v1

filelists/ljs_audio_text_val_filelist.txt

+DUMMY1/LJ022-0023.wav|The overwhelming majority of people in this country know how to sift the wheat from the chaff in what they hear and what they read.
+DUMMY1/LJ043-0030.wav|If somebody did that to me, a lousy trick like that, to take my wife away, and all the furniture, I would be mad as hell, too.
+DUMMY1/LJ005-0201.wav|as is shown by the report of the Commissioners to inquire into the state of the municipal corporations in eighteen thirty-five.
+DUMMY1/LJ001-0110.wav|Even the Caslon type when enlarged shows great shortcomings in this respect:
+DUMMY1/LJ003-0345.wav|All the committee could do in this respect was to throw the responsibility on others.
+DUMMY1/LJ007-0154.wav|These pungent and well-grounded strictures applied with still greater force to the unconvicted prisoner, the man who came to the prison innocent, and still uncontaminated,
+DUMMY1/LJ018-0098.wav|and recognized as one of the frequenters of the bogus law-stationers. His arrest led to that of others.
+DUMMY1/LJ047-0044.wav|Oswald was, however, willing to discuss his contacts with Soviet authorities. He denied having any involvement with Soviet intelligence agencies
+DUMMY1/LJ031-0038.wav|The first physician to see the President at Parkland Hospital was Dr. Charles J. Carrico, a resident in general surgery.
+DUMMY1/LJ048-0194.wav|during the morning of November twenty-two prior to the motorcade.
+DUMMY1/LJ049-0026.wav|On occasion the Secret Service has been permitted to have an agent riding in the passenger compartment with the President.
+DUMMY1/LJ004-0152.wav|although at Mr. Buxton's visit a new jail was in process of erection, the first step towards reform since Howard's visitation in seventeen seventy-four.
+DUMMY1/LJ008-0278.wav|or theirs might be one of many, and it might be considered necessary to "make an example."
+DUMMY1/LJ043-0002.wav|The Warren Commission Report. By The President's Commission on the Assassination of President Kennedy. Chapter seven. Lee Harvey Oswald:
+DUMMY1/LJ009-0114.wav|Mr. Wakefield winds up his graphic but somewhat sensational account by describing another religious service, which may appropriately be inserted here.
+DUMMY1/LJ028-0506.wav|A modern artist would have difficulty in doing such accurate work.
+DUMMY1/LJ050-0168.wav|with the particular purposes of the agency involved. The Commission recognizes that this is a controversial area
+DUMMY1/LJ039-0223.wav|Oswald's Marine training in marksmanship, his other rifle experience and his established familiarity with this particular weapon
+DUMMY1/LJ029-0032.wav|According to O'Donnell, quote, we had a motorcade wherever we went, end quote.
+DUMMY1/LJ031-0070.wav|Dr. Clark, who most closely observed the head wound,
+DUMMY1/LJ034-0198.wav|Euins, who was on the southwest corner of Elm and Houston Streets testified that he could not describe the man he saw in the window.
+DUMMY1/LJ026-0068.wav|Energy enters the plant, to a small extent,
+DUMMY1/LJ039-0075.wav|once you know that you must put the crosshairs on the target and that is all that is necessary.
+DUMMY1/LJ004-0096.wav|the fatal consequences whereof might be prevented if the justices of the peace were duly authorized
+DUMMY1/LJ005-0014.wav|Speaking on a debate on prison matters, he declared that
+DUMMY1/LJ012-0161.wav|he was reported to have fallen away to a shadow.
+DUMMY1/LJ018-0239.wav|His disappearance gave color and substance to evil reports already in circulation that the will and conveyance above referred to
+DUMMY1/LJ019-0257.wav|Here the tread-wheel was in use, there cellular cranks, or hard-labor machines.
+DUMMY1/LJ028-0008.wav|you tap gently with your heel upon the shoulder of the dromedary to urge her on.
+DUMMY1/LJ024-0083.wav|This plan of mine is no attack on the Court;
+DUMMY1/LJ042-0129.wav|No night clubs or bowling alleys, no places of recreation except the trade union dances. I have had enough.
+DUMMY1/LJ036-0103.wav|The police asked him whether he could pick out his passenger from the lineup.
+DUMMY1/LJ046-0058.wav|During his Presidency, Franklin D. Roosevelt made almost four hundred journeys and traveled more than three hundred fifty thousand miles.
+DUMMY1/LJ014-0076.wav|He was seen afterwards smoking and talking with his hosts in their back parlor, and never seen again alive.
+DUMMY1/LJ002-0043.wav|long narrow rooms -- one thirty-six feet, six twenty-three feet, and the eighth eighteen,
+DUMMY1/LJ009-0076.wav|We come to the sermon.
+DUMMY1/LJ017-0131.wav|even when the high sheriff had told him there was no possibility of a reprieve, and within a few hours of execution.
+DUMMY1/LJ046-0184.wav|but there is a system for the immediate notification of the Secret Service by the confining institution when a subject is released or escapes.
+DUMMY1/LJ014-0263.wav|When other pleasures palled he took a theatre, and posed as a munificent patron of the dramatic art.
+DUMMY1/LJ042-0096.wav|(old exchange rate) in addition to his factory salary of approximately equal amount
+DUMMY1/LJ049-0050.wav|Hill had both feet on the car and was climbing aboard to assist President and Mrs. Kennedy.
+DUMMY1/LJ019-0186.wav|seeing that since the establishment of the Central Criminal Court, Newgate received prisoners for trial from several counties,
+DUMMY1/LJ028-0307.wav|then let twenty days pass, and at the end of that time station near the Chaldasan gates a body of four thousand.
+DUMMY1/LJ012-0235.wav|While they were in a state of insensibility the murder was committed.
+DUMMY1/LJ034-0053.wav|reached the same conclusion as Latona that the prints found on the cartons were those of Lee Harvey Oswald.
+DUMMY1/LJ014-0030.wav|These were damnatory facts which well supported the prosecution.
+DUMMY1/LJ015-0203.wav|but were the precautions too minute, the vigilance too close to be eluded or overcome?
+DUMMY1/LJ028-0093.wav|but his scribe wrote it in the manner customary for the scribes of those days to write of their royal masters.
+DUMMY1/LJ002-0018.wav|The inadequacy of the jail was noticed and reported upon again and again by the grand juries of the city of London,
+DUMMY1/LJ028-0275.wav|At last, in the twentieth month,
+DUMMY1/LJ012-0042.wav|which he kept concealed in a hiding-place with a trap-door just under his bed.
+DUMMY1/LJ011-0096.wav|He married a lady also belonging to the Society of Friends, who brought him a large fortune, which, and his own money, he put into a city firm,
+DUMMY1/LJ036-0077.wav|Roger D. Craig, a deputy sheriff of Dallas County,
+DUMMY1/LJ016-0318.wav|Other officials, great lawyers, governors of prisons, and chaplains supported this view.
+DUMMY1/LJ013-0164.wav|who came from his room ready dressed, a suspicious circumstance, as he was always late in the morning.
+DUMMY1/LJ027-0141.wav|is closely reproduced in the life-history of existing deer. Or, in other words,
+DUMMY1/LJ028-0335.wav|accordingly they committed to him the command of their whole army, and put the keys of their city into his hands.
+DUMMY1/LJ031-0202.wav|Mrs. Kennedy chose the hospital in Bethesda for the autopsy because the President had served in the Navy.
+DUMMY1/LJ021-0145.wav|From those willing to join in establishing this hoped-for period of peace,
+DUMMY1/LJ016-0288.wav|"Müller, Müller, He's the man," till a diversion was created by the appearance of the gallows, which was received with continuous yells.
+DUMMY1/LJ028-0081.wav|Years later, when the archaeologists could readily distinguish the false from the true,
+DUMMY1/LJ018-0081.wav|his defense being that he had intended to commit suicide, but that, on the appearance of this officer who had wronged him,
+DUMMY1/LJ021-0066.wav|together with a great increase in the payrolls, there has come a substantial rise in the total of industrial profits
+DUMMY1/LJ009-0238.wav|After this the sheriffs sent for another rope, but the spectators interfered, and the man was carried back to jail.
+DUMMY1/LJ005-0079.wav|and improve the morals of the prisoners, and shall insure the proper measure of punishment to convicted offenders.
+DUMMY1/LJ035-0019.wav|drove to the northwest corner of Elm and Houston, and parked approximately ten feet from the traffic signal.
+DUMMY1/LJ036-0174.wav|This is the approximate time he entered the roominghouse, according to Earlene Roberts, the housekeeper there.
+DUMMY1/LJ046-0146.wav|The criteria in effect prior to November twenty-two, nineteen sixty-three, for determining whether to accept material for the PRS general files
+DUMMY1/LJ017-0044.wav|and the deepest anxiety was felt that the crime, if crime there had been, should be brought home to its perpetrator.
+DUMMY1/LJ017-0070.wav|but his sporting operations did not prosper, and he became a needy man, always driven to desperate straits for cash.
+DUMMY1/LJ014-0020.wav|He was soon afterwards arrested on suspicion, and a search of his lodgings brought to light several garments saturated with blood;
+DUMMY1/LJ016-0020.wav|He never reached the cistern, but fell back into the yard, injuring his legs severely.
+DUMMY1/LJ045-0230.wav|when he was finally apprehended in the Texas Theatre. Although it is not fully corroborated by others who were present,
+DUMMY1/LJ035-0129.wav|and she must have run down the stairs ahead of Oswald and would probably have seen or heard him.
+DUMMY1/LJ008-0307.wav|afterwards express a wish to murder the Recorder for having kept them so long in suspense.
+DUMMY1/LJ008-0294.wav|nearly indefinitely deferred.
+DUMMY1/LJ047-0148.wav|On October twenty-five,
+DUMMY1/LJ008-0111.wav|They entered a "stone cold room," and were presently joined by the prisoner.
+DUMMY1/LJ034-0042.wav|that he could only testify with certainty that the print was less than three days old.
+DUMMY1/LJ037-0234.wav|Mrs. Mary Brock, the wife of a mechanic who worked at the station, was there at the time and she saw a white male,
+DUMMY1/LJ040-0002.wav|Chapter seven. Lee Harvey Oswald: Background and Possible Motives, Part one.
+DUMMY1/LJ045-0140.wav|The arguments he used to justify his use of the alias suggest that Oswald may have come to think that the whole world was becoming involved
+DUMMY1/LJ012-0035.wav|the number and names on watches, were carefully removed or obliterated after the goods passed out of his hands.
+DUMMY1/LJ012-0250.wav|On the seventh July, eighteen thirty-seven,
+DUMMY1/LJ016-0179.wav|contracted with sheriffs and conveners to work by the job.
+DUMMY1/LJ016-0138.wav|at a distance from the prison.
+DUMMY1/LJ027-0052.wav|These principles of homology are essential to a correct interpretation of the facts of morphology.
+DUMMY1/LJ031-0134.wav|On one occasion Mrs. Johnson, accompanied by two Secret Service agents, left the room to see Mrs. Kennedy and Mrs. Connally.
+DUMMY1/LJ019-0273.wav|which Sir Joshua Jebb told the committee he considered the proper elements of penal discipline.
+DUMMY1/LJ014-0110.wav|At the first the boxes were impounded, opened, and found to contain many of O'Connor's effects.
+DUMMY1/LJ034-0160.wav|on Brennan's subsequent certain identification of Lee Harvey Oswald as the man he saw fire the rifle.
+DUMMY1/LJ038-0199.wav|eleven. If I am alive and taken prisoner,
+DUMMY1/LJ014-0010.wav|yet he could not overcome the strange fascination it had for him, and remained by the side of the corpse till the stretcher came.
+DUMMY1/LJ033-0047.wav|I noticed when I went out that the light was on, end quote,
+DUMMY1/LJ040-0027.wav|He was never satisfied with anything.
+DUMMY1/LJ048-0228.wav|and others who were present say that no agent was inebriated or acted improperly.
+DUMMY1/LJ003-0111.wav|He was in consequence put out of the protection of their internal law, end quote. Their code was a subject of some curiosity.
+DUMMY1/LJ008-0258.wav|Let me retrace my steps, and speak more in detail of the treatment of the condemned in those bloodthirsty and brutally indifferent days,
+DUMMY1/LJ029-0022.wav|The original plan called for the President to spend only one day in the State, making whirlwind visits to Dallas, Fort Worth, San Antonio, and Houston.
+DUMMY1/LJ004-0045.wav|Mr. Sturges Bourne, Sir James Mackintosh, Sir James Scarlett, and William Wilberforce.

filelists/ljs_audio_text_val_filelist.txt.cleaned

+DUMMY1/LJ022-0023.wav|ðɪ ˌoʊvɚwˈɛlmɪŋ mədʒˈɔːɹɪɾi ʌv pˈiːpəl ɪn ðɪs kˈʌntɹi nˈoʊ hˌaʊ tə sˈɪft ðə wˈiːt fɹʌmðə tʃˈæf ɪn wˌʌt ðeɪ hˈɪɹ ænd wˌʌt ðeɪ ɹˈiːd.
+DUMMY1/LJ043-0030.wav|ɪf sˈʌmbɑːdi dˈɪd ðˈæt tə mˌiː, ɐ lˈaʊsi tɹˈɪk lˈaɪk ðˈæt, tə tˈeɪk maɪ wˈaɪf ɐwˈeɪ, ænd ˈɔːl ðə fˈɜːnɪtʃɚ, ˈaɪ wʊd biː mˈæd æz hˈɛl, tˈuː.
+DUMMY1/LJ005-0201.wav|ˌæzˌɪz ʃˈoʊn baɪ ðə ɹɪpˈoːɹt ʌvðə kəmˈɪʃənɚz tʊ ɪnkwˈaɪɚɹ ˌɪntʊ ðə stˈeɪt ʌvðə mjuːnˈɪsɪpəl kˌɔːɹpɚɹˈeɪʃənz ɪn eɪtˈiːn θˈɜːɾifˈaɪv.
+DUMMY1/LJ001-0110.wav|ˈiːvən ðə kˈæslɑːn tˈaɪp wɛn ɛnlˈɑːɹdʒd ʃˈoʊz ɡɹˈeɪt ʃˈɔːɹtkʌmɪŋz ɪn ðɪs ɹɪspˈɛkt:
+DUMMY1/LJ003-0345.wav|ˈɔːl ðə kəmˈɪɾi kʊd dˈuː ɪn ðɪs ɹɪspˈɛkt wʌz tə θɹˈoʊ ðə ɹɪspˌɑːnsəbˈɪlɪɾi ˌɑːn ˈʌðɚz.
+DUMMY1/LJ007-0154.wav|ðiːz pˈʌndʒənt ænd wˈɛlɡɹˈaʊndᵻd stɹˈɪktʃɚz ɐplˈaɪd wɪð stˈɪl ɡɹˈeɪɾɚ fˈoːɹs tə ðɪ ʌnkənvˈɪktᵻd pɹˈɪzənɚ, ðə mˈæn hˌuː kˈeɪm tə ðə pɹˈɪzən ˈɪnəsənt, ænd stˈɪl ʌnkəntˈæmᵻnˌeɪɾᵻd,
+DUMMY1/LJ018-0098.wav|ænd ɹˈɛkəɡnˌaɪzd æz wˈʌn ʌvðə fɹˈiːkwɛntɚz ʌvðə bˈoʊɡəs lˈɔːstˈeɪʃənɚz. hɪz ɐɹˈɛst lˈɛd tə ðæt ʌv ˈʌðɚz.
+DUMMY1/LJ047-0044.wav|ˈɑːswəld wʌz, haʊˈɛvɚ, wˈɪlɪŋ tə dɪskˈʌs hɪz kˈɑːntækts wɪð sˈoʊviət ɐθˈɔːɹɪɾiz. hiː dɪnˈaɪd hˌævɪŋ ˌɛni ɪnvˈɑːlvmənt wɪð sˈoʊviət ɪntˈɛlɪdʒəns ˈeɪdʒənsiz
+DUMMY1/LJ031-0038.wav|ðə fˈɜːst fɪzˈɪʃən tə sˈiː ðə pɹˈɛzɪdənt æt pˈɑːɹklənd hˈɑːspɪɾəl wʌz dˈɑːktɚ tʃˈɑːɹlz dʒˈeɪ. kˈæɹɪkˌoʊ, ɐ ɹˈɛzɪdənt ɪn dʒˈɛnɚɹəl sˈɜːdʒɚɹi.
+DUMMY1/LJ048-0194.wav|dˈʊɹɪŋ ðə mˈɔːɹnɪŋ ʌv noʊvˈɛmbɚ twˈɛntitˈuː pɹˈaɪɚ tə ðə mˈoʊɾɚkˌeɪd.
+DUMMY1/LJ049-0026.wav|ˌɑːn əkˈeɪʒən ðə sˈiːkɹət sˈɜːvɪs hɐzbɪn pɚmˈɪɾᵻd tə hæv ɐn ˈeɪdʒənt ɹˈaɪdɪŋ ɪnðə pˈæsɪndʒɚ kəmpˈɑːɹtmənt wɪððə pɹˈɛzɪdənt.
+DUMMY1/LJ004-0152.wav|ɑːlðˈoʊ æt mˈɪstɚ bˈʌkstənz vˈɪzɪt ɐ nˈuː dʒˈeɪl wʌz ɪn pɹˈɑːsɛs ʌv ɪɹˈɛkʃən, ðə fˈɜːst stˈɛp tʊwˈɔːɹdz ɹɪfˈɔːɹm sˈɪns hˈaʊɚdz vˌɪzɪtˈeɪʃən ɪn sˌɛvəntˈiːn sˈɛvəntifˈoːɹ.
+DUMMY1/LJ008-0278.wav|ɔːɹ ðˈɛɹz mˌaɪt biː wˈʌn ʌv mˈɛni, ænd ɪt mˌaɪt biː kənsˈɪdɚd nˈɛsəsɚɹi tuː "mˌeɪk ɐn ɛɡzˈæmpəl."
+DUMMY1/LJ043-0002.wav|ðə wˈɔːɹən kəmˈɪʃən ɹɪpˈoːɹt. baɪ ðə pɹˈɛzɪdənts kəmˈɪʃən ɑːnðɪ ɐsˌæsᵻnˈeɪʃən ʌv pɹˈɛzɪdənt kˈɛnədi. tʃˈæptɚ sˈɛvən. lˈiː hˈɑːɹvi ˈɑːswəld:
+DUMMY1/LJ009-0114.wav|mˈɪstɚ wˈeɪkfiːld wˈaɪndz ˈʌp hɪz ɡɹˈæfɪk bˌʌt sˈʌmwʌt sɛnsˈeɪʃənəl ɐkˈaʊnt baɪ dɪskɹˈaɪbɪŋ ɐnˈʌðɚ ɹɪlˈɪdʒəs sˈɜːvɪs, wˌɪtʃ mˈeɪ ɐpɹˈoʊpɹɪətli biː ɪnsˈɜːɾᵻd hˈɪɹ.
+DUMMY1/LJ028-0506.wav|ɐ mˈɑːdɚn ˈɑːɹɾɪst wʊdhɐv dˈɪfɪkˌʌlti ɪn dˌuːɪŋ sˈʌtʃ ˈækjʊɹət wˈɜːk.
+DUMMY1/LJ050-0168.wav|wɪððə pɚtˈɪkjʊlɚ pˈɜːpəsᵻz ʌvðɪ ˈeɪdʒənsi ɪnvˈɑːlvd. ðə kəmˈɪʃən ɹˈɛkəɡnˌaɪzɪz ðæt ðɪs ɪz ɐ kˌɑːntɹəvˈɜːʃəl ˈɛɹiə
+DUMMY1/LJ039-0223.wav|ˈɑːswəldz mɚɹˈiːn tɹˈeɪnɪŋ ɪn mˈɑːɹksmənʃˌɪp, hɪz ˈʌðɚ ɹˈaɪfəl ɛkspˈiəɹɪəns ænd hɪz ɪstˈæblɪʃt fəmˌɪlɪˈæɹɪɾi wɪð ðɪs pɚtˈɪkjʊlɚ wˈɛpən
+DUMMY1/LJ029-0032.wav|ɐkˈoːɹdɪŋ tʊ oʊdˈɑːnəl, kwˈoʊt, wiː hɐd ɐ mˈoʊɾɚkˌeɪd wɛɹɹˈɛvɚ wiː wˈɛnt, ˈɛnd kwˈoʊt.
+DUMMY1/LJ031-0070.wav|dˈɑːktɚ klˈɑːɹk, hˌuː mˈoʊst klˈoʊsli ɑːbzˈɜːvd ðə hˈɛd wˈuːnd,
+DUMMY1/LJ034-0198.wav|jˈuːɪnz, hˌuː wʌz ɑːnðə saʊθwˈɛst kˈɔːɹnɚɹ ʌv ˈɛlm ænd hjˈuːstən stɹˈiːts tˈɛstɪfˌaɪd ðæt hiː kʊd nˌɑːt dɪskɹˈaɪb ðə mˈæn hiː sˈɔː ɪnðə wˈɪndoʊ.
+DUMMY1/LJ026-0068.wav|ˈɛnɚdʒi ˈɛntɚz ðə plˈænt, tʊ ɐ smˈɔːl ɛkstˈɛnt,
+DUMMY1/LJ039-0075.wav|wˈʌns juː nˈoʊ ðæt juː mˈʌst pˌʊt ðə kɹˈɔshɛɹz ɑːnðə tˈɑːɹɡɪt ænd ðæt ɪz ˈɔːl ðæt ɪz nˈɛsəsɚɹi.
+DUMMY1/LJ004-0096.wav|ðə fˈeɪɾəl kˈɑːnsɪkwənsᵻz wˈɛɹɑːf mˌaɪt biː pɹɪvˈɛntᵻd ɪf ðə dʒˈʌstɪsᵻz ʌvðə pˈiːs wɜː djˈuːli ˈɔːθɚɹˌaɪzd
+DUMMY1/LJ005-0014.wav|spˈiːkɪŋ ˌɑːn ɐ dɪbˈeɪt ˌɑːn pɹˈɪzən mˈæɾɚz, hiː dᵻklˈɛɹd ðˈæt
+DUMMY1/LJ012-0161.wav|hiː wʌz ɹɪpˈoːɹɾᵻd tə hæv fˈɔːlən ɐwˈeɪ tʊ ɐ ʃˈædoʊ.
+DUMMY1/LJ018-0239.wav|hɪz dˌɪsɐpˈɪɹəns ɡˈeɪv kˈʌlɚ ænd sˈʌbstəns tʊ ˈiːvəl ɹɪpˈoːɹts ɔːlɹˌɛdi ɪn sˌɜːkjʊlˈeɪʃən ðætðə wɪl ænd kənvˈeɪəns əbˌʌv ɹɪfˈɜːd tuː
+DUMMY1/LJ019-0257.wav|hˈɪɹ ðə tɹˈɛdwˈiːl wʌz ɪn jˈuːs, ðɛɹ sˈɛljʊlɚ kɹˈæŋks, ɔːɹ hˈɑːɹdlˈeɪbɚ məʃˈiːnz.
+DUMMY1/LJ028-0008.wav|juː tˈæp dʒˈɛntli wɪð jʊɹ hˈiːl əpˌɑːn ðə ʃˈoʊldɚɹ ʌvðə dɹˈoʊmdɚɹi tʊ ˈɜːdʒ hɜːɹ ˈɑːn.
+DUMMY1/LJ024-0083.wav|ðɪs plˈæn ʌv mˈaɪn ɪz nˈoʊ ɐtˈæk ɑːnðə kˈoːɹt;
+DUMMY1/LJ042-0129.wav|nˈoʊ nˈaɪt klˈʌbz ɔːɹ bˈoʊlɪŋ ˈælɪz, nˈoʊ plˈeɪsᵻz ʌv ɹˌɛkɹiːˈeɪʃən ɛksˈɛpt ðə tɹˈeɪd jˈuːniən dˈænsᵻz. ˈaɪ hæv hɐd ɪnˈʌf.
+DUMMY1/LJ036-0103.wav|ðə pəlˈiːs ˈæskt hˌɪm wˈɛðɚ hiː kʊd pˈɪk ˈaʊt hɪz pˈæsɪndʒɚ fɹʌmðə lˈaɪnʌp.
+DUMMY1/LJ046-0058.wav|dˈʊɹɪŋ hɪz pɹˈɛzɪdənsi, fɹˈæŋklɪn dˈiː. ɹˈoʊzəvˌɛlt mˌeɪd ˈɔːlmoʊst fˈoːɹ hˈʌndɹəd dʒˈɜːnɪz ænd tɹˈævəld mˈoːɹ ðɐn θɹˈiː hˈʌndɹəd fˈɪfti θˈaʊzənd mˈaɪlz.
+DUMMY1/LJ014-0076.wav|hiː wʌz sˈiːn ˈæftɚwɚdz smˈoʊkɪŋ ænd tˈɔːkɪŋ wɪð hɪz hˈoʊsts ɪn ðɛɹ bˈæk pˈɑːɹlɚ, ænd nˈɛvɚ sˈiːn ɐɡˈɛn ɐlˈaɪv.
+DUMMY1/LJ002-0043.wav|lˈɑːŋ nˈæɹoʊ ɹˈuːmz wˈʌn θˈɜːɾisˈɪks fˈiːt, sˈɪks twˈɛntiθɹˈiː fˈiːt, ænd ðɪ ˈeɪtθ eɪtˈiːn,
+DUMMY1/LJ009-0076.wav|wiː kˈʌm tə ðə sˈɜːmən.
+DUMMY1/LJ017-0131.wav|ˈiːvən wɛn ðə hˈaɪ ʃˈɛɹɪf hɐd tˈoʊld hˌɪm ðɛɹwˌʌz nˈoʊ pˌɑːsəbˈɪlɪɾi əvɚ ɹɪpɹˈiːv, ænd wɪðˌɪn ɐ fjˈuː ˈaɪʊɹz ʌv ˌɛksɪkjˈuːʃən.
+DUMMY1/LJ046-0184.wav|bˌʌt ðɛɹ ɪz ɐ sˈɪstəm fɚðɪ ɪmˈiːdɪət nˌoʊɾɪfɪkˈeɪʃən ʌvðə sˈiːkɹət sˈɜːvɪs baɪ ðə kənfˈaɪnɪŋ ˌɪnstɪtˈuːʃən wɛn ɐ sˈʌbdʒɛkt ɪz ɹɪlˈiːsd ɔːɹ ɛskˈeɪps.
+DUMMY1/LJ014-0263.wav|wˌɛn ˈʌðɚ plˈɛʒɚz pˈɔːld hiː tˈʊk ɐ θˈiəɾɚ, ænd pˈoʊzd æz ɐ mjuːnˈɪfɪsənt pˈeɪtɹən ʌvðə dɹəmˈæɾɪk ˈɑːɹt.
+DUMMY1/LJ042-0096.wav| ˈoʊld ɛkstʃˈeɪndʒ ɹˈeɪt ɪn ɐdˈɪʃən tə hɪz fˈæktɚɹi sˈælɚɹi ʌv ɐpɹˈɑːksɪmətli ˈiːkwəl ɐmˈaʊnt
+DUMMY1/LJ049-0050.wav|hˈɪl hɐd bˈoʊθ fˈiːt ɑːnðə kˈɑːɹ ænd wʌz klˈaɪmɪŋ ɐbˈoːɹd tʊ ɐsˈɪst pɹˈɛzɪdənt ænd mɪsˈɛs kˈɛnədi.
+DUMMY1/LJ019-0186.wav|sˈiːɪŋ ðæt sˈɪns ðɪ ɪstˈæblɪʃmənt ʌvðə sˈɛntɹəl kɹˈɪmɪnəl kˈoːɹt, nˈuːɡeɪt ɹɪsˈiːvd pɹˈɪzənɚz fɔːɹ tɹˈaɪəl fɹʌm sˈɛvɹəl kˈaʊntɪz,
+DUMMY1/LJ028-0307.wav|ðˈɛn lˈɛt twˈɛnti dˈeɪz pˈæs, ænd æt ðɪ ˈɛnd ʌv ðæt tˈaɪm stˈeɪʃən nˌɪɹ ðə tʃˈældæsən ɡˈeɪts ɐ bˈɑːdi ʌv fˈoːɹ θˈaʊzənd.
+DUMMY1/LJ012-0235.wav|wˌaɪl ðeɪ wɜːɹ ɪn ɐ stˈeɪt ʌv ɪnsˌɛnsəbˈɪlɪɾi ðə mˈɜːdɚ wʌz kəmˈɪɾᵻd.
+DUMMY1/LJ034-0053.wav|ɹˈiːtʃt ðə sˈeɪm kənklˈuːʒən æz lætˈoʊnə ðætðə pɹˈɪnts fˈaʊnd ɑːnðə kˈɑːɹtənz wɜː ðoʊz ʌv lˈiː hˈɑːɹvi ˈɑːswəld.
+DUMMY1/LJ014-0030.wav|ðiːz wɜː dˈæmnətˌoːɹi fˈækts wˌɪtʃ wˈɛl səpˈoːɹɾᵻd ðə pɹˌɑːsɪkjˈuːʃən.
+DUMMY1/LJ015-0203.wav|bˌʌt wɜː ðə pɹɪkˈɔːʃənz tˈuː mˈɪnɪt, ðə vˈɪdʒɪləns tˈuː klˈoʊs təbi ɪlˈuːdᵻd ɔːɹ ˌoʊvɚkˈʌm?
+DUMMY1/LJ028-0093.wav|bˌʌt hɪz skɹˈaɪb ɹˈoʊt ɪt ɪnðə mˈænɚ kˈʌstəmˌɛɹi fɚðə skɹˈaɪbz ʌv ðoʊz dˈeɪz tə ɹˈaɪt ʌv ðɛɹ ɹˈɔɪəl mˈæstɚz.
+DUMMY1/LJ002-0018.wav|ðɪ ɪnˈædɪkwəsi ʌvðə dʒˈeɪl wʌz nˈoʊɾɪsd ænd ɹɪpˈoːɹɾᵻd əpˌɑːn ɐɡˈɛn ænd ɐɡˈɛn baɪ ðə ɡɹˈænd dʒˈʊɹɪz ʌvðə sˈɪɾi ʌv lˈʌndən,
+DUMMY1/LJ028-0275.wav|æt lˈæst, ɪnðə twˈɛntiəθ mˈʌnθ,
+DUMMY1/LJ012-0042.wav|wˌɪtʃ hiː kˈɛpt kənsˈiːld ɪn ɐ hˈaɪdɪŋplˈeɪs wɪð ɐ tɹˈæpdˈoːɹ dʒˈʌst ˌʌndɚ hɪz bˈɛd.
+DUMMY1/LJ011-0096.wav|hiː mˈæɹɪd ɐ lˈeɪdi ˈɑːlsoʊ bɪlˈɑːŋɪŋ tə ðə səsˈaɪəɾi ʌv fɹˈɛndz, hˌuː bɹˈɔːt hˌɪm ɐ lˈɑːɹdʒ fˈɔːɹtʃən, wˈɪtʃ, ænd hɪz ˈoʊn mˈʌni, hiː pˌʊt ˌɪntʊ ɐ sˈɪɾi fˈɜːm,
+DUMMY1/LJ036-0077.wav|ɹˈɑːdʒɚ dˈiː. kɹˈeɪɡ, ɐ dˈɛpjuːɾi ʃˈɛɹɪf ʌv dˈæləs kˈaʊnti,
+DUMMY1/LJ016-0318.wav|ˈʌðɚɹ əfˈɪʃəlz, ɡɹˈeɪt lˈɔɪɚz, ɡˈʌvɚnɚz ʌv pɹˈɪzənz, ænd tʃˈæplɪnz səpˈoːɹɾᵻd ðɪs vjˈuː.
+DUMMY1/LJ013-0164.wav|hˌuː kˈeɪm fɹʌm hɪz ɹˈuːm ɹˈɛdi dɹˈɛst, ɐ səspˈɪʃəs sˈɜːkəmstˌæns, æz hiː wʌz ˈɔːlweɪz lˈeɪt ɪnðə mˈɔːɹnɪŋ.
+DUMMY1/LJ027-0141.wav|ɪz klˈoʊsli ɹɪpɹədˈuːst ɪnðə lˈaɪfhˈɪstɚɹi ʌv ɛɡzˈɪstɪŋ dˈɪɹ. ˈɔːɹ, ɪn ˈʌðɚ wˈɜːdz,
+DUMMY1/LJ028-0335.wav|ɐkˈoːɹdɪŋli ðeɪ kəmˈɪɾᵻd tə hˌɪm ðə kəmˈænd ʌv ðɛɹ hˈoʊl ˈɑːɹmi, ænd pˌʊt ðə kˈiːz ʌv ðɛɹ sˈɪɾi ˌɪntʊ hɪz hˈændz.
+DUMMY1/LJ031-0202.wav|mɪsˈɛs kˈɛnədi tʃˈoʊz ðə hˈɑːspɪɾəl ɪn bəθˈɛzdə fɚðɪ ˈɔːtɑːpsi bɪkˈʌz ðə pɹˈɛzɪdənt hɐd sˈɜːvd ɪnðə nˈeɪvi.
+DUMMY1/LJ021-0145.wav|fɹʌm ðoʊz wˈɪlɪŋ tə dʒˈɔɪn ɪn ɪstˈæblɪʃɪŋ ðɪs hˈoʊptfɔːɹ pˈiəɹɪəd ʌv pˈiːs,
+DUMMY1/LJ016-0288.wav|"mˈʌlɚ, mˈʌlɚ, hiːz ðə mˈæn," tˈɪl ɐ daɪvˈɜːʒən wʌz kɹiːˈeɪɾᵻd baɪ ðɪ ɐpˈɪɹəns ʌvðə ɡˈæloʊz, wˌɪtʃ wʌz ɹɪsˈiːvd wɪð kəntˈɪnjuːəs jˈɛlz.
+DUMMY1/LJ028-0081.wav|jˈɪɹz lˈeɪɾɚ, wˌɛn ðɪ ˌɑːɹkiːˈɑːlədʒˌɪsts kʊd ɹˈɛdɪli dɪstˈɪŋɡwɪʃ ðə fˈɑːls fɹʌmðə tɹˈuː,
+DUMMY1/LJ018-0081.wav|hɪz dɪfˈɛns bˌiːɪŋ ðæt hiː hɐd ɪntˈɛndᵻd tə kəmˈɪt sˈuːɪsˌaɪd, bˌʌt ðˈæt, ɑːnðɪ ɐpˈɪɹəns ʌv ðɪs ˈɑːfɪsɚ hˌuː hɐd ɹˈɔŋd hˌɪm,
+DUMMY1/LJ021-0066.wav|təɡˌɛðɚ wɪð ɐ ɡɹˈeɪt ˈɪnkɹiːs ɪnðə pˈeɪɹoʊlz, ðɛɹ hɐz kˈʌm ɐ səbstˈænʃəl ɹˈaɪz ɪnðə tˈoʊɾəl ʌv ɪndˈʌstɹɪəl pɹˈɑːfɪts
+DUMMY1/LJ009-0238.wav|ˈæftɚ ðɪs ðə ʃˈɛɹɪfs sˈɛnt fɔːɹ ɐnˈʌðɚ ɹˈoʊp, bˌʌt ðə spɛktˈeɪɾɚz ˌɪntəfˈɪɹd, ænd ðə mˈæn wʌz kˈæɹɪd bˈæk tə dʒˈeɪl.
+DUMMY1/LJ005-0079.wav|ænd ɪmpɹˈuːv ðə mˈɔːɹəlz ʌvðə pɹˈɪzənɚz, ænd ʃˌæl ɪnʃˈʊɹ ðə pɹˈɑːpɚ mˈɛʒɚɹ ʌv pˈʌnɪʃmənt tə kənvˈɪktᵻd əfˈɛndɚz.
+DUMMY1/LJ035-0019.wav|dɹˈoʊv tə ðə nɔːɹθwˈɛst kˈɔːɹnɚɹ ʌv ˈɛlm ænd hjˈuːstən, ænd pˈɑːɹkt ɐpɹˈɑːksɪmətli tˈɛn fˈiːt fɹʌmðə tɹˈæfɪk sˈɪɡnəl.
+DUMMY1/LJ036-0174.wav|ðɪs ɪz ðɪ ɐpɹˈɑːksɪmət tˈaɪm hiː ˈɛntɚd ðə ɹˈuːmɪŋhˌaʊs, ɐkˈoːɹdɪŋ tʊ ˈɜːliːn ɹˈɑːbɚts, ðə hˈaʊskiːpɚ ðˈɛɹ.
+DUMMY1/LJ046-0146.wav|ðə kɹaɪtˈiəɹɪə ɪn ɪfˈɛkt pɹˈaɪɚ tə noʊvˈɛmbɚ twˈɛntitˈuː, naɪntˈiːn sˈɪkstiθɹˈiː, fɔːɹ dɪtˈɜːmɪnɪŋ wˈɛðɚ tʊ ɐksˈɛpt mətˈiəɹɪəl fɚðə pˌiːˌɑːɹˈɛs dʒˈɛnɚɹəl fˈaɪlz
+DUMMY1/LJ017-0044.wav|ænd ðə dˈiːpəst æŋzˈaɪəɾi wʌz fˈɛlt ðætðə kɹˈaɪm, ɪf kɹˈaɪm ðˈɛɹ hɐdbɪn, ʃˌʊd biː bɹˈɔːt hˈoʊm tʊ ɪts pˈɜːpɪtɹˌeɪɾɚ.
+DUMMY1/LJ017-0070.wav|bˌʌt hɪz spˈoːɹɾɪŋ ˌɑːpɚɹˈeɪʃənz dɪdnˌɑːt pɹˈɑːspɚ, ænd hiː bɪkˌeɪm ɐ nˈiːdi mˈæn, ˈɔːlweɪz dɹˈɪvən tə dˈɛspɚɹət stɹˈeɪts fɔːɹ kˈæʃ.
+DUMMY1/LJ014-0020.wav|hiː wʌz sˈuːn ˈæftɚwɚdz ɐɹˈɛstᵻd ˌɑːn səspˈɪʃən, ænd ɐ sˈɜːtʃ ʌv hɪz lˈɑːdʒɪŋz bɹˈɔːt tə lˈaɪt sˈɛvɹəl ɡˈɑːɹmənts sˈætʃɚɹˌeɪɾᵻd wɪð blˈʌd;
+DUMMY1/LJ016-0020.wav|hiː nˈɛvɚ ɹˈiːtʃt ðə sˈɪstɚn, bˌʌt fˈɛl bˈæk ˌɪntʊ ðə jˈɑːɹd, ˈɪndʒɚɹɪŋ hɪz lˈɛɡz sɪvˈɪɹli.
+DUMMY1/LJ045-0230.wav|wˌɛn hiː wʌz fˈaɪnəli ˌæpɹɪhˈɛndᵻd ɪnðə tˈɛksəs θˈiəɾɚ. ɑːlðˈoʊ ɪt ɪz nˌɑːt fˈʊli kɚɹˈɑːbɚɹˌeɪɾᵻd baɪ ˈʌðɚz hˌuː wɜː pɹˈɛzənt,
+DUMMY1/LJ035-0129.wav|ænd ʃiː mˈʌstɐv ɹˈʌn dˌaʊn ðə stˈɛɹz ɐhˈɛd ʌv ˈɑːswəld ænd wʊd pɹˈɑːbəbli hæv sˈiːn ɔːɹ hˈɜːd hˌɪm.
+DUMMY1/LJ008-0307.wav|ˈæftɚwɚdz ɛkspɹˈɛs ɐ wˈɪʃ tə mˈɜːdɚ ðə ɹɪkˈoːɹdɚ fɔːɹ hˌævɪŋ kˈɛpt ðˌɛm sˌoʊ lˈɑːŋ ɪn səspˈɛns.
+DUMMY1/LJ008-0294.wav|nˌɪɹli ɪndˈɛfɪnətli dɪfˈɜːd.
+DUMMY1/LJ047-0148.wav|ˌɑːn ɑːktˈoʊbɚ twˈɛntifˈaɪv,
+DUMMY1/LJ008-0111.wav|ðeɪ ˈɛntɚd ˈeɪ "stˈoʊn kˈoʊld ɹˈuːm," ænd wɜː pɹˈɛzəntli dʒˈɔɪnd baɪ ðə pɹˈɪzənɚ.
+DUMMY1/LJ034-0042.wav|ðæt hiː kʊd ˈoʊnli tˈɛstɪfˌaɪ wɪð sˈɜːtənti ðætðə pɹˈɪnt wʌz lˈɛs ðɐn θɹˈiː dˈeɪz ˈoʊld.
+DUMMY1/LJ037-0234.wav|mɪsˈɛs mˈɛɹi bɹˈɑːk, ðə wˈaɪf əvə mɪkˈænɪk hˌuː wˈɜːkt æt ðə stˈeɪʃən, wʌz ðɛɹ æt ðə tˈaɪm ænd ʃiː sˈɔː ɐ wˈaɪt mˈeɪl,
+DUMMY1/LJ040-0002.wav|tʃˈæptɚ sˈɛvən. lˈiː hˈɑːɹvi ˈɑːswəld: bˈækɡɹaʊnd ænd pˈɑːsəbəl mˈoʊɾɪvz, pˈɑːɹt wˌʌn.
+DUMMY1/LJ045-0140.wav|ðɪ ˈɑːɹɡjuːmənts hiː jˈuːzd tə dʒˈʌstɪfˌaɪ hɪz jˈuːs ʌvðɪ ˈeɪliəs sədʒˈɛst ðæt ˈɑːswəld mˌeɪhɐv kˈʌm tə θˈɪŋk ðætðə hˈoʊl wˈɜːld wʌz bɪkˈʌmɪŋ ɪnvˈɑːlvd
+DUMMY1/LJ012-0035.wav|ðə nˈʌmbɚ ænd nˈeɪmz ˌɑːn wˈɑːtʃᵻz, wɜː kˈɛɹfəli ɹɪmˈuːvd ɔːɹ əblˈɪɾɚɹˌeɪɾᵻd ˈæftɚ ðə ɡˈʊdz pˈæst ˌaʊɾəv hɪz hˈændz.
+DUMMY1/LJ012-0250.wav|ɑːnðə sˈɛvənθ dʒuːlˈaɪ, eɪtˈiːn θˈɜːɾisˈɛvən,
+DUMMY1/LJ016-0179.wav|kəntɹˈæktᵻd wɪð ʃˈɛɹɪfs ænd kənvˈɛnɚz tə wˈɜːk baɪ ðə dʒˈɑːb.
+DUMMY1/LJ016-0138.wav|æɾə dˈɪstəns fɹʌmðə pɹˈɪzən.
+DUMMY1/LJ027-0052.wav|ðiːz pɹˈɪnsɪpəlz ʌv həmˈɑːlədʒi ɑːɹ ɪsˈɛnʃəl tʊ ɐ kɚɹˈɛkt ɪntˌɜːpɹɪtˈeɪʃən ʌvðə fˈækts ʌv mɔːɹfˈɑːlədʒi.
+DUMMY1/LJ031-0134.wav|ˌɑːn wˈʌn əkˈeɪʒən mɪsˈɛs dʒˈɑːnsən, ɐkˈʌmpənɪd baɪ tˈuː sˈiːkɹət sˈɜːvɪs ˈeɪdʒənts, lˈɛft ðə ɹˈuːm tə sˈiː mɪsˈɛs kˈɛnədi ænd mɪsˈɛs kənˈæli.
+DUMMY1/LJ019-0273.wav|wˌɪtʃ sˌɜː dʒˈɑːʃjuːə dʒˈɛb tˈoʊld ðə kəmˈɪɾi hiː kənsˈɪdɚd ðə pɹˈɑːpɚɹ ˈɛlɪmənts ʌv pˈiːnəl dˈɪsɪplˌɪn.
+DUMMY1/LJ014-0110.wav|æt ðə fˈɜːst ðə bˈɑːksᵻz wɜːɹ ɪmpˈaʊndᵻd, ˈoʊpənd, ænd fˈaʊnd tə kəntˈeɪn mˈɛnɪəv oʊkˈɑːnɚz ɪfˈɛkts.
+DUMMY1/LJ034-0160.wav|ˌɑːn bɹˈɛnənz sˈʌbsɪkwənt sˈɜːtən aɪdˈɛntɪfɪkˈeɪʃən ʌv lˈiː hˈɑːɹvi ˈɑːswəld æz ðə mˈæn hiː sˈɔː fˈaɪɚ ðə ɹˈaɪfəl.
+DUMMY1/LJ038-0199.wav|ɪlˈɛvən. ɪf ˈaɪ æm ɐlˈaɪv ænd tˈeɪkən pɹˈɪzənɚ,
+DUMMY1/LJ014-0010.wav|jˈɛt hiː kʊd nˌɑːt ˌoʊvɚkˈʌm ðə stɹˈeɪndʒ fˌæsᵻnˈeɪʃən ɪt hˈɐd fɔːɹ hˌɪm, ænd ɹɪmˈeɪnd baɪ ðə sˈaɪd ʌvðə kˈɔːɹps tˈɪl ðə stɹˈɛtʃɚ kˈeɪm.
+DUMMY1/LJ033-0047.wav|ˈaɪ nˈoʊɾɪsd wɛn ˈaɪ wɛnt ˈaʊt ðætðə lˈaɪt wʌz ˈɑːn, ˈɛnd kwˈoʊt,
+DUMMY1/LJ040-0027.wav|hiː wʌz nˈɛvɚ sˈæɾɪsfˌaɪd wɪð ˈɛnɪθˌɪŋ.
+DUMMY1/LJ048-0228.wav|ænd ˈʌðɚz hˌuː wɜː pɹˈɛzənt sˈeɪ ðæt nˈoʊ ˈeɪdʒənt wʌz ɪnˈiːbɹɪˌeɪɾᵻd ɔːɹ ˈæktᵻd ɪmpɹˈɑːpɚli.
+DUMMY1/LJ003-0111.wav|hiː wʌz ɪn kˈɑːnsɪkwəns pˌʊt ˌaʊɾəv ðə pɹətˈɛkʃən ʌv ðɛɹ ɪntˈɜːnəl lˈɔː, ˈɛnd kwˈoʊt. ðɛɹ kˈoʊd wʌzɐ sˈʌbdʒɛkt ʌv sˌʌm kjˌʊɹɪˈɑːsɪɾi.
+DUMMY1/LJ008-0258.wav|lˈɛt mˌiː ɹɪtɹˈeɪs maɪ stˈɛps, ænd spˈiːk mˈoːɹ ɪn diːtˈeɪl ʌvðə tɹˈiːtmənt ʌvðə kəndˈɛmd ɪn ðoʊz blˈʌdθɜːsti ænd bɹˈuːɾəli ɪndˈɪfɹənt dˈeɪz,
+DUMMY1/LJ029-0022.wav|ðɪ ɚɹˈɪdʒɪnəl plˈæn kˈɔːld fɚðə pɹˈɛzɪdənt tə spˈɛnd ˈoʊnli wˈʌn dˈeɪ ɪnðə stˈeɪt, mˌeɪkɪŋ wˈɜːlwɪnd vˈɪzɪts tə dˈæləs, fˈɔːɹt wˈɜːθ, sˌæn æntˈoʊnɪˌoʊ, ænd hjˈuːstən.
+DUMMY1/LJ004-0045.wav|mˈɪstɚ stˈɜːdʒᵻz bˈoːɹn, sˌɜː dʒˈeɪmz mˈækɪntˌɑːʃ, sˌɜː dʒˈeɪmz skˈɑːɹlɪt, ænd wˈɪljəm wˈɪlbɚfˌoːɹs.

filelists/vctk_audio_sid_text_val_filelist.txt

+DUMMY2/p364/p364_240.wav|88|It had happened to him.
+DUMMY2/p280/p280_148.wav|52|It is open season on the Old Firm.
+DUMMY2/p231/p231_320.wav|50|However, he is a coach, and he remains a coach at heart.
+DUMMY2/p282/p282_129.wav|83|It is not a U-turn.
+DUMMY2/p254/p254_015.wav|41|The Greeks used to imagine that it was a sign from the gods to foretell war or heavy rain.
+DUMMY2/p228/p228_285.wav|57|The songs are just so good.
+DUMMY2/p334/p334_307.wav|38|If they don't, they can expect their funding to be cut.
+DUMMY2/p287/p287_081.wav|77|I've never seen anything like it.
+DUMMY2/p247/p247_083.wav|14|It is a job creation scheme.)
+DUMMY2/p264/p264_051.wav|65|We were leading by two goals.)
+DUMMY2/p335/p335_058.wav|49|Let's see that increase over the years.
+DUMMY2/p236/p236_225.wav|75|There is no quick fix.
+DUMMY2/p374/p374_353.wav|11|And that brings us to the point.
+DUMMY2/p272/p272_076.wav|69|Sounds like The Sixth Sense?
+DUMMY2/p271/p271_152.wav|27|The petition was formally presented at Downing Street yesterday.
+DUMMY2/p228/p228_127.wav|57|They've got to account for it.
+DUMMY2/p276/p276_223.wav|106|It's been a humbling year.
+DUMMY2/p262/p262_248.wav|45|The project has already secured the support of Sir Sean Connery.
+DUMMY2/p314/p314_086.wav|51|The team this year is going places.
+DUMMY2/p225/p225_038.wav|101|Diving is no part of football.
+DUMMY2/p279/p279_088.wav|25|The shareholders will vote to wind up the company on Friday morning.
+DUMMY2/p272/p272_018.wav|69|Aristotle thought that the rainbow was caused by reflection of the sun's rays by the rain.
+DUMMY2/p256/p256_098.wav|90|She told The Herald.
+DUMMY2/p261/p261_218.wav|100|All will be revealed in due course.
+DUMMY2/p265/p265_063.wav|73|IT shouldn't come as a surprise, but it does.
+DUMMY2/p314/p314_042.wav|51|It is all about people being assaulted, abused.
+DUMMY2/p241/p241_188.wav|86|I wish I could say something.
+DUMMY2/p283/p283_111.wav|95|It's good to have a voice.
+DUMMY2/p275/p275_006.wav|40|When the sunlight strikes raindrops in the air, they act as a prism and form a rainbow.
+DUMMY2/p228/p228_092.wav|57|Today I couldn't run on it.
+DUMMY2/p295/p295_343.wav|92|The atmosphere is businesslike.
+DUMMY2/p228/p228_187.wav|57|They will run a mile.
+DUMMY2/p294/p294_317.wav|104|It didn't put me off.
+DUMMY2/p231/p231_445.wav|50|It sounded like a bomb.
+DUMMY2/p272/p272_086.wav|69|Today she has been released.
+DUMMY2/p255/p255_210.wav|31|It was worth a photograph.
+DUMMY2/p229/p229_060.wav|67|And a film maker was born.
+DUMMY2/p260/p260_232.wav|81|The Home Office would not release any further details about the group.
+DUMMY2/p245/p245_025.wav|59|Johnson was pretty low.
+DUMMY2/p333/p333_185.wav|64|This area is perfect for children.
+DUMMY2/p244/p244_242.wav|78|He is a man of the people.
+DUMMY2/p376/p376_187.wav|71|"It is a terrible loss."
+DUMMY2/p239/p239_156.wav|48|It is a good lifestyle.
+DUMMY2/p307/p307_037.wav|22|He released a half-dozen solo albums.
+DUMMY2/p305/p305_185.wav|54|I am not even thinking about that.
+DUMMY2/p272/p272_081.wav|69|It was magic.
+DUMMY2/p302/p302_297.wav|30|I'm trying to stay open on that.
+DUMMY2/p275/p275_320.wav|40|We are in the end game.
+DUMMY2/p239/p239_231.wav|48|Then we will face the Danish champions.
+DUMMY2/p268/p268_301.wav|87|It was only later that the condition was diagnosed.
+DUMMY2/p336/p336_088.wav|98|They failed to reach agreement yesterday.
+DUMMY2/p278/p278_255.wav|10|They made such decisions in London.
+DUMMY2/p361/p361_132.wav|79|That got me out.
+DUMMY2/p307/p307_146.wav|22|You hope he prevails.
+DUMMY2/p244/p244_147.wav|78|They could not ignore the will of parliament, he claimed.
+DUMMY2/p294/p294_283.wav|104|This is our unfinished business.
+DUMMY2/p283/p283_300.wav|95|I would have the hammer in the crowd.
+DUMMY2/p239/p239_079.wav|48|I can understand the frustrations of our fans.
+DUMMY2/p264/p264_009.wav|65|There is , according to legend, a boiling pot of gold at one end. )
+DUMMY2/p307/p307_348.wav|22|He did not oppose the divorce.
+DUMMY2/p304/p304_308.wav|72|We are the gateway to justice.
+DUMMY2/p281/p281_056.wav|36|None has ever been found.
+DUMMY2/p267/p267_158.wav|0|We were given a warm and friendly reception.
+DUMMY2/p300/p300_169.wav|102|Who do these people think they are?
+DUMMY2/p276/p276_177.wav|106|They exist in name alone.
+DUMMY2/p228/p228_245.wav|57|It is a policy which has the full support of the minister.
+DUMMY2/p300/p300_303.wav|102|I'm wondering what you feel about the youngest.
+DUMMY2/p362/p362_247.wav|15|This would give Scotland around eight members.
+DUMMY2/p326/p326_031.wav|28|United were in control without always being dominant.
+DUMMY2/p361/p361_288.wav|79|I did not think it was very proper.
+DUMMY2/p286/p286_145.wav|63|Tiger is not the norm.
+DUMMY2/p234/p234_071.wav|3|She did that for the rest of her life.
+DUMMY2/p263/p263_296.wav|39|The decision was announced at its annual conference in Dunfermline.
+DUMMY2/p323/p323_228.wav|34|She became a heroine of my childhood.
+DUMMY2/p280/p280_346.wav|52|It was a bit like having children.
+DUMMY2/p333/p333_080.wav|64|But the tragedy did not stop there.
+DUMMY2/p226/p226_268.wav|43|That decision is for the British Parliament and people.
+DUMMY2/p362/p362_314.wav|15|Is that right?
+DUMMY2/p240/p240_047.wav|93|It is so sad.
+DUMMY2/p250/p250_207.wav|24|You could feel the heat.
+DUMMY2/p273/p273_176.wav|56|Neither side would reveal the details of the offer.
+DUMMY2/p316/p316_147.wav|85|And frankly, it's been a while.
+DUMMY2/p265/p265_047.wav|73|It is unique.
+DUMMY2/p336/p336_353.wav|98|Sometimes you get them, sometimes you don't.
+DUMMY2/p230/p230_376.wav|35|This hasn't happened in a vacuum.
+DUMMY2/p308/p308_209.wav|107|There is great potential on this river.
+DUMMY2/p250/p250_442.wav|24|We have not yet received a letter from the Irish.
+DUMMY2/p260/p260_037.wav|81|It's a fact.
+DUMMY2/p299/p299_345.wav|58|We're very excited and challenged by the project.
+DUMMY2/p269/p269_218.wav|94|A Grampian Police spokesman said.
+DUMMY2/p306/p306_014.wav|12|To the Hebrews it was a token that there would be no more universal floods.
+DUMMY2/p271/p271_292.wav|27|It's a record label, not a form of music.
+DUMMY2/p247/p247_225.wav|14|I am considered a teenager.)
+DUMMY2/p294/p294_094.wav|104|It should be a condition of employment.
+DUMMY2/p269/p269_031.wav|94|Is this accurate?
+DUMMY2/p275/p275_116.wav|40|It's not fair.
+DUMMY2/p265/p265_006.wav|73|When the sunlight strikes raindrops in the air, they act as a prism and form a rainbow.
+DUMMY2/p285/p285_072.wav|2|Mr Irvine said Mr Rafferty was now in good spirits.
+DUMMY2/p270/p270_167.wav|8|We did what we had to do.
+DUMMY2/p360/p360_397.wav|60|It is a relief.

filelists/vctk_audio_sid_text_val_filelist.txt.cleaned

+DUMMY2/p364/p364_240.wav|88|ɪt hɐd hˈæpənd tə hˌɪm.
+DUMMY2/p280/p280_148.wav|52|ɪt ɪz ˈoʊpən sˈiːzən ɑːnðɪ ˈoʊld fˈɜːm.
+DUMMY2/p231/p231_320.wav|50|haʊˈɛvɚ, hiː ɪz ɐ kˈoʊtʃ, ænd hiː ɹɪmˈeɪnz ɐ kˈoʊtʃ æt hˈɑːɹt.
+DUMMY2/p282/p282_129.wav|83|ɪt ɪz nˌɑːɾə jˈuːtˈɜːn.
+DUMMY2/p254/p254_015.wav|41|ðə ɡɹˈiːks jˈuːzd tʊ ɪmˈædʒɪn ðˌɐɾɪt wʌzɐ sˈaɪn fɹʌmðə ɡˈɑːdz tə foːɹtˈɛl wˈɔːɹ ɔːɹ hˈɛvi ɹˈeɪn.
+DUMMY2/p228/p228_285.wav|57|ðə sˈɔŋz ɑːɹ dʒˈʌst sˌoʊ ɡˈʊd.
+DUMMY2/p334/p334_307.wav|38|ɪf ðeɪ dˈoʊnt, ðeɪ kæn ɛkspˈɛkt ðɛɹ fˈʌndɪŋ təbi kˈʌt.
+DUMMY2/p287/p287_081.wav|77|aɪv nˈɛvɚ sˈiːn ˈɛnɪθˌɪŋ lˈaɪk ɪt.
+DUMMY2/p247/p247_083.wav|14|ɪt ɪz ɐ dʒˈɑːb kɹiːˈeɪʃən skˈiːm.
+DUMMY2/p264/p264_051.wav|65|wiː wɜː lˈiːdɪŋ baɪ tˈuː ɡˈoʊlz.
+DUMMY2/p335/p335_058.wav|49|lˈɛts sˈiː ðæt ˈɪnkɹiːs ˌoʊvɚ ðə jˈɪɹz.
+DUMMY2/p236/p236_225.wav|75|ðɛɹ ɪz nˈoʊ kwˈɪk fˈɪks.
+DUMMY2/p374/p374_353.wav|11|ænd ðæt bɹˈɪŋz ˌʌs tə ðə pˈɔɪnt.
+DUMMY2/p272/p272_076.wav|69|sˈaʊndz lˈaɪk ðə sˈɪksθ sˈɛns?
+DUMMY2/p271/p271_152.wav|27|ðə pətˈɪʃən wʌz fˈɔːɹməli pɹɪzˈɛntᵻd æt dˈaʊnɪŋ stɹˈiːt jˈɛstɚdˌeɪ.
+DUMMY2/p228/p228_127.wav|57|ðeɪv ɡɑːt tʊ ɐkˈaʊnt fɔːɹ ɪt.
+DUMMY2/p276/p276_223.wav|106|ɪts bˌɪn ɐ hˈʌmblɪŋ jˈɪɹ.
+DUMMY2/p262/p262_248.wav|45|ðə pɹˈɑːdʒɛkt hɐz ɔːlɹˌɛdi sɪkjˈʊɹd ðə səpˈoːɹt ʌv sˌɜː ʃˈɔːn kɑːnɚɹi.
+DUMMY2/p314/p314_086.wav|51|ðə tˈiːm ðɪs jˈɪɹ ɪz ɡˌoʊɪŋ plˈeɪsᵻz.
+DUMMY2/p225/p225_038.wav|101|dˈaɪvɪŋ ɪz nˈoʊ pˈɑːɹt ʌv fˈʊtbɔːl.
+DUMMY2/p279/p279_088.wav|25|ðə ʃˈɛɹhoʊldɚz wɪl vˈoʊt tə wˈaɪnd ˈʌp ðə kˈʌmpəni ˌɑːn fɹˈaɪdeɪ mˈɔːɹnɪŋ.
+DUMMY2/p272/p272_018.wav|69|ˈæɹɪstˌɑːɾəl θˈɔːt ðætðə ɹˈeɪnboʊ wʌz kˈɔːzd baɪ ɹɪflˈɛkʃən ʌvðə sˈʌnz ɹˈeɪz baɪ ðə ɹˈeɪn.
+DUMMY2/p256/p256_098.wav|90|ʃiː tˈoʊld ðə hˈɛɹəld.
+DUMMY2/p261/p261_218.wav|100|ˈɔːl wɪl biː ɹɪvˈiːld ɪn dˈuː kˈoːɹs.
+DUMMY2/p265/p265_063.wav|73|ɪt ʃˌʊdənt kˈʌm æz ɐ sɚpɹˈaɪz, bˌʌt ɪt dˈʌz.
+DUMMY2/p314/p314_042.wav|51|ɪt ɪz ˈɔːl ɐbˌaʊt pˈiːpəl bˌiːɪŋ ɐsˈɑːltᵻd, ɐbjˈuːsd.
+DUMMY2/p241/p241_188.wav|86|ˈaɪ wˈɪʃ ˈaɪ kʊd sˈeɪ sˈʌmθɪŋ.
+DUMMY2/p283/p283_111.wav|95|ɪts ɡˈʊd tə hæv ɐ vˈɔɪs.
+DUMMY2/p275/p275_006.wav|40|wˌɛn ðə sˈʌnlaɪt stɹˈaɪks ɹˈeɪndɹɑːps ɪnðɪ ˈɛɹ, ðeɪ ˈækt æz ɐ pɹˈɪzəm ænd fˈɔːɹm ɐ ɹˈeɪnboʊ.
+DUMMY2/p228/p228_092.wav|57|tədˈeɪ ˈaɪ kˌʊdənt ɹˈʌn ˈɑːn ɪt.
+DUMMY2/p295/p295_343.wav|92|ðɪ ˈætməsfˌɪɹ ɪz bˈɪznəslˌaɪk.
+DUMMY2/p228/p228_187.wav|57|ðeɪ wɪl ɹˈʌn ɐ mˈaɪl.
+DUMMY2/p294/p294_317.wav|104|ɪt dˈɪdnt pˌʊt mˌiː ˈɔf.
+DUMMY2/p231/p231_445.wav|50|ɪt sˈaʊndᵻd lˈaɪk ɐ bˈɑːm.
+DUMMY2/p272/p272_086.wav|69|tədˈeɪ ʃiː hɐzbɪn ɹɪlˈiːsd.
+DUMMY2/p255/p255_210.wav|31|ɪt wʌz wˈɜːθ ɐ fˈoʊɾəɡɹˌæf.
+DUMMY2/p229/p229_060.wav|67|ænd ɐ fˈɪlm mˈeɪkɚ wʌz bˈɔːɹn.
+DUMMY2/p260/p260_232.wav|81|ðə hˈoʊm ˈɑːfɪs wʊd nˌɑːt ɹɪlˈiːs ˌɛni fˈɜːðɚ diːtˈeɪlz ɐbˌaʊt ðə ɡɹˈuːp.
+DUMMY2/p245/p245_025.wav|59|dʒˈɑːnsən wʌz pɹˈɪɾi lˈoʊ.
+DUMMY2/p333/p333_185.wav|64|ðɪs ˈɛɹiə ɪz pˈɜːfɛkt fɔːɹ tʃˈɪldɹən.
+DUMMY2/p244/p244_242.wav|78|hiː ɪz ɐ mˈæn ʌvðə pˈiːpəl.
+DUMMY2/p376/p376_187.wav|71|"ɪt ɪz ɐ tˈɛɹəbəl lˈɔs."
+DUMMY2/p239/p239_156.wav|48|ɪt ɪz ɐ ɡˈʊd lˈaɪfstaɪl.
+DUMMY2/p307/p307_037.wav|22|hiː ɹɪlˈiːsd ɐ hˈæfdˈʌzən sˈoʊloʊ ˈælbəmz.
+DUMMY2/p305/p305_185.wav|54|ˈaɪ æm nˌɑːt ˈiːvən θˈɪŋkɪŋ ɐbˌaʊt ðˈæt.
+DUMMY2/p272/p272_081.wav|69|ɪt wʌz mˈædʒɪk.
+DUMMY2/p302/p302_297.wav|30|aɪm tɹˈaɪɪŋ tə stˈeɪ ˈoʊpən ˌɑːn ðˈæt.
+DUMMY2/p275/p275_320.wav|40|wiː ɑːɹ ɪnðɪ ˈɛnd ɡˈeɪm.
+DUMMY2/p239/p239_231.wav|48|ðˈɛn wiː wɪl fˈeɪs ðə dˈeɪnɪʃ tʃˈæmpiənz.
+DUMMY2/p268/p268_301.wav|87|ɪt wʌz ˈoʊnli lˈeɪɾɚ ðætðə kəndˈɪʃən wʌz dˌaɪəɡnˈoʊzd.
+DUMMY2/p336/p336_088.wav|98|ðeɪ fˈeɪld tə ɹˈiːtʃ ɐɡɹˈiːmənt jˈɛstɚdˌeɪ.
+DUMMY2/p278/p278_255.wav|10|ðeɪ mˌeɪd sˈʌtʃ dᵻsˈɪʒənz ɪn lˈʌndən.
+DUMMY2/p361/p361_132.wav|79|ðæt ɡɑːt mˌiː ˈaʊt.
+DUMMY2/p307/p307_146.wav|22|juː hˈoʊp hiː pɹɪvˈeɪlz.
+DUMMY2/p244/p244_147.wav|78|ðeɪ kʊd nˌɑːt ɪɡnˈoːɹ ðə wɪl ʌv pˈɑːɹləmənt, hiː klˈeɪmd.
+DUMMY2/p294/p294_283.wav|104|ðɪs ɪz ˌaʊɚɹ ʌnfˈɪnɪʃt bˈɪznəs.
+DUMMY2/p283/p283_300.wav|95|ˈaɪ wʊdhɐv ðə hˈæmɚɹ ɪnðə kɹˈaʊd.
+DUMMY2/p239/p239_079.wav|48|ˈaɪ kæn ˌʌndɚstˈænd ðə fɹʌstɹˈeɪʃənz ʌv ˌaʊɚ fˈænz.
+DUMMY2/p264/p264_009.wav|65|ðɛɹˈɪz , ɐkˈoːɹdɪŋ tə lˈɛdʒənd, ɐ bˈɔɪlɪŋ pˈɑːt ʌv ɡˈoʊld æt wˈʌn ˈɛnd. 
+DUMMY2/p307/p307_348.wav|22|hiː dɪdnˌɑːt əpˈoʊz ðə dɪvˈoːɹs.
+DUMMY2/p304/p304_308.wav|72|wiː ɑːɹ ðə ɡˈeɪtweɪ tə dʒˈʌstɪs.
+DUMMY2/p281/p281_056.wav|36|nˈʌn hɐz ˈɛvɚ bˌɪn fˈaʊnd.
+DUMMY2/p267/p267_158.wav|0|wiː wɜː ɡˈɪvən ɐ wˈɔːɹm ænd fɹˈɛndli ɹɪsˈɛpʃən.
+DUMMY2/p300/p300_169.wav|102|hˌuː dˈuː ðiːz pˈiːpəl θˈɪŋk ðeɪ ɑːɹ?
+DUMMY2/p276/p276_177.wav|106|ðeɪ ɛɡzˈɪst ɪn nˈeɪm ɐlˈoʊn.
+DUMMY2/p228/p228_245.wav|57|ɪt ɪz ɐ pˈɑːlɪsi wˌɪtʃ hɐz ðə fˈʊl səpˈoːɹt ʌvðə mˈɪnɪstɚ.
+DUMMY2/p300/p300_303.wav|102|aɪm wˈʌndɚɹɪŋ wˌʌt juː fˈiːl ɐbˌaʊt ðə jˈʌŋɡəst.
+DUMMY2/p362/p362_247.wav|15|ðɪs wʊd ɡˈɪv skˈɑːtlənd ɐɹˈaʊnd ˈeɪt mˈɛmbɚz.
+DUMMY2/p326/p326_031.wav|28|juːnˈaɪɾᵻd wɜːɹ ɪn kəntɹˈoʊl wɪðˌaʊt ˈɔːlweɪz bˌiːɪŋ dˈɑːmɪnənt.
+DUMMY2/p361/p361_288.wav|79|ˈaɪ dɪdnˌɑːt θˈɪŋk ɪt wʌz vˈɛɹi pɹˈɑːpɚ.
+DUMMY2/p286/p286_145.wav|63|tˈaɪɡɚɹ ɪz nˌɑːt ðə nˈɔːɹm.
+DUMMY2/p234/p234_071.wav|3|ʃiː dˈɪd ðæt fɚðə ɹˈɛst ʌv hɜː lˈaɪf.
+DUMMY2/p263/p263_296.wav|39|ðə dᵻsˈɪʒən wʌz ɐnˈaʊnst æt ɪts ˈænjuːəl kˈɑːnfɹəns ɪn dˈʌnfɚmlˌaɪn.
+DUMMY2/p323/p323_228.wav|34|ʃiː bɪkˌeɪm ɐ hˈɛɹoʊˌɪn ʌv maɪ tʃˈaɪldhʊd.
+DUMMY2/p280/p280_346.wav|52|ɪt wʌzɐ bˈɪt lˈaɪk hˌævɪŋ tʃˈɪldɹən.
+DUMMY2/p333/p333_080.wav|64|bˌʌt ðə tɹˈædʒədi dɪdnˌɑːt stˈɑːp ðˈɛɹ.
+DUMMY2/p226/p226_268.wav|43|ðæt dᵻsˈɪʒən ɪz fɚðə bɹˈɪɾɪʃ pˈɑːɹləmənt ænd pˈiːpəl.
+DUMMY2/p362/p362_314.wav|15|ɪz ðæt ɹˈaɪt?
+DUMMY2/p240/p240_047.wav|93|ɪt ɪz sˌoʊ sˈæd.
+DUMMY2/p250/p250_207.wav|24|juː kʊd fˈiːl ðə hˈiːt.
+DUMMY2/p273/p273_176.wav|56|nˈiːðɚ sˈaɪd wʊd ɹɪvˈiːl ðə diːtˈeɪlz ʌvðɪ ˈɑːfɚ.
+DUMMY2/p316/p316_147.wav|85|ænd fɹˈæŋkli, ɪts bˌɪn ɐ wˈaɪl.
+DUMMY2/p265/p265_047.wav|73|ɪt ɪz juːnˈiːk.
+DUMMY2/p336/p336_353.wav|98|sˈʌmtaɪmz juː ɡˈɛt ðˌɛm, sˈʌmtaɪmz juː dˈoʊnt.
+DUMMY2/p230/p230_376.wav|35|ðɪs hˈæzənt hˈæpənd ɪn ɐ vˈækjuːm.
+DUMMY2/p308/p308_209.wav|107|ðɛɹ ɪz ɡɹˈeɪt pətˈɛnʃəl ˌɑːn ðɪs ɹˈɪvɚ.
+DUMMY2/p250/p250_442.wav|24|wiː hɐvnˌɑːt jˈɛt ɹɪsˈiːvd ɐ lˈɛɾɚ fɹʌmðɪ ˈaɪɹɪʃ.
+DUMMY2/p260/p260_037.wav|81|ɪts ɐ fˈækt.
+DUMMY2/p299/p299_345.wav|58|wɪɹ vˈɛɹi ɛksˈaɪɾᵻd ænd tʃˈælɪndʒd baɪ ðə pɹˈɑːdʒɛkt.
+DUMMY2/p269/p269_218.wav|94|ɐ ɡɹˈæmpiən pəlˈiːs spˈoʊksmən sˈɛd.
+DUMMY2/p306/p306_014.wav|12|tə ðə hˈiːbɹuːz ɪt wʌzɐ tˈoʊkən ðæt ðɛɹ wʊd biː nˈoʊmˌoːɹ jˌuːnɪvˈɜːsəl flˈʌdz.
+DUMMY2/p271/p271_292.wav|27|ɪts ɐ ɹˈɛkɚd lˈeɪbəl, nˌɑːɾə fˈɔːɹm ʌv mjˈuːzɪk.
+DUMMY2/p247/p247_225.wav|14|ˈaɪ æm kənsˈɪdɚd ɐ tˈiːneɪdʒɚ.
+DUMMY2/p294/p294_094.wav|104|ɪt ʃˌʊd biː ɐ kəndˈɪʃən ʌv ɛmplˈɔɪmənt.
+DUMMY2/p269/p269_031.wav|94|ɪz ðɪs ˈækjʊɹət?
+DUMMY2/p275/p275_116.wav|40|ɪts nˌɑːt fˈɛɹ.
+DUMMY2/p265/p265_006.wav|73|wˌɛn ðə sˈʌnlaɪt stɹˈaɪks ɹˈeɪndɹɑːps ɪnðɪ ˈɛɹ, ðeɪ ˈækt æz ɐ pɹˈɪzəm ænd fˈɔːɹm ɐ ɹˈeɪnboʊ.
+DUMMY2/p285/p285_072.wav|2|mˈɪstɚɹ ˈɜːvaɪn sˈɛd mˈɪstɚ ɹˈæfɚɾi wʌz nˈaʊ ɪn ɡˈʊd spˈɪɹɪts.
+DUMMY2/p270/p270_167.wav|8|wiː dˈɪd wˌʌt wiː hædtə dˈuː.
+DUMMY2/p360/p360_397.wav|60|ɪt ɪz ɐ ɹɪlˈiːf.

inference.ipynb

+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%matplotlib inline\n",
+    "import matplotlib.pyplot as plt\n",
+    "import IPython.display as ipd\n",
+    "\n",
+    "import os\n",
+    "import json\n",
+    "import math\n",
+    "import torch\n",
+    "from torch import nn\n",
+    "from torch.nn import functional as F\n",
+    "from torch.utils.data import DataLoader\n",
+    "\n",
+    "import commons\n",
+    "import utils\n",
+    "from data_utils import TextAudioLoader, TextAudioCollate, TextAudioSpeakerLoader, TextAudioSpeakerCollate\n",
+    "from models import SynthesizerTrn\n",
+    "from text.symbols import symbols\n",
+    "from text import text_to_sequence\n",
+    "\n",
+    "from scipy.io.wavfile import write\n",
+    "\n",
+    "\n",
+    "def get_text(text, hps):\n",
+    "    text_norm = text_to_sequence(text, hps.data.text_cleaners)\n",
+    "    if hps.data.add_blank:\n",
+    "        text_norm = commons.intersperse(text_norm, 0)\n",
+    "    text_norm = torch.LongTensor(text_norm)\n",
+    "    return text_norm"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## LJ Speech"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "hps = utils.get_hparams_from_file(\"./configs/ljs_base.json\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "net_g = SynthesizerTrn(\n",
+    "    len(symbols),\n",
+    "    hps.data.filter_length // 2 + 1,\n",
+    "    hps.train.segment_size // hps.data.hop_length,\n",
+    "    **hps.model).cuda()\n",
+    "_ = net_g.eval()\n",
+    "\n",
+    "_ = utils.load_checkpoint(\"/path/to/pretrained_ljs.pth\", net_g, None)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "stn_tst = get_text(\"VITS is Awesome!\", hps)\n",
+    "with torch.no_grad():\n",
+    "    x_tst = stn_tst.cuda().unsqueeze(0)\n",
+    "    x_tst_lengths = torch.LongTensor([stn_tst.size(0)]).cuda()\n",
+    "    audio = net_g.infer(x_tst, x_tst_lengths, noise_scale=.667, noise_scale_w=0.8, length_scale=1)[0][0,0].data.cpu().float().numpy()\n",
+    "ipd.display(ipd.Audio(audio, rate=hps.data.sampling_rate, normalize=False))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## VCTK"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "hps = utils.get_hparams_from_file(\"./configs/vctk_base.json\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "net_g = SynthesizerTrn(\n",
+    "    len(symbols),\n",
+    "    hps.data.filter_length // 2 + 1,\n",
+    "    hps.train.segment_size // hps.data.hop_length,\n",
+    "    n_speakers=hps.data.n_speakers,\n",
+    "    **hps.model).cuda()\n",
+    "_ = net_g.eval()\n",
+    "\n",
+    "_ = utils.load_checkpoint(\"/path/to/pretrained_vctk.pth\", net_g, None)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "stn_tst = get_text(\"VITS is Awesome!\", hps)\n",
+    "with torch.no_grad():\n",
+    "    x_tst = stn_tst.cuda().unsqueeze(0)\n",
+    "    x_tst_lengths = torch.LongTensor([stn_tst.size(0)]).cuda()\n",
+    "    sid = torch.LongTensor([4]).cuda()\n",
+    "    audio = net_g.infer(x_tst, x_tst_lengths, sid=sid, noise_scale=.667, noise_scale_w=0.8, length_scale=1)[0][0,0].data.cpu().float().numpy()\n",
+    "ipd.display(ipd.Audio(audio, rate=hps.data.sampling_rate, normalize=False))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Voice Conversion"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dataset = TextAudioSpeakerLoader(hps.data.validation_files, hps.data)\n",
+    "collate_fn = TextAudioSpeakerCollate()\n",
+    "loader = DataLoader(dataset, num_workers=8, shuffle=False,\n",
+    "    batch_size=1, pin_memory=True,\n",
+    "    drop_last=True, collate_fn=collate_fn)\n",
+    "data_list = list(loader)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with torch.no_grad():\n",
+    "    x, x_lengths, spec, spec_lengths, y, y_lengths, sid_src = [x.cuda() for x in data_list[0]]\n",
+    "    sid_tgt1 = torch.LongTensor([1]).cuda()\n",
+    "    sid_tgt2 = torch.LongTensor([2]).cuda()\n",
+    "    sid_tgt3 = torch.LongTensor([4]).cuda()\n",
+    "    audio1 = net_g.voice_conversion(spec, spec_lengths, sid_src=sid_src, sid_tgt=sid_tgt1)[0][0,0].data.cpu().float().numpy()\n",
+    "    audio2 = net_g.voice_conversion(spec, spec_lengths, sid_src=sid_src, sid_tgt=sid_tgt2)[0][0,0].data.cpu().float().numpy()\n",
+    "    audio3 = net_g.voice_conversion(spec, spec_lengths, sid_src=sid_src, sid_tgt=sid_tgt3)[0][0,0].data.cpu().float().numpy()\n",
+    "print(\"Original SID: %d\" % sid_src.item())\n",
+    "ipd.display(ipd.Audio(y[0].cpu().numpy(), rate=hps.data.sampling_rate, normalize=False))\n",
+    "print(\"Converted SID: %d\" % sid_tgt1.item())\n",
+    "ipd.display(ipd.Audio(audio1, rate=hps.data.sampling_rate, normalize=False))\n",
+    "print(\"Converted SID: %d\" % sid_tgt2.item())\n",
+    "ipd.display(ipd.Audio(audio2, rate=hps.data.sampling_rate, normalize=False))\n",
+    "print(\"Converted SID: %d\" % sid_tgt3.item())\n",
+    "ipd.display(ipd.Audio(audio3, rate=hps.data.sampling_rate, normalize=False))"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}

losses.py

+import torch 
+from torch.nn import functional as F
+
+import commons
+
+
+def feature_loss(fmap_r, fmap_g):
+  loss = 0
+  for dr, dg in zip(fmap_r, fmap_g):
+    for rl, gl in zip(dr, dg):
+      rl = rl.float().detach()
+      gl = gl.float()
+      loss += torch.mean(torch.abs(rl - gl))
+
+  return loss * 2 
+
+
+def discriminator_loss(disc_real_outputs, disc_generated_outputs):
+  loss = 0
+  r_losses = []
+  g_losses = []
+  for dr, dg in zip(disc_real_outputs, disc_generated_outputs):
+    dr = dr.float()
+    dg = dg.float()
+    r_loss = torch.mean((1-dr)**2)
+    g_loss = torch.mean(dg**2)
+    loss += (r_loss + g_loss)
+    r_losses.append(r_loss.item())
+    g_losses.append(g_loss.item())
+
+  return loss, r_losses, g_losses
+
+
+def generator_loss(disc_outputs):
+  loss = 0
+  gen_losses = []
+  for dg in disc_outputs:
+    dg = dg.float()
+    l = torch.mean((1-dg)**2)
+    gen_losses.append(l)
+    loss += l
+
+  return loss, gen_losses
+
+
+def kl_loss(z_p, logs_q, m_p, logs_p, z_mask):
+  """
+  z_p, logs_q: [b, h, t_t]
+  m_p, logs_p: [b, h, t_t]
+  """
+  z_p = z_p.float()
+  logs_q = logs_q.float()
+  m_p = m_p.float()
+  logs_p = logs_p.float()
+  z_mask = z_mask.float()
+
+  kl = logs_p - logs_q - 0.5
+  kl += 0.5 * ((z_p - m_p)**2) * torch.exp(-2. * logs_p)
+  kl = torch.sum(kl * z_mask)
+  l = kl / torch.sum(z_mask)
+  return l

mel_processing.py

+import math
+import os
+import random
+import torch
+from torch import nn
+import torch.nn.functional as F
+import torch.utils.data
+import numpy as np
+import librosa
+import librosa.util as librosa_util
+from librosa.util import normalize, pad_center, tiny
+from scipy.signal import get_window
+from scipy.io.wavfile import read
+from librosa.filters import mel as librosa_mel_fn
+MAX_WAV_VALUE = 32768.0
+
+def dynamic_range_compression_torch(x, C=1, clip_val=1e-5):
+    """
+    PARAMS
+    ------
+    C: compression factor
+    """
+    return torch.log(torch.clamp(x, min=clip_val) * C)
+
+
+def dynamic_range_decompression_torch(x, C=1):
+    """
+    PARAMS
+    ------
+    C: compression factor used to compress
+    """
+    return torch.exp(x) / C
+
+
+def spectral_normalize_torch(magnitudes):
+    output = dynamic_range_compression_torch(magnitudes)
+    return output
+
+
+def spectral_de_normalize_torch(magnitudes):
+    output = dynamic_range_decompression_torch(magnitudes)
+    return output
+
+
+mel_basis = {}
+hann_window = {}
+
+
+def spectrogram_torch(y, n_fft, sampling_rate, hop_size, win_size, center=False):
+    if torch.min(y) < -1.:
+        print('min value is ', torch.min(y))
+    if torch.max(y) > 1.:
+        print('max value is ', torch.max(y))
+
+    global hann_window
+    dtype_device = str(y.dtype) + '_' + str(y.device)
+    wnsize_dtype_device = str(win_size) + '_' + dtype_device
+    if wnsize_dtype_device not in hann_window:
+        hann_window[wnsize_dtype_device] = torch.hann_window(win_size).to(dtype=y.dtype, device=y.device)
+
+    y = torch.nn.functional.pad(y.unsqueeze(1), (int((n_fft-hop_size)/2), int((n_fft-hop_size)/2)), mode='reflect')
+    y = y.squeeze(1)
+
+    spec = torch.stft(y, n_fft, hop_length=hop_size, win_length=win_size, window=hann_window[wnsize_dtype_device],
+                      center=center, pad_mode='reflect', normalized=False, onesided=True)
+    spec = torch.view_as_real(spec)
+
+    spec = torch.sqrt(spec.pow(2).sum(-1) + 1e-6)
+    return spec
+
+
+def spec_to_mel_torch(spec, n_fft, num_mels, sampling_rate, fmin, fmax):
+    global mel_basis
+    dtype_device = str(spec.dtype) + '_' + str(spec.device)
+    fmax_dtype_device = str(fmax) + '_' + dtype_device
+    if fmax_dtype_device not in mel_basis:
+        mel = librosa_mel_fn(sr=sampling_rate, n_fft=n_fft, n_mels=num_mels, fmin=fmin, fmax=fmax)
+        mel_basis[fmax_dtype_device] = torch.from_numpy(mel).to(dtype=spec.dtype, device=spec.device)
+    spec = torch.matmul(mel_basis[fmax_dtype_device], spec)
+    spec = spectral_normalize_torch(spec)
+    return spec
+
+
+def mel_spectrogram_torch(y, n_fft, num_mels, sampling_rate, hop_size, win_size, fmin, fmax, center=False):
+    if torch.min(y) < -1.:
+        print('min value is ', torch.min(y))
+    if torch.max(y) > 1.:
+        print('max value is ', torch.max(y))
+
+    global mel_basis, hann_window
+    dtype_device = str(y.dtype) + '_' + str(y.device)
+    fmax_dtype_device = str(fmax) + '_' + dtype_device
+    wnsize_dtype_device = str(win_size) + '_' + dtype_device
+    if fmax_dtype_device not in mel_basis:
+        mel = librosa_mel_fn(sr=sampling_rate, n_fft=n_fft, n_mels=num_mels, fmin=fmin, fmax=fmax)
+        mel_basis[fmax_dtype_device] = torch.from_numpy(mel).to(dtype=y.dtype, device=y.device)
+    if wnsize_dtype_device not in hann_window:
+        hann_window[wnsize_dtype_device] = torch.hann_window(win_size).to(dtype=y.dtype, device=y.device)
+
+    y = torch.nn.functional.pad(y.unsqueeze(1), (int((n_fft-hop_size)/2), int((n_fft-hop_size)/2)), mode='reflect')
+    y = y.squeeze(1)
+
+    spec = torch.stft(y, n_fft, hop_length=hop_size, win_length=win_size, window=hann_window[wnsize_dtype_device],
+                      center=center, pad_mode='reflect', normalized=False, onesided=True)
+    spec = torch.view_as_real(spec)
+
+    spec = torch.sqrt(spec.pow(2).sum(-1) + 1e-6)
+
+    spec = torch.matmul(mel_basis[fmax_dtype_device], spec)
+    spec = spectral_normalize_torch(spec)
+
+    return spec

models.py

+import copy
+import math
+import torch
+from torch import nn
+from torch.nn import functional as F
+
+import commons
+import modules
+import attentions
+import monotonic_align
+
+from torch.nn import Conv1d, ConvTranspose1d, AvgPool1d, Conv2d
+from torch.nn.utils import weight_norm, remove_weight_norm, spectral_norm
+from commons import init_weights, get_padding
+
+
+class StochasticDurationPredictor(nn.Module):
+  def __init__(self, in_channels, filter_channels, kernel_size, p_dropout, n_flows=4, gin_channels=0):
+    super().__init__()
+    filter_channels = in_channels # it needs to be removed from future version.
+    self.in_channels = in_channels
+    self.filter_channels = filter_channels
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+    self.n_flows = n_flows
+    self.gin_channels = gin_channels
+
+    self.log_flow = modules.Log()
+    self.flows = nn.ModuleList()
+    self.flows.append(modules.ElementwiseAffine(2))
+    for i in range(n_flows):
+      self.flows.append(modules.ConvFlow(2, filter_channels, kernel_size, n_layers=3))
+      self.flows.append(modules.Flip())
+
+    self.post_pre = nn.Conv1d(1, filter_channels, 1)
+    self.post_proj = nn.Conv1d(filter_channels, filter_channels, 1)
+    self.post_convs = modules.DDSConv(filter_channels, kernel_size, n_layers=3, p_dropout=p_dropout)
+    self.post_flows = nn.ModuleList()
+    self.post_flows.append(modules.ElementwiseAffine(2))
+    for i in range(4):
+      self.post_flows.append(modules.ConvFlow(2, filter_channels, kernel_size, n_layers=3))
+      self.post_flows.append(modules.Flip())
+
+    self.pre = nn.Conv1d(in_channels, filter_channels, 1)
+    self.proj = nn.Conv1d(filter_channels, filter_channels, 1)
+    self.convs = modules.DDSConv(filter_channels, kernel_size, n_layers=3, p_dropout=p_dropout)
+    if gin_channels != 0:
+      self.cond = nn.Conv1d(gin_channels, filter_channels, 1)
+
+  def forward(self, x, x_mask, w=None, g=None, reverse=False, noise_scale=1.0):
+    x = torch.detach(x)
+    x = self.pre(x)
+    if g is not None:
+      g = torch.detach(g)
+      x = x + self.cond(g)
+    x = self.convs(x, x_mask)
+    x = self.proj(x) * x_mask
+
+    if not reverse:
+      flows = self.flows
+      assert w is not None
+
+      logdet_tot_q = 0 
+      h_w = self.post_pre(w)
+      h_w = self.post_convs(h_w, x_mask)
+      h_w = self.post_proj(h_w) * x_mask
+      e_q = torch.randn(w.size(0), 2, w.size(2)).to(device=x.device, dtype=x.dtype) * x_mask
+      z_q = e_q
+      for flow in self.post_flows:
+        z_q, logdet_q = flow(z_q, x_mask, g=(x + h_w))
+        logdet_tot_q += logdet_q
+      z_u, z1 = torch.split(z_q, [1, 1], 1) 
+      u = torch.sigmoid(z_u) * x_mask
+      z0 = (w - u) * x_mask
+      logdet_tot_q += torch.sum((F.logsigmoid(z_u) + F.logsigmoid(-z_u)) * x_mask, [1,2])
+      logq = torch.sum(-0.5 * (math.log(2*math.pi) + (e_q**2)) * x_mask, [1,2]) - logdet_tot_q
+
+      logdet_tot = 0
+      z0, logdet = self.log_flow(z0, x_mask)
+      logdet_tot += logdet
+      z = torch.cat([z0, z1], 1)
+      for flow in flows:
+        z, logdet = flow(z, x_mask, g=x, reverse=reverse)
+        logdet_tot = logdet_tot + logdet
+      nll = torch.sum(0.5 * (math.log(2*math.pi) + (z**2)) * x_mask, [1,2]) - logdet_tot
+      return nll + logq # [b]
+    else:
+      flows = list(reversed(self.flows))
+      ###DEBUG
+      flows = flows[:-2] + [flows[-1]] # remove a useless vflow
+      ###DEBUG
+      z = torch.randn(x.size(0), 2, x.size(2)).to(device=x.device, dtype=x.dtype) * noise_scale
+      for flow in flows:
+        z = flow(z, x_mask, g=x, reverse=reverse)
+      z0, z1 = torch.split(z, [1, 1], 1)
+      logw = z0
+      return logw
+
+
+class DurationPredictor(nn.Module):
+  def __init__(self, in_channels, filter_channels, kernel_size, p_dropout, gin_channels=0):
+    super().__init__()
+
+    self.in_channels = in_channels
+    self.filter_channels = filter_channels
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+    self.gin_channels = gin_channels
+
+    self.drop = nn.Dropout(p_dropout)
+    self.conv_1 = nn.Conv1d(in_channels, filter_channels, kernel_size, padding=kernel_size//2)
+    self.norm_1 = modules.LayerNorm(filter_channels)
+    self.conv_2 = nn.Conv1d(filter_channels, filter_channels, kernel_size, padding=kernel_size//2)
+    self.norm_2 = modules.LayerNorm(filter_channels)
+    self.proj = nn.Conv1d(filter_channels, 1, 1)
+
+    if gin_channels != 0:
+      self.cond = nn.Conv1d(gin_channels, in_channels, 1)
+
+  def forward(self, x, x_mask, g=None):
+    x = torch.detach(x)
+    if g is not None:
+      g = torch.detach(g)
+      x = x + self.cond(g)
+    x = self.conv_1(x * x_mask)
+    x = torch.relu(x)
+    x = self.norm_1(x)
+    x = self.drop(x)
+    x = self.conv_2(x * x_mask)
+    x = torch.relu(x)
+    x = self.norm_2(x)
+    x = self.drop(x)
+    x = self.proj(x * x_mask)
+    return x * x_mask
+
+
+class TextEncoder(nn.Module):
+  def __init__(self,
+      n_vocab,
+      out_channels,
+      hidden_channels,
+      filter_channels,
+      n_heads,
+      n_layers,
+      kernel_size,
+      p_dropout):
+    super().__init__()
+    self.n_vocab = n_vocab
+    self.out_channels = out_channels
+    self.hidden_channels = hidden_channels
+    self.filter_channels = filter_channels
+    self.n_heads = n_heads
+    self.n_layers = n_layers
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+
+    self.emb = nn.Embedding(n_vocab, hidden_channels)
+    nn.init.normal_(self.emb.weight, 0.0, hidden_channels**-0.5)
+
+    self.encoder = attentions.Encoder(
+      hidden_channels,
+      filter_channels,
+      n_heads,
+      n_layers,
+      kernel_size,
+      p_dropout)
+    self.proj= nn.Conv1d(hidden_channels, out_channels * 2, 1)
+
+  def forward(self, x, x_lengths):
+    x = self.emb(x) * math.sqrt(self.hidden_channels) # [b, t, h]
+    ##DEBUG
+    x = torch.transpose(x, 1, -1).contiguous() # [b, h, t]
+    ##DEBUG
+    x_mask = torch.unsqueeze(commons.sequence_mask(x_lengths, x.size(2)), 1).to(x.dtype)
+
+    x = self.encoder(x * x_mask, x_mask)
+    stats = self.proj(x) * x_mask
+
+    m, logs = torch.split(stats, self.out_channels, dim=1)
+    return x, m, logs, x_mask
+
+
+class ResidualCouplingBlock(nn.Module):
+  def __init__(self,
+      channels,
+      hidden_channels,
+      kernel_size,
+      dilation_rate,
+      n_layers,
+      n_flows=4,
+      gin_channels=0):
+    super().__init__()
+    self.channels = channels
+    self.hidden_channels = hidden_channels
+    self.kernel_size = kernel_size
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.n_flows = n_flows
+    self.gin_channels = gin_channels
+
+    self.flows = nn.ModuleList()
+    for i in range(n_flows):
+      self.flows.append(modules.ResidualCouplingLayer(channels, hidden_channels, kernel_size, dilation_rate, n_layers, gin_channels=gin_channels, mean_only=True))
+      self.flows.append(modules.Flip())
+
+  def forward(self, x, x_mask, g=None, reverse=False):
+    if not reverse:
+      for flow in self.flows:
+        x, _ = flow(x, x_mask, g=g, reverse=reverse)
+    else:
+      for flow in reversed(self.flows):
+        x = flow(x, x_mask, g=g, reverse=reverse)
+    return x
+
+
+class PosteriorEncoder(nn.Module):
+  def __init__(self,
+      in_channels,
+      out_channels,
+      hidden_channels,
+      kernel_size,
+      dilation_rate,
+      n_layers,
+      gin_channels=0):
+    super().__init__()
+    self.in_channels = in_channels
+    self.out_channels = out_channels
+    self.hidden_channels = hidden_channels
+    self.kernel_size = kernel_size
+    self.dilation_rate = dilation_rate
+    self.n_layers = n_layers
+    self.gin_channels = gin_channels
+
+    self.pre = nn.Conv1d(in_channels, hidden_channels, 1)
+    self.enc = modules.WN(hidden_channels, kernel_size, dilation_rate, n_layers, gin_channels=gin_channels)
+    self.proj = nn.Conv1d(hidden_channels, out_channels * 2, 1)
+
+  def forward(self, x, x_lengths, g=None):
+    x_mask = torch.unsqueeze(commons.sequence_mask(x_lengths, x.size(2)), 1).to(x.dtype)
+    x = self.pre(x) * x_mask
+    x = self.enc(x, x_mask, g=g)
+    stats = self.proj(x) * x_mask
+    m, logs = torch.split(stats, self.out_channels, dim=1)
+    z = (m + torch.randn_like(m) * torch.exp(logs)) * x_mask
+    return z, m, logs, x_mask
+
+
+class Generator(torch.nn.Module):
+    def __init__(self, initial_channel, resblock, resblock_kernel_sizes, resblock_dilation_sizes, upsample_rates, upsample_initial_channel, upsample_kernel_sizes, gin_channels=0):
+        super(Generator, self).__init__()
+        self.num_kernels = len(resblock_kernel_sizes)
+        self.num_upsamples = len(upsample_rates)
+        self.conv_pre = Conv1d(initial_channel, upsample_initial_channel, 7, 1, padding=3)
+        resblock = modules.ResBlock1 if resblock == '1' else modules.ResBlock2
+
+        self.ups = nn.ModuleList()
+        for i, (u, k) in enumerate(zip(upsample_rates, upsample_kernel_sizes)):
+            self.ups.append(weight_norm(
+                ConvTranspose1d(upsample_initial_channel//(2**i), upsample_initial_channel//(2**(i+1)),
+                                k, u, padding=(k-u)//2)))
+
+        self.resblocks = nn.ModuleList()
+        for i in range(len(self.ups)):
+            ch = upsample_initial_channel//(2**(i+1))
+            for j, (k, d) in enumerate(zip(resblock_kernel_sizes, resblock_dilation_sizes)):
+                self.resblocks.append(resblock(ch, k, d))
+
+        self.conv_post = Conv1d(ch, 1, 7, 1, padding=3, bias=False)
+        self.ups.apply(init_weights)
+
+        if gin_channels != 0:
+            self.cond = nn.Conv1d(gin_channels, upsample_initial_channel, 1)
+
+    def forward(self, x, g=None):
+        x = self.conv_pre(x)
+        if g is not None:
+          x = x + self.cond(g)
+
+        for i in range(self.num_upsamples):
+            x = F.leaky_relu(x, modules.LRELU_SLOPE)
+            x = self.ups[i](x)
+            xs = None
+            for j in range(self.num_kernels):
+                if xs is None:
+                    xs = self.resblocks[i*self.num_kernels+j](x)
+                else:
+                    xs += self.resblocks[i*self.num_kernels+j](x)
+            x = xs / self.num_kernels
+        x = F.leaky_relu(x)
+        x = self.conv_post(x)
+        x = torch.tanh(x)
+
+        return x
+
+    def remove_weight_norm(self):
+        print('Removing weight norm...')
+        for l in self.ups:
+            remove_weight_norm(l)
+        for l in self.resblocks:
+            l.remove_weight_norm()
+
+
+class DiscriminatorP(torch.nn.Module):
+    def __init__(self, period, kernel_size=5, stride=3, use_spectral_norm=False):
+        super(DiscriminatorP, self).__init__()
+        self.period = period
+        self.use_spectral_norm = use_spectral_norm
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        self.convs = nn.ModuleList([
+            norm_f(Conv2d(1, 32, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(32, 128, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(128, 512, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(512, 1024, (kernel_size, 1), (stride, 1), padding=(get_padding(kernel_size, 1), 0))),
+            norm_f(Conv2d(1024, 1024, (kernel_size, 1), 1, padding=(get_padding(kernel_size, 1), 0))),
+        ])
+        self.conv_post = norm_f(Conv2d(1024, 1, (3, 1), 1, padding=(1, 0)))
+
+    def forward(self, x):
+        fmap = []
+
+        # 1d to 2d
+        b, c, t = x.shape
+        if t % self.period != 0: # pad first
+            n_pad = self.period - (t % self.period)
+            x = F.pad(x, (0, n_pad), "reflect")
+            t = t + n_pad
+        ###DEBUG
+        x = x.view(b, c, t // self.period, self.period).contiguous()
+        ###DEBUG
+        for l in self.convs:
+            x = l(x)
+            x = F.leaky_relu(x, modules.LRELU_SLOPE)
+            fmap.append(x)
+        x = self.conv_post(x)
+        fmap.append(x)
+        x = torch.flatten(x, 1, -1)
+
+        return x, fmap
+
+
+class DiscriminatorS(torch.nn.Module):
+    def __init__(self, use_spectral_norm=False):
+        super(DiscriminatorS, self).__init__()
+        norm_f = weight_norm if use_spectral_norm == False else spectral_norm
+        self.convs = nn.ModuleList([
+            norm_f(Conv1d(1, 16, 15, 1, padding=7)),
+            norm_f(Conv1d(16, 64, 41, 4, groups=4, padding=20)),
+            norm_f(Conv1d(64, 256, 41, 4, groups=16, padding=20)),
+            norm_f(Conv1d(256, 1024, 41, 4, groups=64, padding=20)),
+            norm_f(Conv1d(1024, 1024, 41, 4, groups=256, padding=20)),
+            norm_f(Conv1d(1024, 1024, 5, 1, padding=2)),
+        ])
+        self.conv_post = norm_f(Conv1d(1024, 1, 3, 1, padding=1))
+
+    def forward(self, x):
+        fmap = []
+
+        for l in self.convs:
+            x = l(x)
+            x = F.leaky_relu(x, modules.LRELU_SLOPE)
+            fmap.append(x)
+        x = self.conv_post(x)
+        fmap.append(x)
+        x = torch.flatten(x, 1, -1)
+
+        return x, fmap
+
+
+class MultiPeriodDiscriminator(torch.nn.Module):
+    def __init__(self, use_spectral_norm=False):
+        super(MultiPeriodDiscriminator, self).__init__()
+        periods = [2,3,5,7,11]
+
+        discs = [DiscriminatorS(use_spectral_norm=use_spectral_norm)]
+        discs = discs + [DiscriminatorP(i, use_spectral_norm=use_spectral_norm) for i in periods]
+        self.discriminators = nn.ModuleList(discs)
+
+    def forward(self, y, y_hat):
+        y_d_rs = []
+        y_d_gs = []
+        fmap_rs = []
+        fmap_gs = []
+        for i, d in enumerate(self.discriminators):
+            y_d_r, fmap_r = d(y)
+            y_d_g, fmap_g = d(y_hat)
+            y_d_rs.append(y_d_r)
+            y_d_gs.append(y_d_g)
+            fmap_rs.append(fmap_r)
+            fmap_gs.append(fmap_g)
+
+        return y_d_rs, y_d_gs, fmap_rs, fmap_gs
+
+
+
+class SynthesizerTrn(nn.Module):
+  """
+  Synthesizer for Training
+  """
+
+  def __init__(self, 
+    n_vocab,
+    spec_channels,
+    segment_size,
+    inter_channels,
+    hidden_channels,
+    filter_channels,
+    n_heads,
+    n_layers,
+    kernel_size,
+    p_dropout,
+    resblock, 
+    resblock_kernel_sizes, 
+    resblock_dilation_sizes, 
+    upsample_rates, 
+    upsample_initial_channel, 
+    upsample_kernel_sizes,
+    n_speakers=0,
+    gin_channels=0,
+    use_sdp=True,
+    **kwargs):
+
+    super().__init__()
+    self.n_vocab = n_vocab
+    self.spec_channels = spec_channels
+    self.inter_channels = inter_channels
+    self.hidden_channels = hidden_channels
+    self.filter_channels = filter_channels
+    self.n_heads = n_heads
+    self.n_layers = n_layers
+    self.kernel_size = kernel_size
+    self.p_dropout = p_dropout
+    self.resblock = resblock
+    self.resblock_kernel_sizes = resblock_kernel_sizes
+    self.resblock_dilation_sizes = resblock_dilation_sizes
+    self.upsample_rates = upsample_rates
+    self.upsample_initial_channel = upsample_initial_channel
+    self.upsample_kernel_sizes = upsample_kernel_sizes
+    self.segment_size = segment_size
+    self.n_speakers = n_speakers
+    self.gin_channels = gin_channels
+
+    self.use_sdp = use_sdp
+
+    self.enc_p = TextEncoder(n_vocab,
+        inter_channels,
+        hidden_channels,
+        filter_channels,
+        n_heads,
+        n_layers,
+        kernel_size,
+        p_dropout)
+    self.dec = Generator(inter_channels, resblock, resblock_kernel_sizes, resblock_dilation_sizes, upsample_rates, upsample_initial_channel, upsample_kernel_sizes, gin_channels=gin_channels)
+    self.enc_q = PosteriorEncoder(spec_channels, inter_channels, hidden_channels, 5, 1, 16, gin_channels=gin_channels)
+    self.flow = ResidualCouplingBlock(inter_channels, hidden_channels, 5, 1, 4, gin_channels=gin_channels)
+
+    if use_sdp:
+      self.dp = StochasticDurationPredictor(hidden_channels, 192, 3, 0.5, 4, gin_channels=gin_channels)
+    else:
+      self.dp = DurationPredictor(hidden_channels, 256, 3, 0.5, gin_channels=gin_channels)
+
+    if n_speakers > 1:
+      self.emb_g = nn.Embedding(n_speakers, gin_channels)
+
+  def forward(self, x, x_lengths, y, y_lengths, sid=None):
+
+    x, m_p, logs_p, x_mask = self.enc_p(x, x_lengths)
+    if self.n_speakers > 0:
+      g = self.emb_g(sid).unsqueeze(-1) # [b, h, 1]
+    else:
+      g = None
+
+    z, m_q, logs_q, y_mask = self.enc_q(y, y_lengths, g=g)
+    z_p = self.flow(z, y_mask, g=g)
+
+    with torch.no_grad():
+      # negative cross-entropy
+      s_p_sq_r = torch.exp(-2 * logs_p) # [b, d, t]
+      neg_cent1 = torch.sum(-0.5 * math.log(2 * math.pi) - logs_p, [1], keepdim=True) # [b, 1, t_s]
+      ###DEBUG
+      neg_cent2 = torch.matmul(-0.5 * (z_p ** 2).transpose(1, 2).contiguous(), s_p_sq_r) # [b, t_t, d] x [b, d, t_s] = [b, t_t, t_s]
+      neg_cent3 = torch.matmul(z_p.transpose(1, 2).contiguous(), (m_p * s_p_sq_r)) # [b, t_t, d] x [b, d, t_s] = [b, t_t, t_s]
+      ###
+      neg_cent4 = torch.sum(-0.5 * (m_p ** 2) * s_p_sq_r, [1], keepdim=True) # [b, 1, t_s]
+      neg_cent = neg_cent1 + neg_cent2 + neg_cent3 + neg_cent4
+
+      attn_mask = torch.unsqueeze(x_mask, 2) * torch.unsqueeze(y_mask, -1)
+      attn = monotonic_align.maximum_path(neg_cent, attn_mask.squeeze(1)).unsqueeze(1).detach()
+
+    w = attn.sum(2)
+    if self.use_sdp:
+      l_length = self.dp(x, x_mask, w, g=g)
+      l_length = l_length / torch.sum(x_mask)
+    else:
+      logw_ = torch.log(w + 1e-6) * x_mask
+      logw = self.dp(x, x_mask, g=g)
+      l_length = torch.sum((logw - logw_)**2, [1,2]) / torch.sum(x_mask) # for averaging 
+
+    # expand prior
+    ###DEBUG
+    m_p = torch.matmul(attn.squeeze(1), m_p.transpose(1, 2).contiguous()).transpose(1, 2).contiguous()
+    logs_p = torch.matmul(attn.squeeze(1), logs_p.transpose(1, 2).contiguous()).transpose(1, 2).contiguous()
+    ###
+
+    z_slice, ids_slice = commons.rand_slice_segments(z, y_lengths, self.segment_size)
+    o = self.dec(z_slice, g=g)
+    return o, l_length, attn, ids_slice, x_mask, y_mask, (z, z_p, m_p, logs_p, m_q, logs_q)
+
+  def infer(self, x, x_lengths, sid=None, noise_scale=1, length_scale=1, noise_scale_w=1., max_len=None):
+    x, m_p, logs_p, x_mask = self.enc_p(x, x_lengths)
+    if self.n_speakers > 0:
+      g = self.emb_g(sid).unsqueeze(-1) # [b, h, 1]
+    else:
+      g = None
+
+    if self.use_sdp:
+      logw = self.dp(x, x_mask, g=g, reverse=True, noise_scale=noise_scale_w)
+    else:
+      logw = self.dp(x, x_mask, g=g)
+    w = torch.exp(logw) * x_mask * length_scale
+    w_ceil = torch.ceil(w)
+    y_lengths = torch.clamp_min(torch.sum(w_ceil, [1, 2]), 1).long()
+    y_mask = torch.unsqueeze(commons.sequence_mask(y_lengths, None), 1).to(x_mask.dtype)
+    attn_mask = torch.unsqueeze(x_mask, 2) * torch.unsqueeze(y_mask, -1)
+    attn = commons.generate_path(w_ceil, attn_mask)
+    ###DEBUG
+    m_p = torch.matmul(attn.squeeze(1), m_p.transpose(1, 2).contiguous()).transpose(1, 2).contiguous() # [b, t', t], [b, t, d] -> [b, d, t']
+    logs_p = torch.matmul(attn.squeeze(1), logs_p.transpose(1, 2).contiguous()).transpose(1, 2).contiguous() # [b, t', t], [b, t, d] -> [b, d, t']
+    ###DEBUG
+    z_p = m_p + torch.randn_like(m_p) * torch.exp(logs_p) * noise_scale
+    z = self.flow(z_p, y_mask, g=g, reverse=True)
+    o = self.dec((z * y_mask)[:,:,:max_len], g=g)
+    return o, attn, y_mask, (z, z_p, m_p, logs_p)
+
+  def voice_conversion(self, y, y_lengths, sid_src, sid_tgt):
+    assert self.n_speakers > 0, "n_speakers have to be larger than 0."
+    g_src = self.emb_g(sid_src).unsqueeze(-1)
+    g_tgt = self.emb_g(sid_tgt).unsqueeze(-1)
+    z, m_q, logs_q, y_mask = self.enc_q(y, y_lengths, g=g_src)
+    z_p = self.flow(z, y_mask, g=g_src)
+    z_hat = self.flow(z_p, y_mask, g=g_tgt, reverse=True)
+    o_hat = self.dec(z_hat * y_mask, g=g_tgt)
+    return o_hat, y_mask, (z, z_p, z_hat)
+

monotonic_align/__init__.py

+import numpy as np
+import torch
+from .monotonic_align.core import maximum_path_c
+
+
+def maximum_path(neg_cent, mask):
+  """ Cython optimized version.
+  neg_cent: [b, t_t, t_s]
+  mask: [b, t_t, t_s]
+  """
+  device = neg_cent.device
+  dtype = neg_cent.dtype
+  neg_cent = neg_cent.data.cpu().numpy().astype(np.float32)
+  path = np.zeros(neg_cent.shape, dtype=np.int32)
+
+  t_t_max = mask.sum(1)[:, 0].data.cpu().numpy().astype(np.int32)
+  t_s_max = mask.sum(2)[:, 0].data.cpu().numpy().astype(np.int32)
+  maximum_path_c(path, neg_cent, t_t_max, t_s_max)
+  return torch.from_numpy(path).to(device=device, dtype=dtype)