatever. | train | 21 |
train-00022-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | great. I've been doing it for about two weeks No my, | great. I've been doing it for about two weeks No my, - | train | 22 |
train-00022-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers",
"repetitions"
] | great. I've been doing it for about two weeks No my, | {D well, } great. I've been doing it for about two weeks [ no my, + No my, ] | train | 22 |
train-00022-02 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "light",
"phenomena": [
"fillers",
"discourse markers",
"editing terms",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers",
"fillers",
"repetitions"
] | great. I've been doing it for about two weeks No my, | {F Oh, } {D well, } great. I've been doing it for about two weeks {F uh, } [ no my, + No my, ] | train | 22 |
train-00022-03 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"discourse markers",
"repetitions",
"repairs"
],
"non_speech_events": false
} | [
"discourse markers",
"repetitions"
] | great. I've been doing it for about two weeks No my, | {D well, } great. I've been doing it for about two weeks [ no my, + No my, ] | train | 22 |
train-00023-00 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "heavy",
"phenomena": [
"fillers",
"repairs"
],
"non_speech_events": true
} | [
"fillers",
"non-speech events",
"repairs"
] | While I'm at work sometimes, on the computer network I read the news sometimes, that's kind of helpful. | While I'm at work sometimes, [ I, + ] on the computer network -- -- {F um, } I read the news sometimes, that's kind of helpful. | train | 23 |
train-00023-01 | [
{
"role": "system",
"content": "You are a spoken-transcript simulator. Given a clean written transcript and a control instruction line, produce the natural spoken version of that transcript with realistic disfluencies inserted.\n\nThe control instruction is a single line of bracketed fields:\n[Disfluency le... | {
"level": "moderate",
"phenomena": [
"editing terms",
"repetitions"
],
"non_speech_events": true
} | [
"non-speech events"
] | While I'm at work sometimes, on the computer network I read the news sometimes, that's kind of helpful. | While I'm at work sometimes, on the computer network -- -- I read the news sometimes, that's kind of helpful. | train | 23 |
End of preview. Expand in Data Studio
README.md exists but content is empty.
- Downloads last month
- 47