cnbeining/OpenHermes-2.5-Mistral-7B-Sentence-Segmentation
Updated
prompt
stringlengths 437
3.49k
| chosen
stringlengths 44
3.23k
| rejected
stringlengths 152
3.2k
|
|---|---|---|
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["We-we", "suffer", "a", "great", "amount", "of", "pain,", "we", "create", "a", "world", "view", "to", "help", "us", "escape", "that", "pain,", "and", "then", "we", "forget", "that", "we", "created", "the", "world", "view,", "and", "so", "were", "doomed", "to", "repeat", "a", "lot", "of", "our", "failures."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["We-we", "suffer", "a", "great", "amount", "of", "pain,"], ["we", "create", "a", "world", "view", "to", "help", "us", "escape", "that", "pain,"], ["and", "then", "we", "forget", "that", "we", "created", "the", "world", "view,"], ["and", "so", "were", "doomed", "to", "repeat", "a", "lot", "of", "our", "failures."]]}
```<end_of_turn> |
```json
{"output": [["We-we", "suffer", "a", "great", "amount", "of", "pain,", "we", "create", "a", "world", "view", "to", "help", "us", "escape", "that", "pain,", "and", "then", "we", "forget", "that", "we", "created", "the", "world", "view,", "and", "so", "were", "doomed", "to", "repeat", "a", "lot", "of", "our", "failures."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["You", "know,", "hunting,", "my", "first", "bow,", "fishing,", "anything", "that", "I", "ever", "did", "was,", "it", "all", "started", "right", "here", "at", "the", "Indian", "village."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["You", "know,", "hunting,", "my", "first", "bow,", "fishing,"], ["anything", "that", "I", "ever", "did", "was,", "it", "all", "started", "right", "here", "at", "the", "Indian", "village."]]}
```<end_of_turn> |
```json
{"output": [["You", "know,", "hunting,", "my", "first", "bow,", "fishing,", "anything", "that", "I", "ever", "did", "was,", "it", "all", "started", "right", "here", "at", "the", "Indian", "village."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["We", "basically", "are", "selling", "machines", "that", "print", "money,", "thats", "what", "this", "is.", "Bitcoin-mining", "hardware", "is", "a", "machine", "that", "prints", "money."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["We", "basically", "are", "selling", "machines", "that", "print", "money,"], ["thats", "what", "this", "is.", "Bitcoin-mining", "hardware", "is", "a", "machine", "that", "prints", "money."]]}
```<end_of_turn> |
```json
{"output": [["We", "basically", "are", "selling", "machines", "that", "print", "money,", "thats", "what", "this", "is.", "Bitcoin-mining", "hardware", "is", "a", "machine", "that", "prints", "money."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Ladies", "and", "gentlemen,", "please", "stand", "for", "the", "Canadian", "national", "anthem.", "And", "everybody", "stood,", "and", "then", "they", "just", "played,", "like,", "a", "tape", "somewhere."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Ladies", "and", "gentlemen,", "please", "stand", "for", "the", "Canadian", "national", "anthem."], ["And", "everybody", "stood,", "and", "then", "they", "just", "played,", "like,", "a", "tape", "somewhere."]]}
```<end_of_turn> |
```json
{"output": [["Ladies", "and", "gentlemen,", "please", "stand", "for", "the", "Canadian", "national", "anthem.", "And", "everybody", "stood,", "and", "then", "they", "just", "played,", "like,", "a", "tape", "somewhere."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["If", "there", "is", "a", "narrower", "range", "of", "opinion", "in", "the", "United", "States", "and", "it", "is", "harder", "to", "express", "a", "variety", "of", "different", "opinions,", "why", "do", "you", "live", "in", "the", "US?"]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["If", "there", "is", "a", "narrower", "range", "of", "opinion", "in", "the", "United", "States", "and", "it", "is", "harder", "to", "express"], ["a", "variety", "of", "different", "opinions,", "why", "do", "you", "live", "in", "the", "US?"]]}
```<end_of_turn> |
```json
{"output": [["If", "there", "is", "a", "narrower", "range", "of", "opinion", "in", "the", "United", "States", "and", "it", "is", "harder", "to", "express", "a", "variety", "of", "different", "opinions,", "why", "do", "you", "live", "in", "the", "US?"]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "entitlement,", "its", "when", "you", "feel", "as", "though", "you", "deserve", "the", "benefits", "of", "something", "without", "giving", "up", "the", "cost", "of", "something."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "entitlement,", "its", "when", "you", "feel", "as", "though", "you", "deserve"], ["the", "benefits", "of", "something", "without", "giving", "up", "the", "cost", "of", "something."]]}
```<end_of_turn> |
```json
{"output": [["And", "entitlement,", "its", "when", "you", "feel", "as", "though", "you", "deserve", "the", "benefits", "of", "something", "without", "giving", "up", "the", "cost", "of", "something."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["One", "of", "the", "drugs", "being", "used", "was", "not", "even", "allowed", "to", "be", "used", "in", "animals", "and", "he", "was", "giving", "this", "to", "people."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["One", "of", "the", "drugs", "being", "used", "was", "not", "even", "allowed"], ["to", "be", "used", "in", "animals", "and", "he", "was", "giving", "this", "to", "people."]]}
```<end_of_turn> |
```json
{"output": [["One", "of", "the", "drugs", "being", "used", "was", "not", "even", "allowed", "to", "be", "used", "in", "animals", "and", "he", "was", "giving", "this", "to", "people."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["-", "He", "wants", "to", "use", "ideas", "of", "the", "legendary", "king", "of", "the", "Britons,", "because", "of", "course", "the", "Tudors", "are", "Welsh,", "the", "descendants", "of", "the", "Britons."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["-", "He", "wants", "to", "use", "ideas", "of", "the", "legendary", "king"], ["of", "the", "Britons,", "because", "of", "course", "the", "Tudors", "are", "Welsh,"], ["the", "descendants", "of", "the", "Britons."]]}
```<end_of_turn> |
```json
{"output": [["-", "He", "wants", "to", "use", "ideas", "of", "the", "legendary", "king", "of", "the", "Britons,", "because", "of", "course", "the", "Tudors", "are", "Welsh,", "the", "descendants", "of", "the", "Britons."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["\u266a", "\u266a", "It", "was", "a", "star-studded", "night", "Saturday", "as", "celebrities", "flooded", "Atlanta", "to", "celebrate", "the", "grand", "opening", "of", "Tyler", "Perry", "Studios."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["\u266a", "\u266a", "It", "was", "a", "star-studded", "night", "Saturday", "as", "celebrities"], ["flooded", "Atlanta", "to", "celebrate", "the", "grand", "opening", "of", "Tyler", "Perry", "Studios."]]}
```<end_of_turn> |
```json
{"output": [["\u266a", "\u266a", "It", "was", "a", "star-studded", "night", "Saturday", "as", "celebrities", "flooded", "Atlanta", "to", "celebrate", "the", "grand", "opening", "of", "Tyler", "Perry", "Studios."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["All", "these", "people", "want", "to", "hear", "Shere.", "And", "then", "I", "heard", "some", "of", "the", "people", "say,", "Well,", "they", "want", "to", "see", "what", "shes", "gonna", "do."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["All", "these", "people", "want", "to", "hear", "Shere."], ["And", "then", "I", "heard", "some", "of", "the", "people", "say,"], ["Well,", "they", "want", "to", "see", "what", "shes", "gonna", "do."]]}
```<end_of_turn> |
```json
{"output": [["All", "these", "people", "want", "to", "hear", "Shere.", "And", "then", "I", "heard", "some", "of", "the", "people", "say,", "Well,", "they", "want", "to", "see", "what", "shes", "gonna", "do."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["TOBY", "JONES:", "For", "more", "than", "ten", "years,", "Alan", "Bates", "and", "his", "team", "of", "campaigners", "had", "been", "carefully", "gathering", "evidence", "bit", "by", "bit", "against", "the", "Post", "Office."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["TOBY", "'JONES:'", "For", "more", "than", "ten", "years,"], ["Alan", "Bates", "and", "his", "team", "of", "campaigners"], ["had", "been", "carefully", "gathering", "evidence", "bit", "by", "bit"], ["against", "the", "Post", "Office."]]}
```<end_of_turn> |
```json
{"output": [["TOBY", "JONES:", "For", "more", "than", "ten", "years,", "Alan", "Bates", "and", "his", "team", "of", "campaigners", "had", "been", "carefully", "gathering", "evidence", "bit", "by", "bit", "against", "the", "Post", "Office."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "was", "at", "a", "hospital", "the", "other", "night", "where,", "I", "think", "there", "were", "actually", "a", "few", "coronavirus", "patients,", "and", "I", "shook", "hands", "with", "everybody,", "youll", "be", "pleased", "to", "know,", "and", "I", "continue", "to", "shake", "hands", "and,", "er..."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "was", "at", "a", "hospital", "the", "other", "night", "where,"], ["I", "think", "there", "were", "actually", "a", "few", "coronavirus", "patients,"], ["and", "I", "shook", "hands", "with", "everybody,"], ["youll", "be", "pleased", "to", "know,", "and", "I", "continue", "to", "shake", "hands", "and,", "er..."]]}
```<end_of_turn> |
```json
{"output": [["I", "was", "at", "a", "hospital", "the", "other", "night", "where,", "I", "think", "there", "were", "actually", "a", "few", "coronavirus", "patients,", "and", "I", "shook", "hands", "with", "everybody,", "youll", "be", "pleased", "to", "know,", "and", "I", "continue", "to", "shake", "hands", "and,", "er..."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["It", "was", "of", "a", "300", "million-year-old", "creature", "called", "Pholiderpeton,", "thought", "to", "be", "one", "of", "the", "first", "tetrapods", "to", "venture", "onto", "land."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["It", "was", "of", "a", "'300'", "million-year-old", "creature", "called", "Pholiderpeton,"], ["thought", "to", "be", "one", "of", "the", "first", "tetrapods", "to", "venture", "onto", "land."]]}
```<end_of_turn> |
```json
{"output": [["It", "was", "of", "a", "300", "million-year-old", "creature", "called", "Pholiderpeton,", "thought", "to", "be", "one", "of", "the", "first", "tetrapods", "to", "venture", "onto", "land."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Ueli", "Steck", "had", "no", "photos", "or", "video", "that", "captured", "the", "moment,", "so", "he", "simply", "cannot", "prove", "that", "he", "did", "reach", "the", "summit."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Ueli", "Steck", "had", "'no'", "photos", "or", "video", "that", "captured", "the", "moment,"], ["so", "he", "simply", "cannot", "prove", "that", "he", "did", "reach", "the", "summit."]]}
```<end_of_turn> |
```json
{"output": [["Ueli", "Steck", "had", "no", "photos", "or", "video", "that", "captured", "the", "moment,", "so", "he", "simply", "cannot", "prove", "that", "he", "did", "reach", "the", "summit."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["As", "he", "is", "enrolled", "into", "the", "school", "system,", "despite", "being", "a", "brilliant", "people,", "Petiots", "behavior", "becomes", "increasingly", "disruptive", "and", "dangerous."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["As", "he", "is", "enrolled", "into", "the", "school", "system,"], ["despite", "being", "a", "brilliant", "people,", "Petiots", "behavior", "becomes", "increasingly", "disruptive", "and", "dangerous."]]}
```<end_of_turn> |
```json
{"output": [["As", "he", "is", "enrolled", "into", "the", "school", "system,", "despite", "being", "a", "brilliant", "people,", "Petiots", "behavior", "becomes", "increasingly", "disruptive", "and", "dangerous."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Although", "Ron", "DeSantis", "said", "that", "it", "was", "intended", "to", "protect", "only", "the", "youngest", "children", "from,", "you", "know,", "being", "indoctrinated", "into", "homosexuality,", "the", "laws", "been", "expanded", "to", "apply", "to", "kids", "in", "grades", "K", "through", "12."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Although", "Ron", "DeSantis", "said", "that", "it", "was", "intended", "to", "protect", "only", "the", "youngest", "children", "from,"], ["you", "know,", "being", "indoctrinated", "into", "homosexuality,", "the", "laws", "been", "expanded", "to", "apply", "to", "kids", "in", "grades", "K", "through", "'12.'"]]}
```<end_of_turn> |
```json
{"output": [["Although", "Ron", "DeSantis", "said", "that", "it", "was", "intended", "to", "protect", "only", "the", "youngest", "children", "from,", "you", "know,", "being", "indoctrinated", "into", "homosexuality,", "the", "laws", "been", "expanded", "to", "apply", "to", "kids", "in", "grades", "K", "through", "12."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "a", "screw", "came", "in,", "he", "said,", "censors", "wanna", "know", "if", "this", "is", "true", "or", "not,", "it", "was", "a", "letter", "from", "President", "Nixon."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "a", "screw", "came", "in,", "he", "said,", "censors", "wanna", "know", "if", "this", "is", "'true'", "or", "not,"], ["it", "was", "a", "letter", "from", "President", "Nixon."]]}
```<end_of_turn> |
```json
{"output": [["And", "a", "screw", "came", "in,", "he", "said,", "censors", "wanna", "know", "if", "this", "is", "true", "or", "not,", "it", "was", "a", "letter", "from", "President", "Nixon."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["It", "made", "me", "more", "fearful", "to", "embrace", "who", "I", "was", "because", "it", "reminded", "me", "how", "much", "hate", "there", "is", "out", "there."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["It", "made", "me", "more", "fearful", "to", "embrace", "who", "I", "was", "because"], ["it", "reminded", "me", "how", "much", "hate", "there", "is", "out", "there."]]}
```<end_of_turn> |
```json
{"output": [["It", "made", "me", "more", "fearful", "to", "embrace", "who", "I", "was", "because", "it", "reminded", "me", "how", "much", "hate", "there", "is", "out", "there."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Captive", "bears", "are", "regularly", "drugged", "and", "then", "a", "long", "needle", "is", "used", "to", "repeatedly", "pierce", "their", "abdomen", "and", "extract", "the", "bile."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Captive", "bears", "are", "regularly", "drugged", "and", "then", "a", "long", "needle"], ["is", "used", "to", "repeatedly", "pierce", "their", "abdomen", "and", "extract", "the", "bile."]]}
```<end_of_turn> |
```json
{"output": [["Captive", "bears", "are", "regularly", "drugged", "and", "then", "a", "long", "needle", "is", "used", "to", "repeatedly", "pierce", "their", "abdomen", "and", "extract", "the", "bile."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["But", "if", "youve", "ever", "been", "backstage", "on", "a", "festival,", "look", "down", "the", "row", "and", "see", "all", "the", "guys", "working,", "thats", "like", "the", "boulevard", "of", "broken", "dreams", "back", "there."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["But", "if", "youve", "ever", "been", "backstage", "'on'", "a", "festival,"], ["look", "down", "the", "row", "and", "see", "all", "the", "guys", "working,"], ["thats", "like", "the", "boulevard", "of", "broken", "dreams", "back", "there."]]}
```<end_of_turn> |
```json
{"output": [["But", "if", "youve", "ever", "been", "backstage", "on", "a", "festival,", "look", "down", "the", "row", "and", "see", "all", "the", "guys", "working,", "thats", "like", "the", "boulevard", "of", "broken", "dreams", "back", "there."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["THE", "CHOICE", "IS", "YOURS.", "I\u2019VE", "BEEN", "PRINCIPAL", "HERE", "FOR", "20", "YEARS,", "AND", "HE\u2019S", "IN", "MY", "TOP", "10", "OF", "THE,", "I", "GUESS,", "12,000-15,000", "STUDENTS", "THAT", "HAVE", "GONE", "THROUGH", "HERE", "OVER", "THAT", "TIME,", "SO", "A", "PRETTY", "TERRIFIC", "KID."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["THE", "CHOICE", "IS", "YOURS.", "I\u2019VE", "BEEN", "PRINCIPAL", "HERE", "FOR", "'20'", "YEARS,"], ["AND", "HE\u2019S", "IN", "MY", "TOP", "'10'", "OF", "THE,", "I", "GUESS,", "12,000-15,000"], ["STUDENTS", "THAT", "HAVE", "GONE", "THROUGH", "HERE", "OVER", "THAT", "TIME,", "SO", "A", "PRETTY", "TERRIFIC", "KID."]]}
```<end_of_turn> |
```json
{"output": [["THE", "CHOICE", "IS", "YOURS.", "I\u2019VE", "BEEN", "PRINCIPAL", "HERE", "FOR", "20", "YEARS,", "AND", "HE\u2019S", "IN", "MY", "TOP", "10", "OF", "THE,", "I", "GUESS,", "12,000-15,000", "STUDENTS", "THAT", "HAVE", "GONE", "THROUGH", "HERE", "OVER", "THAT", "TIME,", "SO", "A", "PRETTY", "TERRIFIC", "KID."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Im", "thinking", "to", "myself,", "What", "is", "going", "on", "down", "there?", "And", "so", "I", "had", "to", "go", "look,", "and", "its", "these", "guys", "just", "thrashing", "their", "heads", "and", "the", "hairs", "going,", "the", "crowd", "are", "doing", "the", "same", "thing."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Im", "thinking", "to", "myself,"], ["What", "is", "going", "'on'", "down", "there?"], ["And", "so", "I", "had", "to", "go", "look,"], ["and", "its", "these", "guys", "just", "thrashing", "their"]]}
```<end_of_turn> |
```json
{"output": [["Im", "thinking", "to", "myself,", "What", "is", "going", "on", "down", "there?", "And", "so", "I", "had", "to", "go", "look,", "and", "its", "these", "guys", "just", "thrashing", "their", "heads", "and", "the", "hairs", "going,", "the", "crowd", "are", "doing", "the", "same", "thing."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "thought,", "Oh,", "my", "God,", "I", "really", "have", "no", "idea", "the", "toll", "this", "has", "taken.", "Why", "is", "Shere", "Hite", "so", "sensitive", "about", "criticism?"]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "thought,", "Oh,", "my", "God,", "I", "really", "have", "'no'", "idea"], ["the", "toll", "this", "has", "taken."], ["Why", "is", "Shere", "Hite", "so", "sensitive", "about", "criticism?"]]}
```<end_of_turn> |
```json
{"output": [["I", "thought,", "Oh,", "my", "God,", "I", "really", "have", "no", "idea", "the", "toll", "this", "has", "taken.", "Why", "is", "Shere", "Hite", "so", "sensitive", "about", "criticism?"]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Theyre", "like,", "Hey,", "whats", "up?", "We", "saw", "you", "made", "the", "switch.", "Curious", "if", "you", "could", "do", "a", "quick", "promo", "bit", "for", "us.", "Id", "say,", "Sure.", "And", "theyd", "say,", "Action!", "Id", "just", "be", "like,", "Um...", "Its", "like", "a", "little", "bit", "better.", "I", "dont", "know.", "-", "-"]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Theyre", "like,", "Hey,", "whats", "up?"], ["We", "saw", "you", "made", "the", "switch."], ["Curious", "if", "you", "could", "do", "a", "quick", "promo", "bit", "for", "us."], ["Id", "say,", "Sure.", "And", "theyd", "say,", "Action!"], ["Id", "just", "be", "like,", "Um..."], ["Its", "like", "a", "little", "bit", "better."], ["I", "dont", "know.", "'-'", "'-'"]]}
```<end_of_turn> |
```json
{"output": [["Theyre", "like,", "Hey,", "whats", "up?", "We", "saw", "you", "made", "the", "switch.", "Curious", "if", "you", "could", "do", "a", "quick", "promo", "bit", "for", "us.", "Id", "say,", "Sure.", "And", "theyd", "say,", "Action!", "Id", "just", "be", "like,", "Um...", "Its", "like", "a", "little", "bit", "better.", "I", "dont", "know.", "-", "-"]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["We", "focused", "most", "of", "today", "orienting", "around", "the", "outside", "of", "the", "vessel,", "focusing", "on", "that", "outer", "hull", "planking", "and", "focusing", "on", "the", "condition", "of", "the", "timbers", "themselves."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["We", "focused", "most", "of", "today"], ["orienting", "around", "the", "outside", "of", "the", "vessel,"], ["focusing", "'on'", "that", "outer", "hull", "planking"], ["and", "focusing", "'on'", "the", "condition", "of", "the", "timbers", "themselves."]]}
```<end_of_turn> |
```json
{"output": [["We", "focused", "most", "of", "today", "orienting", "around", "the", "outside", "of", "the", "vessel,", "focusing", "on", "that", "outer", "hull", "planking", "and", "focusing", "on", "the", "condition", "of", "the", "timbers", "themselves."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "crash", "of", "1929", "was", "like", "a", "perfect", "storm", "in", "which", "all", "of", "these", "improbably", "things", "came", "togther", "at", "the", "wrong", "time", "in", "the", "wrong", "way."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "crash", "of", "'1929'", "was", "like", "a", "perfect", "storm"], ["in", "which", "all", "of", "these", "improbably", "things", "came", "togther"], ["at", "the", "wrong", "time", "in", "the", "wrong", "way."]]}
```<end_of_turn> |
```json
{"output": [["The", "crash", "of", "1929", "was", "like", "a", "perfect", "storm", "in", "which", "all", "of", "these", "improbably", "things", "came", "togther", "at", "the", "wrong", "time", "in", "the", "wrong", "way."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["So", "Pierre", "produced", "a", "book", "in", "which", "all", "the", "arguments", "of", "Faurisson", "were", "to", "be", "put", "in", "front", "of", "the", "court."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["So", "Pierre", "produced", "a", "book"], ["in", "which", "all", "the", "arguments", "of", "Faurisson"], ["were", "to", "be", "put", "in", "front", "of", "the", "court."]]}
```<end_of_turn> |
```json
{"output": [["So", "Pierre", "produced", "a", "book", "in", "which", "all", "the", "arguments", "of", "Faurisson", "were", "to", "be", "put", "in", "front", "of", "the", "court."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Its", "trying", "to", "just", "imagine", "what", "they", "were", "like", "and", "what", "they", "were", "doing,", "and", "what", "life", "was", "like", "for", "them,", "and", "just", "wanting", "to", "be", "there."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Its", "trying", "to", "just", "imagine", "what", "they", "were", "like"], ["and", "what", "they", "were", "doing,", "and", "what", "life", "was", "like", "for", "them,"], ["and", "just", "wanting", "to", "be", "there."]]}
```<end_of_turn> |
```json
{"output": [["Its", "trying", "to", "just", "imagine", "what", "they", "were", "like", "and", "what", "they", "were", "doing,", "and", "what", "life", "was", "like", "for", "them,", "and", "just", "wanting", "to", "be", "there."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["-", "Most", "of", "the", "time,", "when", "I", "say", "to", "somebody,", "Im", "a", "guitar", "tech,", "a", "lot", "of", "times", "they", "dont", "know", "what", "that", "is."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["-", "Most", "of", "the", "time,", "when", "I", "say", "to", "somebody,"], ["Im", "a", "guitar", "tech,", "a", "lot", "of", "times"], ["they", "dont", "know", "what", "that", "is."]]}
```<end_of_turn> |
```json
{"output": [["-", "Most", "of", "the", "time,", "when", "I", "say", "to", "somebody,", "Im", "a", "guitar", "tech,", "a", "lot", "of", "times", "they", "dont", "know", "what", "that", "is."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Now", "the", "motorcade", "will", "very", "shortly", "start", "to", "move", "out", "for", "downtown", "Dallas,", "where", "thousands", "should", "already", "be", "on", "the", "street", "right", "now", "waiting", "for", "a", "view", "of", "the", "president", "and", "his", "wife."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Now", "the", "motorcade", "will", "very", "shortly", "start", "to", "move", "out"], ["for", "downtown", "Dallas,", "where", "thousands", "should", "already", "be", "'on'"], ["the", "street", "right", "now", "waiting", "for", "a", "view", "of", "the", "president", "and", "his", "wife."]]}
```<end_of_turn> |
```json
{"output": [["Now", "the", "motorcade", "will", "very", "shortly", "start", "to", "move", "out", "for", "downtown", "Dallas,", "where", "thousands", "should", "already", "be", "on", "the", "street", "right", "now", "waiting", "for", "a", "view", "of", "the", "president", "and", "his", "wife."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["He", "more", "and", "more", "felt", "that", "the", "project", "was", "being", "taken", "away", "from", "him", "because", "the", "project", "got", "so", "big", "that", "he", "suddenly", "wasnt", "the", "only", "one", "involved", "in", "it."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["He", "more", "and", "more", "felt", "that", "the", "project"], ["was", "being", "taken", "away", "from", "him", "because", "the", "project"], ["got", "so", "big", "that", "he", "suddenly", "wasnt", "the", "only", "one", "involved", "in", "it."]]}
```<end_of_turn> |
```json
{"output": [["He", "more", "and", "more", "felt", "that", "the", "project", "was", "being", "taken", "away", "from", "him", "because", "the", "project", "got", "so", "big", "that", "he", "suddenly", "wasnt", "the", "only", "one", "involved", "in", "it."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "Secret", "Service", "decided,", "while", "the", "plane", "was", "en", "route", "to", "Washington,", "that", "the", "body", "would", "go", "by", "helicopter."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "Secret", "Service", "decided,"], ["while", "the", "plane", "was", "en", "route", "to", "Washington,"], ["that", "the", "body", "would", "go", "by", "helicopter."]]}
```<end_of_turn> |
```json
{"output": [["The", "Secret", "Service", "decided,", "while", "the", "plane", "was", "en", "route", "to", "Washington,", "that", "the", "body", "would", "go", "by", "helicopter."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["They", "show", "up", "50", "or", "60", "or", "80", "at", "the", "same", "time", "in", "the", "restaurant", "and", "the", "store.", "And", "we", "were", "not", "used", "to", "that."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["They", "show", "up", "'50'", "or", "'60'", "or", "'80'"], ["at", "the", "same", "time", "in", "the", "restaurant", "and", "the", "store."], ["And", "we", "were", "not", "used", "to", "that."]]}
```<end_of_turn> |
```json
{"output": [["They", "show", "up", "50", "or", "60", "or", "80", "at", "the", "same", "time", "in", "the", "restaurant", "and", "the", "store.", "And", "we", "were", "not", "used", "to", "that."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["For", "centuries,", "astronomers", "and", "mathematicians", "have", "gazed", "in", "awe", "and", "wonder", "at", "the", "heavens,", "seeking", "to", "discover", "if", "theres", "a", "master", "plan."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["For", "centuries,", "astronomers", "and", "mathematicians", "have", "gazed", "in", "awe"], ["and", "wonder", "at", "the", "heavens,", "seeking", "to", "discover"], ["if", "theres", "a", "master", "plan."]]}
```<end_of_turn> |
```json
{"output": [["For", "centuries,", "astronomers", "and", "mathematicians", "have", "gazed", "in", "awe", "and", "wonder", "at", "the", "heavens,", "seeking", "to", "discover", "if", "theres", "a", "master", "plan."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Hed", "say,", "What", "are", "some", "of", "the", "flavor", "profiles", "youre", "getting", "from", "that?", "Im", "a", "sommelier.", "And", "Id", "say,", "Um,", "Im", "getting,", "like,", "salty", "tears.", "And", "jealousy.", "And", "then", "hed", "feel", "bad", "and", "give", "me", "a", "kiss."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Hed", "say,", "What", "are", "some", "of", "the", "flavor", "profiles"], ["youre", "getting", "from", "that?"], ["Im", "a", "sommelier."], ["And", "Id", "say,", "Um,", "Im", "getting,", "like,", "salty", "tears."], ["And", "jealousy."], ["And", "then", "hed", "feel", "bad", "and", "give", "me", "a", "kiss."]]}
```<end_of_turn> |
```json
{"output": [["Hed", "say,", "What", "are", "some", "of", "the", "flavor", "profiles", "youre", "getting", "from", "that?", "Im", "a", "sommelier.", "And", "Id", "say,", "Um,", "Im", "getting,", "like,", "salty", "tears.", "And", "jealousy.", "And", "then", "hed", "feel", "bad", "and", "give", "me", "a", "kiss."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "Knellers", "bodies", "are", "the", "first", "of", "nine", "dissected", "cadavers", "the", "police", "discover", "in", "the", "River", "Seine", "over", "the", "coming", "months."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "Knellers", "bodies", "are", "the", "first", "of", "nine"], ["dissected", "cadavers", "the", "police", "discover", "in", "the", "River", "Seine"], ["over", "the", "coming", "months."]]}
```<end_of_turn> |
```json
{"output": [["The", "Knellers", "bodies", "are", "the", "first", "of", "nine", "dissected", "cadavers", "the", "police", "discover", "in", "the", "River", "Seine", "over", "the", "coming", "months."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["All", "the", "negativity", "hit", "me", "right", "and", "left", "about", "how", "he", "should", "be", "this", "and", "he", "sh...", "hes", "that,", "hes", "horrible,", "hes", "a..."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["All", "the", "negativity", "hit", "me", "right", "and", "left"], ["about", "how", "he", "should", "be", "this", "and", "he", "sh..."], ["hes", "that,", "hes", "horrible,", "hes", "a..."]]}
```<end_of_turn> |
```json
{"output": [["All", "the", "negativity", "hit", "me", "right", "and", "left", "about", "how", "he", "should", "be", "this", "and", "he", "sh...", "hes", "that,", "hes", "horrible,", "hes", "a..."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["My", "boss", "at", "the", "time,", "who", "was", "the", "chief", "trial", "attorney,", "walked", "by", "my", "office", "one", "day", "and", "he", "said,", "Hillary,", "how", "do", "you", "feel", "about", "serial", "killers?", "And", "I", "said,", "Im", "against", "them.", "And", "he", "basically", "said,", "Well,", "would", "you", "like", "to", "work", "with", "me", "on", "the", "case?", "And", "I", "was", "like,", "Absolutely.", "This", "was", "the", "only", "serial", "killer", "that", "I", "can", "recall", "coming", "out", "of", "Ocean", "County."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["My", "boss", "at", "the", "time,"], ["who", "was", "the", "chief", "trial", "attorney,"], ["walked", "by", "my", "office", "one", "day", "and", "he", "said,"], ["Hillary,", "how", "do", "you", "feel", "about", "serial", "killers?"], ["And", "I", "said,", "Im", "against", "them."], ["And", "he", "basically", "said,"], ["Well,", "would", "you", "like", "to", "work", "with", "me", "'on'", "the", "case?"], ["And", "I", "was", "like,", "Absolutely."], ["This", "was", "the", "only", "serial", "killer", "that", "I", "can", "recall", "coming", "out", "of", "Ocean", "County."]]}
```<end_of_turn> |
```json
{"output": [["My", "boss", "at", "the", "time,", "who", "was", "the", "chief", "trial", "attorney,", "walked", "by", "my", "office", "one", "day", "and", "he", "said,", "Hillary,", "how", "do", "you", "feel", "about", "serial", "killers?", "And", "I", "said,", "Im", "against", "them.", "And", "he", "basically", "said,", "Well,", "would", "you", "like", "to", "work", "with", "me", "on", "the", "case?", "And", "I", "was", "like,", "Absolutely.", "This", "was", "the", "only", "serial", "killer", "that", "I", "can", "recall", "coming", "out", "of", "Ocean", "County."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["If", "you", "take", "a", "look", "back", "to", "where", "we", "come", "from", "towards", "the", "foothills,", "youll", "find", "the", "green", "trees", "as", "it", "starts", "breaking", "out", "into", "the", "prairie", "here."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["If", "you", "take", "a", "look", "back", "to", "where", "we", "come", "from", "towards", "the", "foothills,"], ["youll", "find", "the", "green", "trees", "as", "it", "starts", "breaking", "out", "into", "the", "prairie", "here."]]}
```<end_of_turn> |
```json
{"output": [["If", "you", "take", "a", "look", "back", "to", "where", "we", "come", "from", "towards", "the", "foothills,", "youll", "find", "the", "green", "trees", "as", "it", "starts", "breaking", "out", "into", "the", "prairie", "here."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["When", "Tom\u00e1s", "initiated", "the", "movement", "himself,", "yet", "we", "intervened", "with", "the", "TMS,", "Tom\u00e1s", "said,", "That", "wasnt", "me,", "I", "didnt", "initiate", "the", "movement."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["When", "\"Tom\u00e1s\"", "initiated", "the", "movement", "himself,"], ["yet", "we", "intervened", "with", "the", "TMS,"], ["\"Tom\u00e1s\"", "said,", "That", "wasnt", "me,", "I", "didnt", "initiate", "the", "movement."]]}
```<end_of_turn> |
```json
{"output": [["When", "Tom\u00e1s", "initiated", "the", "movement", "himself,", "yet", "we", "intervened", "with", "the", "TMS,", "Tom\u00e1s", "said,", "That", "wasnt", "me,", "I", "didnt", "initiate", "the", "movement."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["We", "didnt", "know,", "when", "we", "were", "called", "from", "Bethesda,", "that", "there", "was", "a", "gunshot", "wound", "to", "the", "back", "just", "to", "the", "right", "of", "the", "midline."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["We", "didnt", "know,", "when", "we", "were", "called", "from", "Bethesda,"], ["that", "there", "was", "a", "gunshot", "wound", "to", "the", "back", "just", "to", "the", "right"], ["of", "the", "midline."]]}
```<end_of_turn> |
```json
{"output": [["We", "didnt", "know,", "when", "we", "were", "called", "from", "Bethesda,", "that", "there", "was", "a", "gunshot", "wound", "to", "the", "back", "just", "to", "the", "right", "of", "the", "midline."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["the", "international", "community", "-", "they", "will", "simply", "digest", "East", "Timor", "and", "turn", "it", "into...", "theyre", "trying", "to", "turn", "it", "into", "cash", "crop."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["the", "international", "community", "-", "they", "will", "simply", "digest"], ["East", "Timor", "and", "turn", "it", "into...", "theyre", "trying", "to", "turn", "it", "into"], ["cash", "crop."]]}
```<end_of_turn> |
```json
{"output": [["the", "international", "community", "-", "they", "will", "simply", "digest", "East", "Timor", "and", "turn", "it", "into...", "theyre", "trying", "to", "turn", "it", "into", "cash", "crop."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["If", "at", "any", "point", "you", "believe", "you", "dont", "have", "any", "problems,", "if", "youre", "ignoring", "your", "problems", "or", "avoiding", "them", "or", "escaping", "them", "in", "some", "way,", "youre", "likely", "harming", "yourself", "over", "the", "long", "run."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["If", "at", "any", "point", "you", "believe", "you", "dont", "have", "any", "problems,"], ["if", "youre", "ignoring", "your", "problems", "or", "avoiding", "them", "or", "escaping", "them", "in", "some", "way,"], ["youre", "likely", "harming", "yourself", "over", "the", "long", "run."]]}
```<end_of_turn> |
```json
{"output": [["If", "at", "any", "point", "you", "believe", "you", "dont", "have", "any", "problems,", "if", "youre", "ignoring", "your", "problems", "or", "avoiding", "them", "or", "escaping", "them", "in", "some", "way,", "youre", "likely", "harming", "yourself", "over", "the", "long", "run."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Man:", "SHE\u2019S", "PROBABLY", "GONNA", "BE", "THE", "NEXT", "DANICA", "PATRICK", "IN", "NASCAR", "IF", "YOU\u2019LL", "JUST", "GIVE", "HER", "ABOUT", "5", "OR", "6", "MORE", "YEARS."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Man:", "SHE\u2019S", "PROBABLY", "GONNA", "BE", "THE", "NEXT", "DANICA", "PATRICK", "IN", "NASCAR"], ["IF", "YOU\u2019LL", "JUST", "GIVE", "HER", "ABOUT", "'5'", "OR", "'6'", "MORE", "YEARS."]]}
```<end_of_turn> |
```json
{"output": [["Man:", "SHE\u2019S", "PROBABLY", "GONNA", "BE", "THE", "NEXT", "DANICA", "PATRICK", "IN", "NASCAR", "IF", "YOU\u2019LL", "JUST", "GIVE", "HER", "ABOUT", "5", "OR", "6", "MORE", "YEARS."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["At", "Christmas-time,", "we", "let", "in", "light,", "and", "we...?", "#", "Banish", "aid?", "#", "Not", "aid.", "Aid?", "Banish", "aid?", "We", "dont", "banish", "aid."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["At", "Christmas-time,", "we", "let", "in", "light,", "and", "we...?", "#"], ["Banish", "aid?", "#", "Not", "aid.", "Aid?", "Banish", "aid?", "We", "dont", "banish", "aid."]]}
```<end_of_turn> |
```json
{"output": [["At", "Christmas-time,", "we", "let", "in", "light,", "and", "we...?", "#", "Banish", "aid?", "#", "Not", "aid.", "Aid?", "Banish", "aid?", "We", "dont", "banish", "aid."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "European", "bottlers,", "they", "hadnt", "been", "told", "they", "needed", "to", "sterilize", "the", "cork", "and", "so", "there", "were", "germs", "and", "bacteria,", "whatever", "getting", "in", "there", "and", "it", "made", "it", "almost", "poisonous."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "European", "bottlers,", "they", "hadnt", "been", "told", "they", "needed", "to", "sterilize", "the", "cork"], ["and", "so", "there", "were", "germs", "and", "bacteria,", "whatever", "getting", "in", "there", "and", "it", "made", "it", "almost", "poisonous."]]}
```<end_of_turn> |
```json
{"output": [["The", "European", "bottlers,", "they", "hadnt", "been", "told", "they", "needed", "to", "sterilize", "the", "cork", "and", "so", "there", "were", "germs", "and", "bacteria,", "whatever", "getting", "in", "there", "and", "it", "made", "it", "almost", "poisonous."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["SEB:", "I", "had", "my", "MRI", "yesterday,", "so,", "uh,", "hear", "the", "results", "maybe", "tomorrow,", "then", "Im", "taking", "some", "painkillers."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["SEB:", "I", "had", "my", "MRI", "yesterday,"], ["so,", "uh,", "hear", "the", "results", "maybe", "tomorrow,"], ["then", "Im", "taking", "some", "painkillers."]]}
```<end_of_turn> |
```json
{"output": [["SEB:", "I", "had", "my", "MRI", "yesterday,", "so,", "uh,", "hear", "the", "results", "maybe", "tomorrow,", "then", "Im", "taking", "some", "painkillers."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["We", "can", "get", "into", "more", "detail,", "but", "at", "the", "first", "level", "of", "approximation,", "theres", "two", "targets", "for", "propaganda."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["We", "can", "get", "into", "more", "detail,"], ["but", "at", "the", "first", "level", "of", "approximation,"], ["theres", "two", "targets", "for", "propaganda."]]}
```<end_of_turn> |
```json
{"output": [["We", "can", "get", "into", "more", "detail,", "but", "at", "the", "first", "level", "of", "approximation,", "theres", "two", "targets", "for", "propaganda."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Just,", "like,", "heavy", "conversations,", "in", "general,", "I", "wanna", "have", "with", "my", "family,", "like", "stuff", "I", "wanna", "talk", "about,", "like,", "my", "brothers", "and", "sisters,", "like,", "what", "the", "fuck", "are", "we", "doing?"]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Just,", "like,", "heavy", "conversations,", "in", "general,"], ["I", "wanna", "have", "with", "my", "family,"], ["like", "stuff", "I", "wanna", "talk", "about,"], ["like,", "my", "brothers", "and", "sisters,"], ["like,", "what", "the", "fuck", "are", "we", "doing?"]]}
```<end_of_turn> |
```json
{"output": [["Just,", "like,", "heavy", "conversations,", "in", "general,", "I", "wanna", "have", "with", "my", "family,", "like", "stuff", "I", "wanna", "talk", "about,", "like,", "my", "brothers", "and", "sisters,", "like,", "what", "the", "fuck", "are", "we", "doing?"]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "know", "you", "dont", "like", "to", "be", "called", "Ternera,", "but", "when", "talking", "to", "people", "who", "know", "you,", "former", "associates", "and", "former", "members", "of", "ETA,", "they", "call", "you", "that."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "know", "you", "dont", "like", "to", "be", "called", "Ternera,"], ["but", "when", "talking", "to", "people", "who", "know", "you,"], ["former", "associates", "and", "former", "members", "of", "ETA,"], ["they", "call", "you", "that."]]}
```<end_of_turn> |
```json
{"output": [["I", "know", "you", "dont", "like", "to", "be", "called", "Ternera,", "but", "when", "talking", "to", "people", "who", "know", "you,", "former", "associates", "and", "former", "members", "of", "ETA,", "they", "call", "you", "that."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "when", "you", "only", "win", "three", "out", "of", "the", "ten", "games,", "theres", "no", "way", "youre", "gonna", "have", "that", "coach", "continue", "on."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "when", "you", "only", "win", "three", "out", "of", "the", "ten", "games,"], ["theres", "'no'", "way", "youre", "gonna", "have", "that", "coach", "continue", "on."]]}
```<end_of_turn> |
```json
{"output": [["And", "when", "you", "only", "win", "three", "out", "of", "the", "ten", "games,", "theres", "no", "way", "youre", "gonna", "have", "that", "coach", "continue", "on."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "because", "his", "mother", "was", "left", "a", "worthless", "deed", "to", "the", "Colorado", "Lode,", "he", "became", "the", "richest", "man", "in", "America."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "because", "his", "mother", "was", "left", "a", "worthless", "deed", "to", "the", "Colorado", "Lode,"], ["he", "became", "the", "richest", "man", "in", "America."]]}
```<end_of_turn> |
```json
{"output": [["And", "because", "his", "mother", "was", "left", "a", "worthless", "deed", "to", "the", "Colorado", "Lode,", "he", "became", "the", "richest", "man", "in", "America."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["We", "lose", "the", "game,", "and", "then", "on", "the", "way", "from", "the", "stadium", "back", "to", "the", "hotel,", "were", "getting", "an", "email", "saying,", "Send", "us", "your", "flights", "when", "you", "wanna", "leave.", "And", "then", "the", "next", "morning,", "people", "are", "leaving."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["We", "lose", "the", "game,"], ["and", "then", "'on'", "the", "way", "from", "the", "stadium"], ["back", "to", "the", "hotel,", "were", "getting", "an", "email", "saying,"], ["Send", "us", "your", "flights", "when", "you", "wanna", "leave."], ["And", "then", "the", "next", "morning,", "people", "are", "leaving."]]}
```<end_of_turn> |
```json
{"output": [["We", "lose", "the", "game,", "and", "then", "on", "the", "way", "from", "the", "stadium", "back", "to", "the", "hotel,", "were", "getting", "an", "email", "saying,", "Send", "us", "your", "flights", "when", "you", "wanna", "leave.", "And", "then", "the", "next", "morning,", "people", "are", "leaving."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["He", "had", "gone", "from", "poster", "child", "for", "marine", "mammal", "captivity", "To", "one", "of", "its", "most", "vocal", "critics."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["He", "had", "gone", "from", "poster", "child"], ["for", "marine", "mammal", "captivity"], ["To", "one", "of", "its", "most", "vocal", "critics."]]}
```<end_of_turn> |
```json
{"output": [["He", "had", "gone", "from", "poster", "child", "for", "marine", "mammal", "captivity", "To", "one", "of", "its", "most", "vocal", "critics."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Many", "believe", "the", "mysterious", "object", "that", "caused", "city-wide", "panic", "was,", "in", "fact,", "a", "UFO,", "but", "the", "military", "vehemently", "denies", "such", "accusations."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Many", "believe", "the", "mysterious", "object"], ["that", "caused", "city-wide", "panic", "was,"], ["in", "fact,", "a", "UFO,"], ["but", "the", "military", "vehemently", "denies", "such", "accusations."]]}
```<end_of_turn> |
```json
{"output": [["Many", "believe", "the", "mysterious", "object", "that", "caused", "city-wide", "panic", "was,", "in", "fact,", "a", "UFO,", "but", "the", "military", "vehemently", "denies", "such", "accusations."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "thought", "it", "looked", "like", "makeup,", "but", "then", "some", "months", "later", "in", "the", "cutting", "room,", "the", "editor,", "Bud", "Smith,", "showed", "me", "the", "makeup", "tests", "again."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "thought", "it", "looked", "like", "makeup,"], ["but", "then", "some", "months", "later", "in", "the", "cutting", "room,"], ["the", "editor,", "Bud", "Smith,", "showed", "me", "the", "makeup", "tests", "again."]]}
```<end_of_turn> |
```json
{"output": [["I", "thought", "it", "looked", "like", "makeup,", "but", "then", "some", "months", "later", "in", "the", "cutting", "room,", "the", "editor,", "Bud", "Smith,", "showed", "me", "the", "makeup", "tests", "again."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Gil", "is", "one", "of", "the", "extremely", "rare", "cases", "of", "people", "who", "experienced", "the", "locked-in", "syndrome", "and", "came", "back", "to", "tell", "us", "about", "it."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Gil", "is", "one", "of", "the", "extremely", "rare", "cases"], ["of", "people", "who", "experienced", "the", "locked-in", "syndrome"], ["and", "came", "back", "to", "tell", "us", "about", "it."]]}
```<end_of_turn> |
```json
{"output": [["Gil", "is", "one", "of", "the", "extremely", "rare", "cases", "of", "people", "who", "experienced", "the", "locked-in", "syndrome", "and", "came", "back", "to", "tell", "us", "about", "it."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["ELVIN:", "By", "the", "time", "we", "got", "to", "the", "end", "of", "the", "weekend,", "the", "theater", "was", "full,", "and", "everybody", "was", "humming."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["'ELVIN:'", "By", "the", "time", "we", "got", "to", "the", "end"], ["of", "the", "weekend,", "the", "theater", "was", "full,"], ["and", "everybody", "was", "humming."]]}
```<end_of_turn> |
```json
{"output": [["ELVIN:", "By", "the", "time", "we", "got", "to", "the", "end", "of", "the", "weekend,", "the", "theater", "was", "full,", "and", "everybody", "was", "humming."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["\u266a", "Let", "it", "go", "\u266a", "\u266a", "Let", "it", "go", "\u266a", "\u266a", "Cant", "hold", "me", "back", "anymore", "\u266a", "Elsa", "was", "originally", "written", "as", "this", "very", "conventional", "nemesis", "character."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["\u266a", "Let", "it", "go", "\u266a"], ["\u266a", "Let", "it", "go", "\u266a"], ["\u266a", "Cant", "hold", "me", "back", "anymore", "\u266a"], ["Elsa", "was", "originally", "written", "as", "this", "very", "conventional", "nemesis", "character."]]}
```<end_of_turn> |
```json
{"output": [["\u266a", "Let", "it", "go", "\u266a", "\u266a", "Let", "it", "go", "\u266a", "\u266a", "Cant", "hold", "me", "back", "anymore", "\u266a", "Elsa", "was", "originally", "written", "as", "this", "very", "conventional", "nemesis", "character."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Having", "two", "governments", "on", "your", "back", "and,", "on", "top", "of", "that,", "all", "the", "resources", "that", "they", "have", "available..."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Having", "two", "governments", "'on'", "your", "back", "and,"], ["'on'", "top", "of", "that,", "all", "the", "resources", "that", "they", "have", "available..."]]}
```<end_of_turn> |
```json
{"output": [["Having", "two", "governments", "on", "your", "back", "and,", "on", "top", "of", "that,", "all", "the", "resources", "that", "they", "have", "available..."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Using", "such", "parameters", "as", "the", "suspects", "age,", "sex", "and", "location,", "they", "then", "winnow", "that", "list", "down", "to", "just", "20."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Using", "such", "parameters", "as", "the", "suspects", "age,", "sex", "and", "location,"], ["they", "then", "winnow", "that", "list", "down", "to", "just", "'20.'"]]}
```<end_of_turn> |
```json
{"output": [["Using", "such", "parameters", "as", "the", "suspects", "age,", "sex", "and", "location,", "they", "then", "winnow", "that", "list", "down", "to", "just", "20."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["It", "wasnt", "particularly", "something", "that", "you", "would", "choose", "to", "pattern", "your", "own", "life", "on,", "but", "it", "was", "kind", "of", "saying,", "if", "this", "is", "what", "youve", "got,", "then", "isnt", "that", "enough?", "And", "that", "was", "quite", "a", "positive", "thing."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["It", "wasnt", "particularly", "something", "that", "you", "would", "choose", "to", "pattern", "your", "own", "life", "on,"], ["but", "it", "was", "kind", "of", "saying,", "if", "this", "is", "what", "youve", "got,"], ["then", "isnt", "that", "enough?", "And", "that", "was", "quite", "a", "positive", "thing."]]}
```<end_of_turn> |
```json
{"output": [["It", "wasnt", "particularly", "something", "that", "you", "would", "choose", "to", "pattern", "your", "own", "life", "on,", "but", "it", "was", "kind", "of", "saying,", "if", "this", "is", "what", "youve", "got,", "then", "isnt", "that", "enough?", "And", "that", "was", "quite", "a", "positive", "thing."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Measuring", "these", "small", "movements,", "the", "trillions", "of", "kilometers", "away,", "one", "can", "prove", "that", "this", "planet", "exists."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Measuring", "these", "small", "movements,", "the", "trillions", "of", "kilometers", "away,"], ["one", "can", "prove", "that", "this", "planet", "exists."]]}
```<end_of_turn> |
```json
{"output": [["Measuring", "these", "small", "movements,", "the", "trillions", "of", "kilometers", "away,", "one", "can", "prove", "that", "this", "planet", "exists."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["There", "was", "a", "guy", "sitting", "in", "a", "half-collapsed", "cardboard", "box,", "and", "he", "had", "a", "plastic", "leg,", "and", "his", "plastic", "foot", "was", "pointed", "in", "the", "wrong", "direction."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["There", "was", "a", "guy", "sitting", "in", "a", "half-collapsed", "cardboard", "box,"], ["and", "he", "had", "a", "plastic", "leg,", "and", "his", "plastic", "foot"], ["was", "pointed", "in", "the", "wrong", "direction."]]}
```<end_of_turn> |
```json
{"output": [["There", "was", "a", "guy", "sitting", "in", "a", "half-collapsed", "cardboard", "box,", "and", "he", "had", "a", "plastic", "leg,", "and", "his", "plastic", "foot", "was", "pointed", "in", "the", "wrong", "direction."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "when", "you", "look", "at", "those", "old", "Ralph", "McQuarrie", "concept", "pieces,", "it", "makes", "sense", "that", "George", "would", "have", "floated", "this", "five-page", "treatment", "to", "his", "key", "collaborator", "on", "A", "New", "Hope."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "when", "you", "look", "at", "those", "old"], ["Ralph", "McQuarrie", "concept", "pieces,", "it", "makes", "sense"], ["that", "George", "would", "have", "floated", "this", "five-page"], ["treatment", "to", "his", "key", "collaborator", "'on'", "A", "New", "Hope."]]}
```<end_of_turn> |
```json
{"output": [["And", "when", "you", "look", "at", "those", "old", "Ralph", "McQuarrie", "concept", "pieces,", "it", "makes", "sense", "that", "George", "would", "have", "floated", "this", "five-page", "treatment", "to", "his", "key", "collaborator", "on", "A", "New", "Hope."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Paolos", "inclination", "towards", "these", "high-danger,", "huge-stake", "surgeries", "reminds", "me", "of", "individuals", "who", "thrill", "seek", "over", "and", "over", "again."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Paolos", "inclination", "towards", "these"], ["high-danger,", "huge-stake", "surgeries", "reminds", "me", "of"], ["individuals", "who", "thrill", "seek", "over", "and", "over", "again."]]}
```<end_of_turn> |
```json
{"output": [["Paolos", "inclination", "towards", "these", "high-danger,", "huge-stake", "surgeries", "reminds", "me", "of", "individuals", "who", "thrill", "seek", "over", "and", "over", "again."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "Im", "hanging", "out", "in", "my", "dorm", "room,", "and", "all", "of", "a", "sudden,", "the", "guy", "across", "the", "hall,", "like,", "blasts", "Billie", "Jean."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "Im", "hanging", "out", "in", "my", "dorm", "room,"], ["and", "all", "of", "a", "sudden,", "the", "guy"], ["across", "the", "hall,", "like,", "blasts", "Billie", "Jean."]]}
```<end_of_turn> |
```json
{"output": [["And", "Im", "hanging", "out", "in", "my", "dorm", "room,", "and", "all", "of", "a", "sudden,", "the", "guy", "across", "the", "hall,", "like,", "blasts", "Billie", "Jean."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["You", "know,", "if", "shit", "goes", "down,", "if", "Pence", "does", "what", "we", "think", "hes", "gonna", "do,", "then", "were", "here", "to", "defend", "this", "city."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["You", "know,", "if", "shit", "goes", "down,"], ["if", "Pence", "does", "what", "we", "think", "hes", "gonna", "do,"], ["then", "were", "here", "to", "defend", "this", "city."]]}
```<end_of_turn> |
```json
{"output": [["You", "know,", "if", "shit", "goes", "down,", "if", "Pence", "does", "what", "we", "think", "hes", "gonna", "do,", "then", "were", "here", "to", "defend", "this", "city."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["THE", "HUMANS", "FINALLY", "TASTED", "SUCCESS.", "PATTERNS", "THAT", "CHANGED", "HUMANKIND", "SUCCESS.", "PATTERNS", "THAT", "CHANGED", "HUMANKIND", "FOREVER."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["THE", "HUMANS", "FINALLY", "TASTED", "SUCCESS."], ["PATTERNS", "THAT", "CHANGED", "HUMANKIND", "SUCCESS."], ["PATTERNS", "THAT", "CHANGED", "HUMANKIND", "FOREVER."]]}
```<end_of_turn> |
```json
{"output": [["THE", "HUMANS", "FINALLY", "TASTED", "SUCCESS.", "PATTERNS", "THAT", "CHANGED", "HUMANKIND", "SUCCESS.", "PATTERNS", "THAT", "CHANGED", "HUMANKIND", "FOREVER."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "emergency", "room", "nurse,", "Doris", "Nelson,", "was", "letting", "people", "in,", "and", "there", "was", "a", "policeman", "or", "a", "Secret", "Service", "agent."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "emergency", "room", "nurse,", "Doris", "Nelson,"], ["was", "letting", "people", "in,", "and", "there", "was", "a", "policeman"], ["or", "a", "Secret", "Service", "agent."]]}
```<end_of_turn> |
```json
{"output": [["The", "emergency", "room", "nurse,", "Doris", "Nelson,", "was", "letting", "people", "in,", "and", "there", "was", "a", "policeman", "or", "a", "Secret", "Service", "agent."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "dont", "think", "theres", "anybody", "who", "has", "done", "the", "body", "of", "work", "that", "hes", "done,", "and", "its", "also", "worked", "with", "as", "many", "collaborators", "while", "maintaining", "that", "level", "of", "consistency."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "dont", "think", "theres", "anybody", "who", "has", "done", "the", "body", "of", "work"], ["that", "hes", "done,", "and", "its", "also", "worked", "with", "as", "many", "collaborators"], ["while", "maintaining", "that", "level", "of", "consistency."]]}
```<end_of_turn> |
```json
{"output": [["I", "dont", "think", "theres", "anybody", "who", "has", "done", "the", "body", "of", "work", "that", "hes", "done,", "and", "its", "also", "worked", "with", "as", "many", "collaborators", "while", "maintaining", "that", "level", "of", "consistency."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "it", "seemed", "like", "we", "were", "always", "trying", "to", "get", "somewhere", "just", "going", "from", "island", "to", "island", "to", "island", "and,", "you", "know,", "just", "staying", "long", "enough", "to", "get", "essentials."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "it", "seemed", "like", "we", "were", "always", "trying", "to", "get", "somewhere"], ["just", "going", "from", "island", "to", "island", "to", "island", "and,", "you", "know,"], ["just", "staying", "long", "enough", "to", "get", "essentials."]]}
```<end_of_turn> |
```json
{"output": [["And", "it", "seemed", "like", "we", "were", "always", "trying", "to", "get", "somewhere", "just", "going", "from", "island", "to", "island", "to", "island", "and,", "you", "know,", "just", "staying", "long", "enough", "to", "get", "essentials."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["As", "soon", "as", "walruses", "start", "to", "show", "signs", "Of", "that", "type", "of", "behaviour,", "a", "decision", "is", "made", "As", "to", "whether", "or", "not", "those", "tusks", "need", "to", "be", "removed."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["As", "soon", "as", "walruses", "start", "to", "show", "signs"], ["Of", "that", "type", "of", "behaviour,", "a", "decision", "is", "made"], ["As", "to", "whether", "or", "not", "those", "tusks", "need", "to", "be", "removed."]]}
```<end_of_turn> |
```json
{"output": [["As", "soon", "as", "walruses", "start", "to", "show", "signs", "Of", "that", "type", "of", "behaviour,", "a", "decision", "is", "made", "As", "to", "whether", "or", "not", "those", "tusks", "need", "to", "be", "removed."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["It", "creates", "a", "threat", "to", "our", "sense", "of", "security", "if", "the", "guy", "next", "door", "is", "really", "the", "evil", "that", "lives", "among", "us."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["It", "creates", "a", "threat", "to", "our", "sense", "of", "security"], ["if", "the", "guy", "next", "door", "is", "really", "the", "evil", "that", "lives", "among", "us."]]}
```<end_of_turn> |
```json
{"output": [["It", "creates", "a", "threat", "to", "our", "sense", "of", "security", "if", "the", "guy", "next", "door", "is", "really", "the", "evil", "that", "lives", "among", "us."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "point", "is", "that", "in", "a", "military", "state,", "or", "a", "feudal", "state,", "or", "what", "we", "would", "nowadays", "call", "a", "totalitarian", "state,", "it", "doesnt", "much", "matter", "what", "people", "think,", "because", "youve", "got", "a", "bludgeon", "over", "their", "head,", "and", "you", "can", "control", "what", "they", "do."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "point", "is", "that", "in", "a", "military", "state,"], ["or", "a", "feudal", "state,", "or", "what", "we", "would", "nowadays", "call", "a", "totalitarian", "state,"], ["it", "doesnt", "much", "matter", "what", "people", "think,", "because", "youve", "got", "a", "bludgeon", "over", "their", "head,"], ["and", "you", "can", "control", "what", "they", "do."]]}
```<end_of_turn> |
```json
{"output": [["The", "point", "is", "that", "in", "a", "military", "state,", "or", "a", "feudal", "state,", "or", "what", "we", "would", "nowadays", "call", "a", "totalitarian", "state,", "it", "doesnt", "much", "matter", "what", "people", "think,", "because", "youve", "got", "a", "bludgeon", "over", "their", "head,", "and", "you", "can", "control", "what", "they", "do."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["This", "bullet", "hole", "in", "his", "neck,", "we", "thought--in", "fact,", "Dr.", "Perry", "said", "to", "the", "newsmen", "right", "after", "we", "got", "out", "of", "Trauma", "Room", "1", "that", "maybe", "we", "thought", "this", "was", "an", "entrance", "wound", "here", "in", "his", "neck,", "near", "his", "windpipe,", "just", "above", "his", "clavicle."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["This", "bullet", "hole", "in", "his", "neck,", "we", "thought--in", "fact,", "Dr.", "Perry", "said", "to", "the", "newsmen", "right", "after", "we", "got", "out", "of", "Trauma", "Room", "'1'"], ["that", "maybe", "we", "thought", "this", "was", "an", "entrance", "wound", "here", "in", "his", "neck,", "near", "his", "windpipe,", "just", "above", "his", "clavicle."]]}
```<end_of_turn> |
```json
{"output": [["This", "bullet", "hole", "in", "his", "neck,", "we", "thought--in", "fact,", "Dr.", "Perry", "said", "to", "the", "newsmen", "right", "after", "we", "got", "out", "of", "Trauma", "Room", "1", "that", "maybe", "we", "thought", "this", "was", "an", "entrance", "wound", "here", "in", "his", "neck,", "near", "his", "windpipe,", "just", "above", "his", "clavicle."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "actually", "think", "it", "was", "kind", "of", "dope", "that", "people", "were,", "like,", "outside", "of", "my", "concert", "protesting,", "like,", "thats", "some", "really", "cool,", "like,", "rock", "star", "shit."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "actually", "think", "it", "was", "kind", "of", "dope"], ["that", "people", "were,", "like,", "outside", "of", "my", "concert", "protesting,", "like,"], ["thats", "some", "really", "cool,", "like,", "rock", "star", "shit."]]}
```<end_of_turn> |
```json
{"output": [["I", "actually", "think", "it", "was", "kind", "of", "dope", "that", "people", "were,", "like,", "outside", "of", "my", "concert", "protesting,", "like,", "thats", "some", "really", "cool,", "like,", "rock", "star", "shit."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["\u266a", "\u266a", "SPIKE", "LEE:", "Each", "artist", "should", "be", "allowed", "to", "pursue", "their", "artistic", "endeavor,", "but", "a", "lot", "of", "stuff", "thats", "out", "today", "is", "coonery", "buffoonery."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["\u266a", "\u266a", "SPIKE", "'LEE:'"], ["Each", "artist", "should", "be", "allowed", "to", "pursue", "their", "artistic", "endeavor,"], ["but", "a", "lot", "of", "stuff", "thats", "out", "today", "is", "coonery", "buffoonery."]]}
```<end_of_turn> |
```json
{"output": [["\u266a", "\u266a", "SPIKE", "LEE:", "Each", "artist", "should", "be", "allowed", "to", "pursue", "their", "artistic", "endeavor,", "but", "a", "lot", "of", "stuff", "thats", "out", "today", "is", "coonery", "buffoonery."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["When", "I", "was", "taking", "the", "shot,", "I", "decided", "to", "strike", "the", "ball", "differently,", "Wow,", "two", "goals", "in", "the", "semi-final,", "and", "then", "the", "third", "goal."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["When", "I", "was", "taking", "the", "shot,"], ["I", "decided", "to", "strike", "the", "ball", "differently,", "Wow,"], ["two", "goals", "in", "the", "semi-final,", "and", "then", "the", "third", "goal."]]}
```<end_of_turn> |
```json
{"output": [["When", "I", "was", "taking", "the", "shot,", "I", "decided", "to", "strike", "the", "ball", "differently,", "Wow,", "two", "goals", "in", "the", "semi-final,", "and", "then", "the", "third", "goal."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["-", "My", "name", "is", "TJ", "Hoffman", "and", "I", "classified", "myself", "on", "the", "road", "as", "tour", "manager", "and", "instrument", "technician."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["-", "My", "name", "is", "TJ", "Hoffman"], ["and", "I", "classified", "myself", "'on'", "the", "road"], ["as", "tour", "manager", "and", "instrument", "technician."]]}
```<end_of_turn> |
```json
{"output": [["-", "My", "name", "is", "TJ", "Hoffman", "and", "I", "classified", "myself", "on", "the", "road", "as", "tour", "manager", "and", "instrument", "technician."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "persistence", "and", "the", "tenacity", "and", "this", "passionate", "faith", "that", "what", "Im", "proposing", "as", "an", "entrepreneur", "is", "really", "what", "the", "world", "needs."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "persistence", "and", "the", "tenacity"], ["and", "this", "passionate", "faith", "that", "what", "Im", "proposing"], ["as", "an", "entrepreneur", "is", "really", "what", "the", "world", "needs."]]}
```<end_of_turn> |
```json
{"output": [["The", "persistence", "and", "the", "tenacity", "and", "this", "passionate", "faith", "that", "what", "Im", "proposing", "as", "an", "entrepreneur", "is", "really", "what", "the", "world", "needs."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "judge", "said", "there", "seemed", "to", "be", "a", "culture", "of", "secrecy", "and", "excessive", "confidentially", "within", "the", "Post", "Office."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "judge", "said", "there", "seemed", "to", "be", "a", "culture"], ["of", "secrecy", "and", "excessive", "confidentially", "within", "the", "Post", "Office."]]}
```<end_of_turn> |
```json
{"output": [["The", "judge", "said", "there", "seemed", "to", "be", "a", "culture", "of", "secrecy", "and", "excessive", "confidentially", "within", "the", "Post", "Office."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "then", "you", "had", "this", "sculpture", "behind", "the", "judge", "which", "was", "a", "man", "that", "had", "tucks", "his", "legs", "into", "him", "and", "was", "sitting", "like", "guilt", "hanging", "on", "him."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "then", "you", "had", "this", "sculpture", "behind", "the", "judge"], ["which", "was", "a", "man", "that", "had", "tucks", "his", "legs", "into", "him"], ["and", "was", "sitting", "like", "guilt", "hanging", "'on'", "him."]]}
```<end_of_turn> |
```json
{"output": [["And", "then", "you", "had", "this", "sculpture", "behind", "the", "judge", "which", "was", "a", "man", "that", "had", "tucks", "his", "legs", "into", "him", "and", "was", "sitting", "like", "guilt", "hanging", "on", "him."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["You", "have", "to", "be", "able", "to", "think", "about", "things", "that", "you", "cant", "see,", "but", "exist", "in", "your", "head,", "and", "for", "a", "palaeontologist,", "its", "a", "question", "of", "imagining", "such", "a", "world", "as", "weve", "never", "been", "a", "part", "of."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["You", "have", "to", "be", "able", "to", "think", "about", "things"], ["that", "you", "cant", "see,", "but", "exist", "in", "your", "head,"], ["and", "for", "a", "palaeontologist,", "its", "a", "question", "of", "imagining"], ["such", "a", "world", "as", "weve", "never", "been", "a", "part", "of."]]}
```<end_of_turn> |
```json
{"output": [["You", "have", "to", "be", "able", "to", "think", "about", "things", "that", "you", "cant", "see,", "but", "exist", "in", "your", "head,", "and", "for", "a", "palaeontologist,", "its", "a", "question", "of", "imagining", "such", "a", "world", "as", "weve", "never", "been", "a", "part", "of."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["BIG-TIME", "GAP", "NOW", "BETWEEN", "FIRST", "AND", "SECOND", "AS", "THEY", "MAKE", "THEIR", "WAY", "DOWN", "THE", "FRONT", "STRETCH", "INTO", "TURN", "ONE."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["BIG-TIME", "GAP", "NOW", "BETWEEN", "FIRST", "AND", "SECOND"], ["AS", "THEY", "MAKE", "THEIR", "WAY", "DOWN", "THE", "FRONT", "STRETCH"], ["INTO", "TURN", "ONE."]]}
```<end_of_turn> |
```json
{"output": [["BIG-TIME", "GAP", "NOW", "BETWEEN", "FIRST", "AND", "SECOND", "AS", "THEY", "MAKE", "THEIR", "WAY", "DOWN", "THE", "FRONT", "STRETCH", "INTO", "TURN", "ONE."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["As", "we", "make", "more", "and", "more", "precise", "experiments", "to", "try", "and", "discover", "and", "say", "very", "precisely", "what", "matter", "is", "made", "of,", "we", "bump", "into", "the", "language", "of", "mathematics."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["As", "we", "make", "more", "and", "more", "precise", "experiments"], ["to", "try", "and", "discover", "and", "say", "very", "precisely"], ["what", "matter", "is", "made", "of,"], ["we", "bump", "into", "the", "language", "of", "mathematics."]]}
```<end_of_turn> |
```json
{"output": [["As", "we", "make", "more", "and", "more", "precise", "experiments", "to", "try", "and", "discover", "and", "say", "very", "precisely", "what", "matter", "is", "made", "of,", "we", "bump", "into", "the", "language", "of", "mathematics."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Paolo", "explains", "Hannahs", "death", "as", "she", "was", "very", "sick,", "she", "was", "out", "of", "options,", "and", "this", "was", "her", "last", "best", "chance."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Paolo", "explains", "Hannahs", "death", "as"], ["she", "was", "very", "sick,", "she", "was", "out", "of", "options,"], ["and", "this", "was", "her", "last", "best", "chance."]]}
```<end_of_turn> |
```json
{"output": [["Paolo", "explains", "Hannahs", "death", "as", "she", "was", "very", "sick,", "she", "was", "out", "of", "options,", "and", "this", "was", "her", "last", "best", "chance."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["It", "was", "kind", "of", "like", "embroidery,", "like", "little", "stitches,", "then", "they", "all", "had", "to", "come", "together", "and", "mean", "something", "and", "form", "a", "larger", "pattern."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["It", "was", "kind", "of", "like", "embroidery,"], ["like", "little", "stitches,", "then", "they", "all", "had", "to", "come", "together"], ["and", "mean", "something", "and", "form", "a", "larger", "pattern."]]}
```<end_of_turn> |
```json
{"output": [["It", "was", "kind", "of", "like", "embroidery,", "like", "little", "stitches,", "then", "they", "all", "had", "to", "come", "together", "and", "mean", "something", "and", "form", "a", "larger", "pattern."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["ETA", "calls", "upon", "the", "governments", "of", "Spain", "and", "France", "to", "open", "a", "direct", "line", "of", "dialogue", "in", "order", "to", "put", "an", "end", "to", "the", "consequences", "of", "the", "conflict", "and", "to", "overcome", "armed", "confrontation."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["ETA", "calls", "upon", "the", "governments", "of", "Spain", "and", "France"], ["to", "open", "a", "direct", "line", "of", "dialogue"], ["in", "order", "to", "put", "an", "end", "to", "the", "consequences", "of", "the", "conflict"], ["and", "to", "overcome", "armed", "confrontation."]]}
```<end_of_turn> |
```json
{"output": [["ETA", "calls", "upon", "the", "governments", "of", "Spain", "and", "France", "to", "open", "a", "direct", "line", "of", "dialogue", "in", "order", "to", "put", "an", "end", "to", "the", "consequences", "of", "the", "conflict", "and", "to", "overcome", "armed", "confrontation."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["-", "Now", "some", "historians", "have", "suggested", "that", "Henry", "was", "furious", "at", "Anne", "for", "the", "delivery", "of", "a", "daughter", "and", "indeed", "that", "this", "was", "the", "beginning", "of", "the", "end", "of", "their", "relationship,", "but", "actually", "we", "have", "good", "evidence", "that", "Henry", "consoled", "Anne", "and", "suggested", "that", "the", "delivery", "of", "a", "healthy", "daughter", "was", "a", "good", "sign", "for", "the", "sons", "that", "would", "follow."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["-", "Now", "some", "historians", "have", "suggested", "that", "Henry", "was", "furious", "at", "Anne"], ["for", "the", "delivery", "of", "a", "daughter", "and", "indeed", "that", "this", "was", "the", "beginning", "of", "the", "end", "of", "their", "relationship,"], ["but", "actually", "we", "have", "good", "evidence", "that", "Henry", "consoled", "Anne"], ["and", "suggested", "that", "the", "delivery", "of", "a", "healthy", "daughter", "was", "a", "good", "sign"], ["for", "the", "sons", "that", "would", "follow."]]}
```<end_of_turn> |
```json
{"output": [["-", "Now", "some", "historians", "have", "suggested", "that", "Henry", "was", "furious", "at", "Anne", "for", "the", "delivery", "of", "a", "daughter", "and", "indeed", "that", "this", "was", "the", "beginning", "of", "the", "end", "of", "their", "relationship,", "but", "actually", "we", "have", "good", "evidence", "that", "Henry", "consoled", "Anne", "and", "suggested", "that", "the", "delivery", "of", "a", "healthy", "daughter", "was", "a", "good", "sign", "for", "the", "sons", "that", "would", "follow."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Still", "determined", "to", "pursue", "a", "career", "in", "medicine,", "upon", "leaving", "the", "army,", "he", "takes", "advantage", "of", "a", "new", "government", "initiative", "and", "enrolls", "at", "the", "University", "of", "Paris", "to", "become", "a", "doctor."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Still", "determined", "to", "pursue", "a", "career", "in", "medicine,"], ["upon", "leaving", "the", "army,", "he", "takes", "advantage", "of", "a", "new", "government", "initiative"], ["and", "enrolls", "at", "the", "University", "of", "Paris", "to", "become", "a", "doctor."]]}
```<end_of_turn> |
```json
{"output": [["Still", "determined", "to", "pursue", "a", "career", "in", "medicine,", "upon", "leaving", "the", "army,", "he", "takes", "advantage", "of", "a", "new", "government", "initiative", "and", "enrolls", "at", "the", "University", "of", "Paris", "to", "become", "a", "doctor."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["UH,", "HERES", "SOME", "DYE", "WE", "PREPARED", "EARLIER.", "SUBMERGE", "YOUR", "SHIRT", "IN", "THE", "DYE", "EARLIER.", "SUBMERGE", "YOUR", "SHIRT", "IN", "THE", "DYE", "AND", "HOLD", "IT", "UNDER", "WITH", "A", "STICK."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["UH,", "HERES", "SOME", "DYE", "WE", "PREPARED", "EARLIER."], ["SUBMERGE", "YOUR", "SHIRT", "IN", "THE", "DYE", "EARLIER."], ["SUBMERGE", "YOUR", "SHIRT", "IN", "THE", "DYE", "AND", "HOLD", "IT", "UNDER", "WITH", "A", "STICK."]]}
```<end_of_turn> |
```json
{"output": [["UH,", "HERES", "SOME", "DYE", "WE", "PREPARED", "EARLIER.", "SUBMERGE", "YOUR", "SHIRT", "IN", "THE", "DYE", "EARLIER.", "SUBMERGE", "YOUR", "SHIRT", "IN", "THE", "DYE", "AND", "HOLD", "IT", "UNDER", "WITH", "A", "STICK."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Dr", "Panchens", "lab", "was", "one", "of", "the", "few", "places", "in", "the", "world", "looking", "at", "the", "origin", "of", "tetrapods,", "the", "broad", "term", "used", "for", "all", "four-limbed", "vertebrates,", "from", "amphibians", "to", "mammals."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Dr", "Panchens", "lab", "was", "one", "of", "the", "few", "places", "in", "the", "world"], ["looking", "at", "the", "origin", "of", "tetrapods,", "the", "broad", "term", "used", "for", "all", "four-limbed", "vertebrates,"], ["from", "amphibians", "to", "mammals."]]}
```<end_of_turn> |
```json
{"output": [["Dr", "Panchens", "lab", "was", "one", "of", "the", "few", "places", "in", "the", "world", "looking", "at", "the", "origin", "of", "tetrapods,", "the", "broad", "term", "used", "for", "all", "four-limbed", "vertebrates,", "from", "amphibians", "to", "mammals."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["I", "think", "you", "were", "talking", "to", "Bob", "Dylan", "on", "the", "phone,", "and", "smoking", "a", "joint.", "-", "Our", "plan\u2014", "There", "was", "a", "knock", "on", "the", "door."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["I", "think", "you", "were", "talking", "to", "Bob", "Dylan", "'on'", "the", "phone,", "and", "smoking", "a", "joint."], ["'-'", "Our", "\"plan\u2014\""], ["There", "was", "a", "knock", "'on'", "the", "door."]]}
```<end_of_turn> |
```json
{"output": [["I", "think", "you", "were", "talking", "to", "Bob", "Dylan", "on", "the", "phone,", "and", "smoking", "a", "joint.", "-", "Our", "plan\u2014", "There", "was", "a", "knock", "on", "the", "door."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["Its", "very", "secretive", "now", "we", "all", "have", "to", "sign", "NDAs.", "And", "weve", "never", "signed", "NDAs", "for", "someone", "who", "wants", "to", "buy", "a", "wedding", "dress."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["Its", "very", "secretive", "now"], ["we", "all", "have", "to", "sign", "NDAs."], ["And", "weve", "never", "signed", "NDAs"], ["for", "someone", "who", "wants", "to", "buy", "a", "wedding", "dress."]]}
```<end_of_turn> |
```json
{"output": [["Its", "very", "secretive", "now", "we", "all", "have", "to", "sign", "NDAs.", "And", "weve", "never", "signed", "NDAs", "for", "someone", "who", "wants", "to", "buy", "a", "wedding", "dress."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["The", "only", "way", "to", "solve", "it", "seems", "to", "be", "to", "convince", "GM,", "Ford,", "Chrysler,", "76,", "Union,", "Shell,", "and", "Standard", "to", "go", "out", "of", "business.", "Mm-hm."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["The", "only", "way", "to", "solve", "it"], ["seems", "to", "be", "to", "convince", "GM,"], ["Ford,", "Chrysler,", "76,", "Union,", "Shell,", "and", "Standard"], ["to", "go", "out", "of", "business."], ["Mm-hm."]]}
```<end_of_turn> |
```json
{"output": [["The", "only", "way", "to", "solve", "it", "seems", "to", "be", "to", "convince", "GM,", "Ford,", "Chrysler,", "76,", "Union,", "Shell,", "and", "Standard", "to", "go", "out", "of", "business.", "Mm-hm."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["-", "Childbirth", "was", "a", "very,", "very", "common", "form", "of", "death", "for", "women", "in", "the", "period", "and", "Jane", "holds", "on", "until", "the", "24th", "of", "October,", "12", "days", "after", "the", "birth", "of", "her", "son,", "before", "sadly", "dying", "in", "the", "night."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["-", "Childbirth", "was", "a", "very,", "very", "common", "form"], ["of", "death", "for", "women", "in", "the", "period", "and"], ["Jane", "holds", "'on'", "until", "the", "24th", "of", "October,", "'12'"], ["days", "after", "the", "birth", "of", "her", "son,"], ["before", "sadly", "dying", "in", "the", "night."]]}
```<end_of_turn> |
```json
{"output": [["-", "Childbirth", "was", "a", "very,", "very", "common", "form", "of", "death", "for", "women", "in", "the", "period", "and", "Jane", "holds", "on", "until", "the", "24th", "of", "October,", "12", "days", "after", "the", "birth", "of", "her", "son,", "before", "sadly", "dying", "in", "the", "night."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["NEWSMAN:", "Its", "debut", "as", "a", "box", "office", "smash", "was", "a", "shock", "to", "critics,", "but", "not", "to", "the", "filmmaker", "and", "star", "who", "claims", "Hollywood", "has", "ignored", "the", "very", "audience", "that", "is", "flocking", "to", "his", "movie."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["NEWSMAN:", "Its", "debut", "as", "a", "box", "office", "smash"], ["was", "a", "shock", "to", "critics,", "but", "not", "to", "the", "filmmaker"], ["and", "star", "who", "claims", "Hollywood", "has", "ignored", "the", "very", "audience"], ["that", "is", "flocking", "to", "his", "movie."]]}
```<end_of_turn> |
```json
{"output": [["NEWSMAN:", "Its", "debut", "as", "a", "box", "office", "smash", "was", "a", "shock", "to", "critics,", "but", "not", "to", "the", "filmmaker", "and", "star", "who", "claims", "Hollywood", "has", "ignored", "the", "very", "audience", "that", "is", "flocking", "to", "his", "movie."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["That", "kind", "of", "reaches", "its", "zenith", "in", "in", "1992,", "at", "which", "The", "Chronic", "marks", "the", "sort", "of", "the", "end", "point", "at", "which", "hip-hop", "is", "established", "now", "as", "a", "mainstream", "pop", "music", "genre."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["That", "kind", "of", "reaches", "its", "zenith", "in", "in", "1992,"], ["at", "which", "The", "Chronic", "marks", "the", "sort", "of", "the", "end"], ["point", "at", "which", "hip-hop", "is", "established", "now", "as", "a", "mainstream"], ["pop", "music", "genre."]]}
```<end_of_turn> |
```json
{"output": [["That", "kind", "of", "reaches", "its", "zenith", "in", "in", "1992,", "at", "which", "The", "Chronic", "marks", "the", "sort", "of", "the", "end", "point", "at", "which", "hip-hop", "is", "established", "now", "as", "a", "mainstream", "pop", "music", "genre."]]}
```<end_of_turn> |
<start_of_turn>user
Segment the following sentence - sometime inside sentences - at natural breaking points for native speakers. ALWAYS provide a RFC compliant JSON response with with ```json and ``` as wrapper. ALWAYS finish the whole input, and NEVER explain yourself. ```json
{"input": ["And", "it", "really", "becomes", "a", "secret", "sauce", "decades", "later", "when", "they", "launched", "the", "Disney", "Channel,", "an", "all", "new", "Mickey", "Mouse", "Club."]}
<end_of_turn>
<start_of_turn>model |
```json
{"output": [["And", "it", "really", "becomes", "a", "secret", "sauce", "decades", "later"], ["when", "they", "launched", "the", "Disney", "Channel,", "an", "all", "new"], ["Mickey", "Mouse", "Club."]]}
```<end_of_turn> |
```json
{"output": [["And", "it", "really", "becomes", "a", "secret", "sauce", "decades", "later", "when", "they", "launched", "the", "Disney", "Channel,", "an", "all", "new", "Mickey", "Mouse", "Club."]]}
```<end_of_turn> |
End of preview. Expand
in Dataset Viewer.
- Downloads last month
- 51
Size of downloaded dataset files:
3.61 MB
Size of the auto-converted Parquet files:
3.61 MB
Number of rows:
9,181
Models trained or fine-tuned on cnbeining/sentence-segmentation-dpo-raw