{ "leaderboard_sha256": "8b8caeb5d33be17c7b40566028a54f3e312728ca95205057ff0d5b52c99514c3", "scope": "All 33 models, all 1088 original records each; scores are unchanged.", "normalization": "whisper_english", "models": [ { "model": "llasa", "checkpoint": "HKUSTAudio/Llasa-1B-Multilingual", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.7399313405342042, "cer": 0.514342676902934, "corpus_metrics_match": true, "word_substitutions": 5386, "word_deletions": 1712, "word_insertions": 1739, "reference_words": 11943, "empty_transcripts": 17, "wer_over_100_percent": 177, "exact_transcripts": 34, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 2.52, 5.0, 6.0, 7.16, 10.24, 34.12 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 3, "top20_fraction_of_word_errors": 0.06834898721285504, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.895910780669145, "empty_transcripts": 3 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.7481558321807285, "empty_transcripts": 11 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.6857994781960493, "empty_transcripts": 3 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.5217391304347826, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "we", "transcript": "i", "count": 3 }, { "reference": "was", "transcript": "is", "count": 3 }, { "reference": "the", "transcript": "a", "count": 3 }, { "reference": "and", "transcript": "in", "count": 3 }, { "reference": "ever", "transcript": "er", "count": 2 }, { "reference": "from outer space", "transcript": "for of us", "count": 2 }, { "reference": "as", "transcript": "this", "count": 2 }, { "reference": "style", "transcript": "stay", "count": 2 }, { "reference": "like", "transcript": "at", "count": 2 }, { "reference": "his", "transcript": "this", "count": 2 }, { "reference": "will no one go", "transcript": "where you are going", "count": 2 }, { "reference": "he must be disguised to avoid encounters", "transcript": "yes it has passed the 1st turn", "count": 2 } ], "common_deletions": [ [ "and", 5 ], [ "in", 5 ], [ "the", 4 ], [ "time", 3 ], [ "name", 3 ], [ "on", 3 ], [ "year", 2 ], [ "home", 2 ] ], "common_insertions": [ [ "a", 23 ], [ "the", 19 ], [ "and", 18 ], [ "i", 11 ], [ "that", 8 ], [ "we", 5 ], [ "in", 5 ], [ "what", 4 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19579295-common_voice_en_19579310", "reference": "Russo was thrown out of his car and died immediately.", "transcript": "Russell was thrown off of his pa-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a", "duration_s": 4.44, "metrics": { "wer": 11.1, "cer": 4.173076923076923, "mer": 0.9652173913043478, "wil": 0.9860869565217392, "wip": 0.01391304347826087, "word_hits": 4, "word_substitutions": 6, "word_deletions": 0, "word_insertions": 105, "reference_words": 10, "char_hits": 31, "char_substitutions": 21, "char_deletions": 0, "char_insertions": 196, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37470753-common_voice_en_37470745", "reference": "Carson greatly admired the talents of Hank Williams.", "transcript": "CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY CURTAIN PLATE TECHNOLOGY", "duration_s": 6.1, "metrics": { "wer": 6.0, "cer": 7.019607843137255, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 40, "reference_words": 8, "char_hits": 41, "char_substitutions": 10, "char_deletions": 0, "char_insertions": 348, "reference_chars": 51, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19671290-common_voice_en_19671292", "reference": "Colonel John Haslet quickly reassigned him to duty as the regimental surgeon.", "transcript": "Cut! I don't know what you're talking about. I don't know what you're talking about. I don't know what you're talking about. I don't know what you're talking about.", "duration_s": 12.08, "metrics": { "wer": 3.0833333333333335, "cer": 1.6578947368421053, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 12, "word_deletions": 0, "word_insertions": 25, "reference_words": 12, "char_hits": 41, "char_substitutions": 35, "char_deletions": 0, "char_insertions": 91, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_557458-common_voice_en_557459", "reference": "The heather was on fire, and you could see a thin blue smoke rising up slowly.", "transcript": "I don't know what's on the screen. I don't know what's on the screen. It's more... I don't know what's on the screen.", "duration_s": 9.98, "metrics": { "wer": 1.75, "cer": 1.0921052631578947, "mer": 0.9333333333333333, "wil": 0.9916666666666667, "wip": 0.008333333333333333, "word_hits": 2, "word_substitutions": 14, "word_deletions": 0, "word_insertions": 14, "reference_words": 16, "char_hits": 35, "char_substitutions": 41, "char_deletions": 0, "char_insertions": 42, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_537766-common_voice_en_537767", "reference": "The heather was on fire, and you could see a thin blue smoke rising up slowly.", "transcript": "I don't know what's on the screen. I don't know what's on the screen. It's more... I don't know what's on the screen.", "duration_s": 9.98, "metrics": { "wer": 1.75, "cer": 1.0921052631578947, "mer": 0.9333333333333333, "wil": 0.9916666666666667, "wip": 0.008333333333333333, "word_hits": 2, "word_substitutions": 14, "word_deletions": 0, "word_insertions": 14, "reference_words": 16, "char_hits": 35, "char_substitutions": 41, "char_deletions": 0, "char_insertions": 42, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25127181-common_voice_en_25127184", "reference": "The style changes very quickly when executed.", "transcript": "This time change is very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very.", "duration_s": 6.74, "metrics": { "wer": 3.857142857142857, "cer": 2.590909090909091, "mer": 0.9642857142857143, "wil": 0.9948979591836735, "wip": 0.00510204081632653, "word_hits": 1, "word_substitutions": 6, "word_deletions": 0, "word_insertions": 21, "reference_words": 7, "char_hits": 26, "char_substitutions": 17, "char_deletions": 1, "char_insertions": 96, "reference_chars": 44, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19816147-common_voice_en_19816145", "reference": "While this solution was workable, it was hardly ideal.", "transcript": "Well, yes, sir, it's a little, it's a little, it's a little, it's a little, it's a little, it's a little.", "duration_s": 9.0, "metrics": { "wer": 2.888888888888889, "cer": 1.4230769230769231, "mer": 0.9629629629629629, "wil": 0.9958847736625515, "wip": 0.004115226337448559, "word_hits": 1, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 18, "reference_words": 9, "char_hits": 28, "char_substitutions": 24, "char_deletions": 0, "char_insertions": 50, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19977527-common_voice_en_19977528", "reference": "Copperopolis was on the main stage road from Sonora to Sacramento.", "transcript": "Corporate police, this is the end of the story, thank you, thank you, this is the end of the story, thank you, thank you, thank you.", "duration_s": 8.6, "metrics": { "wer": 2.272727272727273, "cer": 1.353846153846154, "mer": 0.9615384615384616, "wil": 0.9965034965034965, "wip": 0.003496503496503497, "word_hits": 1, "word_substitutions": 10, "word_deletions": 0, "word_insertions": 15, "reference_words": 11, "char_hits": 36, "char_substitutions": 29, "char_deletions": 0, "char_insertions": 59, "reference_chars": 65, "exact_match_rate": 0.0 } } ] }, { "model": "dia", "checkpoint": "nari-labs/Dia-1.6B-0626", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.673532613246253, "cer": 0.5928256858894245, "corpus_metrics_match": true, "word_substitutions": 1067, "word_deletions": 3554, "word_insertions": 3423, "reference_words": 11943, "empty_transcripts": 242, "wer_over_100_percent": 52, "exact_transcripts": 492, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 2.1130158730158732, 5.593106575963719, 11.075918367346938, 21.559727891156463, 29.976961451247167, 30.023401360544216 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.30034808552958725, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 1.8754646840148699, "empty_transcripts": 29 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.7146150299677271, "empty_transcripts": 194 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.3112187849422288, "empty_transcripts": 19 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "the", "transcript": "0", "count": 3 }, { "reference": "this was the strangest of all things that ever", "transcript": "i do not know what you are talking about", "count": 2 }, { "reference": "when", "transcript": "thank", "count": 2 }, { "reference": "are at", "transcript": "for watching", "count": 2 }, { "reference": "rare", "transcript": "rarer", "count": 2 }, { "reference": "little thieves are hanged but great", "transcript": "i will see you next time", "count": 2 }, { "reference": "this", "transcript": "the", "count": 2 }, { "reference": "the airport is", "transcript": "thanks for watching", "count": 2 }, { "reference": "any tutting that is", "transcript": "thank you for watching", "count": 1 }, { "reference": "requirements are rather high for this job", "transcript": "receive notifications when i upload new videos", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "the railway station is also an important bus interchange station", "transcript": "you enjoy this video i hope you enjoy this video", "count": 1 } ], "common_deletions": [ [ "is", 3 ], [ "came to earth from outer space", 2 ], [ "the trees", 2 ], [ "will no one go", 2 ], [ "a party and somebody says banana bread a sniper might be near you", 2 ], [ "i am trying to explain to my mother how to get pictures off her phone", 2 ], [ "ones escape", 2 ], [ "i have the diet of a kid who found $20", 2 ] ], "common_insertions": [ [ "thank", 3 ], [ "i do not", 2 ], [ "will show you how to", 2 ], [ "secondary schools giving a general practical education", 2 ], [ "hello friends today i will show you how to make a simple easy and easy to prepare cake please subscribe to my channel and press", 1 ], [ "bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to receive notifications when i upload new videos please subscribe to my channel and press the bell icon to", 1 ], [ "hello and welcome back to my channel i hope you are enjoying this video i hope you enjoy this video i hope", 1 ], [ "other infrastructure soon followed", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_17879420-common_voice_en_17879418", "reference": "Every cloud has a silver lining.", "transcript": "Welcome to my channel, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I", "duration_s": 30.00018140589569, "metrics": { "wer": 31.833333333333332, "cer": 27.06451612903226, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 6, "word_deletions": 0, "word_insertions": 185, "reference_words": 6, "char_hits": 21, "char_substitutions": 10, "char_deletions": 0, "char_insertions": 829, "reference_chars": 31, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_17784-common_voice_en_17786", "reference": "The virus spread faster than wildfire.", "transcript": "Welcome to my channel, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I", "duration_s": 26.70294784580499, "metrics": { "wer": 31.666666666666668, "cer": 22.62162162162162, "mer": 0.9947643979057592, "wil": 0.9991273996509599, "wip": 0.0008726003490401396, "word_hits": 1, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 185, "reference_words": 6, "char_hits": 23, "char_substitutions": 14, "char_deletions": 0, "char_insertions": 823, "reference_chars": 37, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18944281-common_voice_en_18944280", "reference": "Recordings of Greene are scarce in the compact disc era.", "transcript": "Welcome to my channel, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I hope you enjoy the video, I", "duration_s": 30.00018140589569, "metrics": { "wer": 19.0, "cer": 15.036363636363637, "mer": 0.9947643979057592, "wil": 0.9994764397905759, "wip": 0.0005235602094240838, "word_hits": 1, "word_substitutions": 9, "word_deletions": 0, "word_insertions": 181, "reference_words": 10, "char_hits": 33, "char_substitutions": 22, "char_deletions": 0, "char_insertions": 805, "reference_chars": 55, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37232389-common_voice_en_37190783", "reference": "Will you be happy when I am in the earth?", "transcript": "Hi, welcome to my channel, I hope you are enjoying this video, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching, I hope you like it, thank you for watching", "duration_s": 11.795736961451247, "metrics": { "wer": 18.1, "cer": 20.475, "mer": 0.9890710382513661, "wil": 0.9978142076502732, "wip": 0.002185792349726776, "word_hits": 2, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 173, "reference_words": 10, "char_hits": 38, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 817, "reference_chars": 40, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19455206-common_voice_en_19455210", "reference": "The land is volcanic in origin with mountains in the interior.", "transcript": "Hello everyone, welcome back to my channel, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I hope you are enjoying this video, I'll see you next time.", "duration_s": 29.291972789115647, "metrics": { "wer": 15.181818181818182, "cer": 12.80327868852459, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 11, "word_deletions": 0, "word_insertions": 156, "reference_words": 11, "char_hits": 55, "char_substitutions": 6, "char_deletions": 0, "char_insertions": 775, "reference_chars": 61, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_275174-common_voice_en_275175", "reference": "The requirements are rather high for this job.", "transcript": "Hello friends. Today I will show you how to make a simple, easy, and easy-to-prepare cake. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos. Please subscribe to my channel and press the bell icon to receive notifications when I upload new videos.", "duration_s": 11.064308390022676, "metrics": { "wer": 20.125, "cer": 19.644444444444446, "mer": 0.9938271604938271, "wil": 0.9992283950617284, "wip": 0.0007716049382716049, "word_hits": 1, "word_substitutions": 7, "word_deletions": 0, "word_insertions": 154, "reference_words": 8, "char_hits": 42, "char_substitutions": 3, "char_deletions": 0, "char_insertions": 881, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19691556-common_voice_en_19691557", "reference": "The law has been replaced twice since then.", "transcript": "Hello friends, welcome to my channel. Today I will show you how to make a very simple and easy to make Christmas tree. I will show you how to make a very simple and easy to make Christmas tree. I will show you how to make a very simple and easy to make Christmas tree. I will show you how to make a very simple and easy to make Christmas tree. I will show you how to make a very simple and easy to make Christmas tree. I will show you how to make a very simple and easy to make Christmas tree. I will show you how to make Christmas tree. I will show you how to make Christmas tree. I will show you how to make Christmas tree. I will show you how to make Christmas tree. I will show you how to make Christmas tree. I will show you how to make Christmas tree.", "duration_s": 10.100680272108843, "metrics": { "wer": 19.625, "cer": 16.761904761904763, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 149, "reference_words": 8, "char_hits": 39, "char_substitutions": 3, "char_deletions": 0, "char_insertions": 701, "reference_chars": 42, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_26952530-common_voice_en_26952588", "reference": "Critics have taken Roney to task for sounding too similar to his idol.", "transcript": "Hello friends, welcome to my channel. Today I will show you how to make a very simple and easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make and very easy-to-make", "duration_s": 13.56045351473923, "metrics": { "wer": 9.538461538461538, "cer": 7.333333333333333, "mer": 0.9841269841269841, "wil": 0.9975579975579976, "wip": 0.002442002442002442, "word_hits": 2, "word_substitutions": 11, "word_deletions": 0, "word_insertions": 113, "reference_words": 13, "char_hits": 56, "char_substitutions": 13, "char_deletions": 0, "char_insertions": 493, "reference_chars": 69, "exact_match_rate": 0.0 } } ] }, { "model": "cosyvoice", "checkpoint": "artifacts/cosyvoice3", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.13815624215021352, "cer": 0.07709932750104148, "corpus_metrics_match": true, "word_substitutions": 1286, "word_deletions": 114, "word_insertions": 250, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 7, "exact_transcripts": 456, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.96, 3.0, 3.72, 4.52, 5.545999999999994, 7.56 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.12545454545454546, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.23048327137546468, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.14061779621945597, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.11367871785314947, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.021739130434782608, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "was", "transcript": "is", "count": 10 }, { "reference": "a", "transcript": "the", "count": 10 }, { "reference": "the", "transcript": "a", "count": 9 }, { "reference": "is", "transcript": "was", "count": 4 }, { "reference": "near", "transcript": "in", "count": 3 }, { "reference": "he had", "transcript": "you have", "count": 2 }, { "reference": "it", "transcript": "he", "count": 2 }, { "reference": "will no one go", "transcript": "you know him though", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "wish", "transcript": "which", "count": 2 }, { "reference": "with", "transcript": "the", "count": 2 }, { "reference": "and", "transcript": "now", "count": 2 } ], "common_deletions": [ [ "is", 6 ], [ "a", 5 ], [ "to", 4 ], [ "in", 4 ], [ "and", 3 ], [ "him", 3 ], [ "it", 3 ], [ "the", 3 ] ], "common_insertions": [ [ "a", 11 ], [ "the", 10 ], [ "is", 6 ], [ "and", 4 ], [ "i", 4 ], [ "in", 4 ], [ "we", 3 ], [ "if", 3 ] ], "worst_examples": [ { "id": "seed_common_voice_en_32377071-common_voice_en_32377070", "reference": "The name means \"common land\" or \"land shared between local inhabitants\".", "transcript": "The way he's coming at it, I want to share it with all the other actors.", "duration_s": 4.04, "metrics": { "wer": 1.4545454545454546, "cer": 0.6417910447761194, "mer": 0.9411764705882353, "wil": 0.9946524064171123, "wip": 0.0053475935828877, "word_hits": 1, "word_substitutions": 10, "word_deletions": 0, "word_insertions": 6, "reference_words": 11, "char_hits": 33, "char_substitutions": 29, "char_deletions": 5, "char_insertions": 9, "reference_chars": 67, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23809069-common_voice_en_23809074", "reference": "The \"Review\" somehow limped on, buoyed up by a narrow but devoted subscription base.", "transcript": "A huge thank you to all of you for joining us. We are on a narrow but devoted subscription base.", "duration_s": 5.0, "metrics": { "wer": 1.0, "cer": 0.525, "mer": 0.7, "wil": 0.8714285714285714, "wip": 0.12857142857142856, "word_hits": 6, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 6, "reference_words": 14, "char_hits": 52, "char_substitutions": 28, "char_deletions": 0, "char_insertions": 14, "reference_chars": 80, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_17406732-common_voice_en_17406734", "reference": "They both stormed out, neither thinking to fill me in on where they were going.", "transcript": "If they've stormed out, you'd be thinking the four-year-old would have a bang.", "duration_s": 3.68, "metrics": { "wer": 0.9333333333333333, "cer": 0.5844155844155844, "mer": 0.7777777777777778, "wil": 0.9372549019607843, "wip": 0.06274509803921569, "word_hits": 4, "word_substitutions": 10, "word_deletions": 1, "word_insertions": 3, "reference_words": 15, "char_hits": 42, "char_substitutions": 27, "char_deletions": 8, "char_insertions": 10, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19665909-common_voice_en_19665911", "reference": "Winter afternoons are cool and frequently sunny, but mornings are freezing to frigid.", "transcript": "We're interacting with the neck line frequently, Sonny, but many of you are freezing to fridge it.", "duration_s": 4.84, "metrics": { "wer": 1.0, "cer": 0.4819277108433735, "mer": 0.7222222222222222, "wil": 0.8931623931623931, "wip": 0.10683760683760685, "word_hits": 5, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 5, "reference_words": 13, "char_hits": 59, "char_substitutions": 21, "char_deletions": 3, "char_insertions": 16, "reference_chars": 83, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_17369337-common_voice_en_17369336", "reference": "Does the new splinter cell game come with a multiplayer mode?", "transcript": "Hey, didn't you expect to see a game coming in multiple mode?", "duration_s": 3.16, "metrics": { "wer": 1.0909090909090908, "cer": 0.5666666666666667, "mer": 0.8571428571428571, "wil": 0.972027972027972, "wip": 0.027972027972027975, "word_hits": 2, "word_substitutions": 8, "word_deletions": 1, "word_insertions": 3, "reference_words": 11, "char_hits": 33, "char_substitutions": 20, "char_deletions": 7, "char_insertions": 7, "reference_chars": 60, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18746637-common_voice_en_18746639", "reference": "The nest is a flimsy platform of sticks is built by both sexes.", "transcript": "And that's the best way to platform a sexy bit about sexes.", "duration_s": 4.08, "metrics": { "wer": 0.8461538461538461, "cer": 0.5806451612903226, "mer": 0.8461538461538461, "wil": 0.9763313609467456, "wip": 0.02366863905325444, "word_hits": 2, "word_substitutions": 11, "word_deletions": 0, "word_insertions": 0, "reference_words": 13, "char_hits": 34, "char_substitutions": 17, "char_deletions": 11, "char_insertions": 8, "reference_chars": 62, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19722359-common_voice_en_19722362", "reference": "Other infrastructure soon followed.", "transcript": "I don't understand what you're saying from root.", "duration_s": 1.68, "metrics": { "wer": 2.5, "cer": 1.0, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 6, "reference_words": 4, "char_hits": 16, "char_substitutions": 17, "char_deletions": 1, "char_insertions": 16, "reference_chars": 34, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_30471104-common_voice_en_30471105", "reference": "The ant colonies move around following the queen\u2019s pheromone trail.", "transcript": "Peter Cullen is moving a man following the Queen's form on the trail.", "duration_s": 3.6, "metrics": { "wer": 0.9090909090909091, "cer": 0.4090909090909091, "mer": 0.7142857142857143, "wil": 0.8961038961038961, "wip": 0.10389610389610389, "word_hits": 4, "word_substitutions": 7, "word_deletions": 0, "word_insertions": 3, "reference_words": 11, "char_hits": 47, "char_substitutions": 14, "char_deletions": 5, "char_insertions": 8, "reference_chars": 66, "exact_match_rate": 0.0 } } ] }, { "model": "vui", "checkpoint": "vui-abraham-100m.pt", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.12149376203633928, "cer": 0.08397310004165923, "corpus_metrics_match": true, "word_substitutions": 528, "word_deletions": 182, "word_insertions": 741, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 19, "exact_transcripts": 540, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 4.133151927437642, 4.31891156462585, 4.690430839002268, 5.433469387755102, 7.058866213151927, 52.61641723356009 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.1509303928325293, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.4349442379182156, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.11837252189949285, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.0667163622810287, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.2391304347826087, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 9 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "his", "transcript": "hit", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "the", "transcript": "a", "count": 2 }, { "reference": "out", "transcript": "outs", "count": 2 }, { "reference": "a", "transcript": "no", "count": 2 }, { "reference": "this", "transcript": "the", "count": 2 }, { "reference": "but great ones", "transcript": "make one is", "count": 2 }, { "reference": "heather", "transcript": "heater", "count": 2 }, { "reference": "and", "transcript": "in", "count": 2 }, { "reference": "his son is", "transcript": "is done his", "count": 1 } ], "common_deletions": [ [ "a", 17 ], [ "it", 8 ], [ "the", 6 ], [ "this", 6 ], [ "in", 3 ], [ "as", 3 ], [ "i", 3 ], [ "his", 2 ] ], "common_insertions": [ [ "and", 22 ], [ "but", 11 ], [ "the", 4 ], [ "an", 3 ], [ "to", 2 ], [ "he", 2 ], [ "just", 2 ], [ "in the", 2 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19049396-common_voice_en_19049395", "reference": "Players begin by selecting one of three factions.", "transcript": "I first did a study on science and I searched for science as seeing which solutions were", "duration_s": 4.458231292517007, "metrics": { "wer": 2.125, "cer": 1.4772727272727273, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 9, "reference_words": 8, "char_hits": 21, "char_substitutions": 23, "char_deletions": 0, "char_insertions": 42, "reference_chars": 44, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20266775-common_voice_en_20266781", "reference": "Brim also had one child from her first marriage.", "transcript": "press on it the beat default meter prints him also and successfully so on", "duration_s": 4.597551020408163, "metrics": { "wer": 1.5555555555555556, "cer": 1.288888888888889, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 9, "word_deletions": 0, "word_insertions": 5, "reference_words": 9, "char_hits": 15, "char_substitutions": 30, "char_deletions": 0, "char_insertions": 28, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_21184954-common_voice_en_21184958", "reference": "Early residents of the village were French Canadians and Menomonee Indians.", "transcript": "to the discipline of some geoffrey london pain that's how that's disamassed", "duration_s": 4.551111111111111, "metrics": { "wer": 1.1818181818181819, "cer": 0.7972972972972973, "mer": 0.9285714285714286, "wil": 0.9935064935064936, "wip": 0.006493506493506493, "word_hits": 1, "word_substitutions": 10, "word_deletions": 0, "word_insertions": 3, "reference_words": 11, "char_hits": 27, "char_substitutions": 38, "char_deletions": 9, "char_insertions": 12, "reference_chars": 74, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22771349-common_voice_en_22771352", "reference": "The Union urgently needed new revenue.", "transcript": "that you need to leave the old ones and put them to fix.", "duration_s": 4.458231292517007, "metrics": { "wer": 2.0, "cer": 1.054054054054054, "mer": 0.9230769230769231, "wil": 0.9871794871794872, "wip": 0.01282051282051282, "word_hits": 1, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 7, "reference_words": 6, "char_hits": 16, "char_substitutions": 21, "char_deletions": 0, "char_insertions": 18, "reference_chars": 37, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_493740-common_voice_en_493738", "reference": "Mrs. Lewis was selected as the new commissioner.", "transcript": "but so that's the toughest thing to actually choose to live", "duration_s": 4.922630385487528, "metrics": { "wer": 1.5, "cer": 0.9795918367346939, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 4, "reference_words": 8, "char_hits": 15, "char_substitutions": 31, "char_deletions": 3, "char_insertions": 14, "reference_chars": 49, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19579295-common_voice_en_19579294", "reference": "Zabuton are found throughout Japan and enter many aspects of the culture.", "transcript": "bullet to some of the stones and start to come because now", "duration_s": 4.690430839002268, "metrics": { "wer": 1.0, "cer": 0.75, "mer": 0.9230769230769231, "wil": 0.9930555555555556, "wip": 0.006944444444444444, "word_hits": 1, "word_substitutions": 10, "word_deletions": 1, "word_insertions": 1, "reference_words": 12, "char_hits": 22, "char_substitutions": 32, "char_deletions": 18, "char_insertions": 4, "reference_chars": 72, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23729047-common_voice_en_23729051", "reference": "Anubis plans to destroy Abydos but Daniel appears and demands that Anubis stop.", "transcript": "Nubis plans to destroy Aeth. I will ask for a sequel, and then meet us. Antivans, stop.", "duration_s": 6.919546485260771, "metrics": { "wer": 0.9230769230769231, "cer": 0.5, "mer": 0.7058823529411765, "wil": 0.8868778280542986, "wip": 0.11312217194570137, "word_hits": 5, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 4, "reference_words": 13, "char_hits": 45, "char_substitutions": 31, "char_deletions": 2, "char_insertions": 6, "reference_chars": 78, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_26933529-common_voice_en_26933531", "reference": "There is also a genetic test that can be done for this disorder.", "transcript": "There is also a genetic test that can be done for this disorder. So I'm going to play this song after I...", "duration_s": 4.829750566893424, "metrics": { "wer": 0.8461538461538461, "cer": 0.6666666666666666, "mer": 0.4583333333333333, "wil": 0.45833333333333337, "wip": 0.5416666666666666, "word_hits": 13, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 11, "reference_words": 13, "char_hits": 63, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 42, "reference_chars": 63, "exact_match_rate": 0.0 } } ] }, { "model": "conversationtts", "checkpoint": "artifacts/conversationtts/model.safetensors", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.08716402913840743, "cer": 0.0824703921918705, "corpus_metrics_match": true, "word_substitutions": 216, "word_deletions": 29, "word_insertions": 796, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 20, "exact_transcripts": 833, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.88, 3.36, 4.24, 5.0600000000000005, 7.84, 41.12 ], "high_clipping_recordings": 25, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.4783861671469741, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.25650557620817843, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.07514983863531581, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.09317927692881103, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.021739130434782608, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 3 }, { "reference": "in", "transcript": "and", "count": 2 }, { "reference": "revenge", "transcript": "ruins", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "and", "transcript": "in", "count": 2 }, { "reference": "her", "transcript": "our", "count": 2 }, { "reference": "parfrey", "transcript": "parfey", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "how about", "transcript": "i bet", "count": 1 }, { "reference": "kick", "transcript": "kicked", "count": 1 }, { "reference": "hypergamy", "transcript": "ghani", "count": 1 }, { "reference": "ant", "transcript": "an", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "a", 2 ], [ "of", 2 ], [ "mart", 1 ], [ "swallow", 1 ], [ "are", 1 ], [ "will always let you down", 1 ], [ "run", 1 ] ], "common_insertions": [ [ "the", 4 ], [ "and", 3 ], [ "mole", 3 ], [ "null", 3 ], [ "a", 2 ], [ "mill", 2 ], [ "so", 2 ], [ "some", 2 ] ], "worst_examples": [ { "id": "seed_common_voice_en_205647-common_voice_en_205648", "reference": "The doctor cried after his birth.", "transcript": "The doctor cried after his birth, and they, ant of his fate, crap, and it's, and pename stover, right? He had various, vil tarroid on the very point, and, and telt, they been in wrecks, and fight Dr. Water, so this adjuvant and this dartig, and Dr. Reporter.", "duration_s": 41.12, "metrics": { "wer": 7.0, "cer": 6.625, "mer": 0.875, "wil": 0.875, "wip": 0.125, "word_hits": 6, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 42, "reference_words": 6, "char_hits": 32, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 212, "reference_chars": 32, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_32697372-common_voice_en_32697449", "reference": "The development sits on of real estate.", "transcript": "The development sits on of real estate and was in the very care. So I asked, I think, on fencing a phone. The things was possibly on some ways in those are cases. So, we will, with the felt, with the strenuousness, to probably put that on decisions.", "duration_s": 33.44, "metrics": { "wer": 5.857142857142857, "cer": 5.2894736842105265, "mer": 0.8541666666666666, "wil": 0.8541666666666666, "wip": 0.14583333333333334, "word_hits": 7, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 41, "reference_words": 7, "char_hits": 38, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 201, "reference_chars": 38, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19748971-common_voice_en_19748972", "reference": "Relief carving can be described as \"carving pictures in wood\".", "transcript": "Relief carving can be described as I'm age on mothers to prevent you even den work me on what and the of meal blub and sem disblub um in fricking it is relief carving of longer can to a fit T prevents be a being", "duration_s": 28.32, "metrics": { "wer": 3.8, "cer": 2.6779661016949152, "mer": 0.8444444444444444, "wil": 0.8911111111111111, "wip": 0.10888888888888888, "word_hits": 7, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 35, "reference_words": 10, "char_hits": 51, "char_substitutions": 8, "char_deletions": 0, "char_insertions": 150, "reference_chars": 59, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_24988318-common_voice_en_24988361", "reference": "It had a lower rate of fire and was used as a siege engine.", "transcript": "It had a lower rate of fire and was used as a siege engine. This way the going off through text my indition. This would did you get invented you pen of you for two thousand will cuss in the ploy a smatter right? I a tust hide this ware in the buildings.", "duration_s": 14.4, "metrics": { "wer": 2.7142857142857144, "cer": 3.1551724137931036, "mer": 0.7307692307692307, "wil": 0.7307692307692308, "wip": 0.2692307692307692, "word_hits": 14, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 38, "reference_words": 14, "char_hits": 58, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 183, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_36558935-common_voice_en_36558813", "reference": "Exports fell after the name became closely associated with terrorist attacks.", "transcript": "exports fell after the name became closely associated with terrorist attacks next preventable the vandal hacks water says hon in the tree touts of fallouts from Monica I can tell you what during good on to s-mending have close so then spare Dalitigate to hat.", "duration_s": 26.8, "metrics": { "wer": 3.1818181818181817, "cer": 2.473684210526316, "mer": 0.7608695652173914, "wil": 0.7608695652173914, "wip": 0.2391304347826087, "word_hits": 11, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 35, "reference_words": 11, "char_hits": 76, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 188, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23381412-common_voice_en_23381413", "reference": "After taking Israeli citizenship, the couple adopted Hebrew names: Adam and Aviva Eden.", "transcript": "After taking Israeli citizenship, the couple adopted Hebrew names Chidepeot, Ovi, Moron, Joya, Six, Othuv, Trins, Thouzas, the one of Arthur, but knowing she'd Ova, that was every Adam to all T people, my love, every people, you had one of the wall.", "duration_s": 26.0, "metrics": { "wer": 2.6153846153846154, "cer": 1.8571428571428572, "mer": 0.7727272727272727, "wil": 0.8251748251748252, "wip": 0.17482517482517482, "word_hits": 10, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 31, "reference_words": 13, "char_hits": 80, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 152, "reference_chars": 84, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23998981-common_voice_en_23998980", "reference": "However, there were no widespread reforms of the common law.", "transcript": "However, there were no widespread reforms of the common law. Plus, this lotus-mall-chid-prime-worn-remold-pathety-that-tatcher-cum-your-ree-this-my-wide-of-clough-the-distance-of-birth-of-the-pat-kay-chin-ten-lerve-of-the-far.", "duration_s": 18.56, "metrics": { "wer": 3.3, "cer": 2.810344827586207, "mer": 0.7674418604651163, "wil": 0.7674418604651163, "wip": 0.23255813953488372, "word_hits": 10, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 33, "reference_words": 10, "char_hits": 58, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 163, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_24819370-common_voice_en_24819368", "reference": "It was also ranked the \"Worst Performing Vehicle\".", "transcript": "It was also ranked the job different there because to everything's teeth the other year's motivation when person he beer was sure had cacked it on their cause. Other great the apples. Paul Fowler's.", "duration_s": 14.32, "metrics": { "wer": 4.0, "cer": 3.3617021276595747, "mer": 0.8648648648648649, "wil": 0.9155405405405406, "wip": 0.08445945945945946, "word_hits": 5, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 29, "reference_words": 8, "char_hits": 42, "char_substitutions": 5, "char_deletions": 0, "char_insertions": 153, "reference_chars": 47, "exact_match_rate": 0.0 } } ] }, { "model": "bark", "checkpoint": "suno/bark-small", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.07736749560411957, "cer": 0.049648872225197885, "corpus_metrics_match": true, "word_substitutions": 456, "word_deletions": 128, "word_insertions": 340, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 11, "exact_transcripts": 703, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.2533333333333334, 4.28, 5.266666666666667, 6.986666666666666, 13.211999999999989, 15.0 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 1, "top20_fraction_of_word_errors": 0.3008658008658009, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.2620817843866171, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.07607192254495158, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.04584420424897503, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 6 }, { "reference": "the", "transcript": "a", "count": 4 }, { "reference": "was", "transcript": "is", "count": 3 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "go", "transcript": "knows", "count": 2 }, { "reference": "we", "transcript": "you", "count": 2 }, { "reference": "returned", "transcript": "return", "count": 2 }, { "reference": "wish", "transcript": "which", "count": 2 }, { "reference": "they", "transcript": "he", "count": 2 }, { "reference": "hanged", "transcript": "hind", "count": 2 }, { "reference": "real", "transcript": "dell", "count": 2 }, { "reference": "mabuchi", "transcript": "chee", "count": 2 } ], "common_deletions": [ [ "his", 6 ], [ "is", 5 ], [ "the", 4 ], [ "this", 4 ], [ "he", 3 ], [ "in", 3 ], [ "to", 2 ], [ "will", 2 ] ], "common_insertions": [ [ "but", 7 ], [ "and", 3 ], [ "as", 2 ], [ "the", 2 ], [ "a", 2 ], [ "if", 2 ], [ "mama", 2 ], [ "is", 2 ] ], "worst_examples": [ { "id": "seed_common_voice_en_205647-common_voice_en_205649", "reference": "Don't take any chances.", "transcript": "Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss Don't you kiss", "duration_s": 3.6666666666666665, "metrics": { "wer": 11.6, "cer": 9.73913043478261, "mer": 0.9666666666666667, "wil": 0.9866666666666667, "wip": 0.013333333333333334, "word_hits": 2, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 55, "reference_words": 5, "char_hits": 15, "char_substitutions": 8, "char_deletions": 0, "char_insertions": 216, "reference_chars": 23, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33985575-common_voice_en_33985578", "reference": "He joined the Marine Raiders and was soon promoted to First Lieutenant.", "transcript": "He joined the Marine Raiders and was soon promoted to First Lieutenant. Consistency in his power for a major statement, well he sent me, with abundant second from them.", "duration_s": 14.44, "metrics": { "wer": 1.4166666666666667, "cer": 1.338235294117647, "mer": 0.5862068965517241, "wil": 0.5862068965517242, "wip": 0.41379310344827586, "word_hits": 12, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 17, "reference_words": 12, "char_hits": 68, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 91, "reference_chars": 68, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18522556-common_voice_en_18522557", "reference": "Any tutting that's required, I'll attend to myself.", "transcript": "any two things that require a lot into my sense. For example, I see though, this is the...", "duration_s": 14.44, "metrics": { "wer": 1.6, "cer": 1.0384615384615385, "mer": 0.8421052631578947, "wil": 0.9526315789473684, "wip": 0.047368421052631574, "word_hits": 3, "word_substitutions": 7, "word_deletions": 0, "word_insertions": 9, "reference_words": 10, "char_hits": 33, "char_substitutions": 16, "char_deletions": 3, "char_insertions": 35, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_575287-common_voice_en_575288", "reference": "A young child is enjoying a drink at the supermarket.", "transcript": "The country sports save the order alley of business and areas, being a bad from the core sounding.", "duration_s": 14.426666666666666, "metrics": { "wer": 1.6, "cer": 1.3461538461538463, "mer": 0.8888888888888888, "wil": 0.9777777777777777, "wip": 0.022222222222222223, "word_hits": 2, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 8, "reference_words": 10, "char_hits": 26, "char_substitutions": 26, "char_deletions": 0, "char_insertions": 44, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22575787-common_voice_en_22575799", "reference": "The bridge is located southeast of the town of Molde.", "transcript": "the bridge is located southeast of the town of mold this fella business table easy is more than less more of more it has", "duration_s": 14.44, "metrics": { "wer": 1.5, "cer": 1.3076923076923077, "mer": 0.625, "wil": 0.6625, "wip": 0.3375, "word_hits": 9, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 14, "reference_words": 10, "char_hits": 52, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 68, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19560742-common_voice_en_19560749", "reference": "Back at Stadium mark he once again found himself on the bench.", "transcript": "Back at Stadium Arc, he once again found himself on the bench. There's unfear that Sane Ward getting along to City and Kresh. Making...", "duration_s": 14.426666666666666, "metrics": { "wer": 1.25, "cer": 1.2131147540983607, "mer": 0.5769230769230769, "wil": 0.6121794871794872, "wip": 0.3878205128205128, "word_hits": 11, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 14, "reference_words": 12, "char_hits": 59, "char_substitutions": 1, "char_deletions": 1, "char_insertions": 72, "reference_chars": 61, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20888046-common_voice_en_20888048", "reference": "Fox won her next seven fights in a row, with two knockouts.", "transcript": "seven bites in a row with two knockouts i'm gonna fall through the road now", "duration_s": 14.986666666666666, "metrics": { "wer": 1.1666666666666667, "cer": 1.0784313725490196, "mer": 0.6666666666666666, "wil": 0.7598039215686274, "wip": 0.24019607843137256, "word_hits": 7, "word_substitutions": 1, "word_deletions": 4, "word_insertions": 9, "reference_words": 12, "char_hits": 20, "char_substitutions": 29, "char_deletions": 2, "char_insertions": 24, "reference_chars": 51, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18962809-common_voice_en_18962814", "reference": "The Larson Brothers Airport is located in the town.", "transcript": "the larson brothers airport is located in the town it was cerebral for her and it was you proceeding before in the town", "duration_s": 12.693333333333333, "metrics": { "wer": 1.5555555555555556, "cer": 1.38, "mer": 0.6086956521739131, "wil": 0.6086956521739131, "wip": 0.391304347826087, "word_hits": 9, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 14, "reference_words": 9, "char_hits": 50, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 69, "reference_chars": 50, "exact_match_rate": 0.0 } } ] }, { "model": "vits", "checkpoint": "facebook/mms-tts-eng", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.062295905551369005, "cer": 0.02673629708980539, "corpus_metrics_match": true, "word_substitutions": 606, "word_deletions": 58, "word_insertions": 80, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 662, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.344, 3.552, 4.272, 5.088, 6.196799999999996, 8.096 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.12231182795698925, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.11524163568773234, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.059474412171507604, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.06038017144986955, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.08695652173913043, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 4 }, { "reference": "and", "transcript": "in", "count": 3 }, { "reference": "have", "transcript": "would", "count": 2 }, { "reference": "beard", "transcript": "baird", "count": 2 }, { "reference": "style", "transcript": "still", "count": 2 }, { "reference": "accidents", "transcript": "excitement", "count": 2 }, { "reference": "sony", "transcript": "sunny", "count": 2 }, { "reference": "will", "transcript": "well", "count": 2 }, { "reference": "go", "transcript": "do", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "that", "transcript": "the", "count": 2 }, { "reference": "a sniper", "transcript": "the snapper", "count": 2 } ], "common_deletions": [ [ "to", 2 ], [ "a", 2 ], [ "i", 2 ], [ "it", 2 ], [ "mart", 1 ], [ "alley", 1 ], [ "singer", 1 ], [ "run", 1 ] ], "common_insertions": [ [ "and", 4 ], [ "in", 3 ], [ "be a", 2 ], [ "his", 2 ], [ "a", 2 ], [ "all", 2 ], [ "as", 2 ], [ "is", 2 ] ], "worst_examples": [ { "id": "seed_common_voice_en_18436934-common_voice_en_18436935", "reference": "Biotechnology, or Biotech, is a very popular degree choice", "transcript": "Be a technology or be a ditch is a very popular degree choice.", "duration_s": 4.24, "metrics": { "wer": 0.6666666666666666, "cer": 0.19642857142857142, "mer": 0.46153846153846156, "wil": 0.5811965811965811, "wip": 0.4188034188034188, "word_hits": 7, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 4, "reference_words": 9, "char_hits": 52, "char_substitutions": 2, "char_deletions": 2, "char_insertions": 7, "reference_chars": 56, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_29890305-common_voice_en_29890301", "reference": "His father used to play Dholak at Radio Pakistan Hyderabad.", "transcript": "His father used to play delicate reggae of Peckist and Heterobud.", "duration_s": 3.968, "metrics": { "wer": 0.6, "cer": 0.3448275862068966, "mer": 0.5454545454545454, "wil": 0.7727272727272727, "wip": 0.22727272727272727, "word_hits": 5, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 1, "reference_words": 10, "char_hits": 46, "char_substitutions": 10, "char_deletions": 2, "char_insertions": 8, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_17259121-common_voice_en_17259122", "reference": "This pepperoni tastes off, she said.", "transcript": "This pepper and it has this off she said", "duration_s": 2.736, "metrics": { "wer": 0.8333333333333334, "cer": 0.2647058823529412, "mer": 0.5555555555555556, "wil": 0.7037037037037037, "wip": 0.2962962962962963, "word_hits": 4, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 3, "reference_words": 6, "char_hits": 31, "char_substitutions": 3, "char_deletions": 0, "char_insertions": 6, "reference_chars": 34, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_650392-common_voice_en_650390", "reference": "Kids and adults walk down path in a tree farm.", "transcript": "It's an adult's walk down path in a tree farm.", "duration_s": 3.728, "metrics": { "wer": 0.5, "cer": 0.15555555555555556, "mer": 0.4166666666666667, "wil": 0.5916666666666667, "wip": 0.4083333333333333, "word_hits": 7, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 2, "reference_words": 10, "char_hits": 42, "char_substitutions": 1, "char_deletions": 2, "char_insertions": 4, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18200641-common_voice_en_18200765", "reference": "Mrs. Travis, when I leave my kids in kindergarten, I expect you to supervise them.", "transcript": "Bruce trade is when I leave my kids in kinder garden. I expect you to supervise them", "duration_s": 5.648, "metrics": { "wer": 0.3333333333333333, "cer": 0.13580246913580246, "mer": 0.29411764705882354, "wil": 0.4352941176470587, "wip": 0.5647058823529413, "word_hits": 12, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 2, "reference_words": 15, "char_hits": 73, "char_substitutions": 7, "char_deletions": 1, "char_insertions": 3, "reference_chars": 81, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19211875-common_voice_en_19211870", "reference": "The airport is served by Calm Air and Perimeter Aviation with flights to Winnipeg.", "transcript": "The airport is served by Comair and Pereneter Aviation with flights to win it.", "duration_s": 5.232, "metrics": { "wer": 0.35714285714285715, "cer": 0.1111111111111111, "mer": 0.3333333333333333, "wil": 0.4897959183673469, "wip": 0.5102040816326531, "word_hits": 10, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 1, "reference_words": 14, "char_hits": 72, "char_substitutions": 5, "char_deletions": 4, "char_insertions": 0, "reference_chars": 81, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23976710-common_voice_en_23976707", "reference": "The style of \"The Seven Valleys\" is highly poetic, though not composed in verse.", "transcript": "The soul of the seven volleys is highly, but it is not composed in verse.", "duration_s": 4.528, "metrics": { "wer": 0.35714285714285715, "cer": 0.20833333333333334, "mer": 0.3333333333333333, "wil": 0.5238095238095238, "wip": 0.47619047619047616, "word_hits": 10, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 1, "reference_words": 14, "char_hits": 57, "char_substitutions": 10, "char_deletions": 5, "char_insertions": 0, "reference_chars": 72, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_21762342-common_voice_en_21762343", "reference": "Houston's recording industry had become the center of Texas R and B music.", "transcript": "Alstone's recording industry had become the center of Texas where it had missed", "duration_s": 4.72, "metrics": { "wer": 0.35714285714285715, "cer": 0.24324324324324326, "mer": 0.35714285714285715, "wil": 0.586734693877551, "wip": 0.41326530612244905, "word_hits": 9, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 0, "reference_words": 14, "char_hits": 63, "char_substitutions": 10, "char_deletions": 1, "char_insertions": 7, "reference_chars": 74, "exact_match_rate": 0.0 } } ] }, { "model": "openvoice", "checkpoint": "myshell-ai/OpenVoiceV2", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.03684166457339027, "cer": 0.015904897934892578, "corpus_metrics_match": true, "word_substitutions": 372, "word_deletions": 35, "word_insertions": 33, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 784, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.6253968253968254, 3.84, 4.574331065759637, 5.328979591836735, 6.2693877551020405, 8.301133786848073 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.17727272727272728, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.06691449814126393, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.03653757491931766, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.03242638837122624, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 3 }, { "reference": "this", "transcript": "the", "count": 3 }, { "reference": "he", "transcript": "it", "count": 2 }, { "reference": "the", "transcript": "a", "count": 2 }, { "reference": "named", "transcript": "made", "count": 2 }, { "reference": "off", "transcript": "of", "count": 2 }, { "reference": "hanged", "transcript": "henned", "count": 2 }, { "reference": "have", "transcript": "help", "count": 2 }, { "reference": "heather", "transcript": "header", "count": 2 }, { "reference": "realized", "transcript": "revised", "count": 2 }, { "reference": "in", "transcript": "and", "count": 2 }, { "reference": "tutting", "transcript": "tidings", "count": 1 } ], "common_deletions": [ [ "is", 3 ], [ "mart", 1 ], [ "run", 1 ], [ "arise", 1 ], [ "lash", 1 ], [ "the leaves", 1 ], [ "the", 1 ], [ "on", 1 ] ], "common_insertions": [ [ "is", 2 ], [ "and to", 1 ], [ "carbonese", 1 ], [ "would", 1 ], [ "lemon", 1 ], [ "queen", 1 ], [ "thomas", 1 ], [ "main", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19748004-common_voice_en_19748007", "reference": "Kanwal is said to mean \"snakes indeed\" in a local Aboriginal language.", "transcript": "Can we all accept the means of X and T in a local Aboriginal language?", "duration_s": 4.899410430839002, "metrics": { "wer": 0.8333333333333334, "cer": 0.3582089552238806, "mer": 0.6666666666666666, "wil": 0.8611111111111112, "wip": 0.1388888888888889, "word_hits": 5, "word_substitutions": 7, "word_deletions": 0, "word_insertions": 3, "reference_words": 12, "char_hits": 48, "char_substitutions": 16, "char_deletions": 3, "char_insertions": 5, "reference_chars": 67, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_15734837-common_voice_en_15734838", "reference": "Tyler, Lucy, Michelle, we're going to space!", "transcript": "Tyre, we'll save the show. We're going to space.", "duration_s": 3.5294331065759637, "metrics": { "wer": 0.75, "cer": 0.4878048780487805, "mer": 0.5454545454545454, "wil": 0.7159090909090909, "wip": 0.28409090909090906, "word_hits": 5, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 3, "reference_words": 8, "char_hits": 28, "char_substitutions": 13, "char_deletions": 0, "char_insertions": 7, "reference_chars": 41, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_590785-common_voice_en_590791", "reference": "Angela is pregnant, she told me.", "transcript": "And to let us pregnant, she told me.", "duration_s": 2.9605442176870747, "metrics": { "wer": 0.6666666666666666, "cer": 0.26666666666666666, "mer": 0.5, "wil": 0.6666666666666667, "wip": 0.3333333333333333, "word_hits": 4, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 2, "reference_words": 6, "char_hits": 26, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 4, "reference_chars": 30, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19867710-common_voice_en_19867712", "reference": "Whilst at college Dave, met singer and songwriter Adrian Snell.", "transcript": "While Step College Day met singer and songwriter Adrienne Snell.", "duration_s": 4.504671201814059, "metrics": { "wer": 0.4, "cer": 0.16393442622950818, "mer": 0.4, "wil": 0.64, "wip": 0.36, "word_hits": 6, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 0, "reference_words": 10, "char_hits": 55, "char_substitutions": 4, "char_deletions": 2, "char_insertions": 4, "reference_chars": 61, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19691845-common_voice_en_19691884", "reference": "There is a growing body of case law on Bayh-Dole.", "transcript": "There is a growing body of kiss love on Beidou.", "duration_s": 3.680362811791383, "metrics": { "wer": 0.36363636363636365, "cer": 0.25, "mer": 0.36363636363636365, "wil": 0.5545454545454546, "wip": 0.44545454545454544, "word_hits": 7, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 37, "char_substitutions": 8, "char_deletions": 3, "char_insertions": 1, "reference_chars": 48, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_650392-common_voice_en_650390", "reference": "Kids and adults walk down path in a tree farm.", "transcript": "Hits an adult swap down path in a tree farm.", "duration_s": 3.564263038548753, "metrics": { "wer": 0.4, "cer": 0.15555555555555556, "mer": 0.4, "wil": 0.64, "wip": 0.36, "word_hits": 6, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 0, "reference_words": 10, "char_hits": 39, "char_substitutions": 3, "char_deletions": 3, "char_insertions": 1, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23982302-common_voice_en_23982305", "reference": "Two theories abound to the origin of the name.", "transcript": "Teeth ear is about to the origin of the name", "duration_s": 3.4597732426303853, "metrics": { "wer": 0.4444444444444444, "cer": 0.20930232558139536, "mer": 0.4, "wil": 0.6000000000000001, "wip": 0.39999999999999997, "word_hits": 6, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 9, "char_hits": 37, "char_substitutions": 4, "char_deletions": 2, "char_insertions": 3, "reference_chars": 43, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18962809-common_voice_en_18962813", "reference": "Examples of contract brewers include Epic Brewing Company, Funk Estate and Yeastie Boys.", "transcript": "Examples of contract brewers include Epic Brewing Company, Funka State, and Yes Dig Boys.", "duration_s": 6.176507936507937, "metrics": { "wer": 0.3076923076923077, "cer": 0.06976744186046512, "mer": 0.2857142857142857, "wil": 0.4505494505494505, "wip": 0.5494505494505495, "word_hits": 10, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 81, "char_substitutions": 4, "char_deletions": 1, "char_insertions": 1, "reference_chars": 86, "exact_match_rate": 0.0 } } ] }, { "model": "voxcpm", "checkpoint": "openbmb/VoxCPM2", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.0334924223394457, "cer": 0.02599238231268226, "corpus_metrics_match": true, "word_substitutions": 99, "word_deletions": 14, "word_insertions": 287, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 910, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.96, 3.04, 3.52, 4.0, 5.063999999999978, 9.28 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3525, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.07992565055762081, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.03734439834024896, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.012299664554603056, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "and", "transcript": "in", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "lleida", "transcript": "leida", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "roads", "transcript": "road", "count": 1 }, { "reference": "lash", "transcript": "lashpark", "count": 1 }, { "reference": "iro", "transcript": "eero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "ramblers", "transcript": "rambler", "count": 1 }, { "reference": "mcdonald", "transcript": "mcdarnold", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "with", 2 ], [ "mart", 1 ], [ "run", 1 ], [ "park", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "jumper", 1 ] ], "common_insertions": [ [ "af prebend dolbeitsy", 2 ], [ "what is up inky and the bartons", 1 ], [ "but i i will leak", 1 ], [ "you know", 1 ], [ "i would peddle like the oldie", 1 ], [ "winkel leads a lake in the temporal", 1 ], [ "can", 1 ], [ "teleni kuzbo", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_20798109-common_voice_en_20798113", "reference": "A combination of rock, punk and pop music, this album made the band famous.", "transcript": "A combination of rock, punk and pop music, this album made the band famous. Vauban, Annapisat and Har Beam mixed his stas to Ralinga the Verpratit Singtion.", "duration_s": 9.28, "metrics": { "wer": 0.9285714285714286, "cer": 1.0972222222222223, "mer": 0.48148148148148145, "wil": 0.4814814814814815, "wip": 0.5185185185185185, "word_hits": 14, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 13, "reference_words": 14, "char_hits": 72, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 79, "reference_chars": 72, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20451876-common_voice_en_20451880", "reference": "He did not live to complete the development of his system.", "transcript": "He did not live to complete the development of his system. He did not live to complete the development of his system.", "duration_s": 6.24, "metrics": { "wer": 1.0, "cer": 1.0175438596491229, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 11, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 11, "reference_words": 11, "char_hits": 57, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 58, "reference_chars": 57, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_31005683-common_voice_en_31005688", "reference": "This has caused an outrage in the Serbian public.", "transcript": "This has caused an outrage in the Serbian public. This has caused an outrage in the Serbian public.", "duration_s": 6.88, "metrics": { "wer": 1.0, "cer": 1.0208333333333333, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 9, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 9, "reference_words": 9, "char_hits": 48, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 49, "reference_chars": 48, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_26031217-common_voice_en_26031215", "reference": "It was an important source of slaves at that time.", "transcript": "It was an important source of slaves at that time. An important source of slaves at that time.", "duration_s": 6.4, "metrics": { "wer": 0.8, "cer": 0.8775510204081632, "mer": 0.4444444444444444, "wil": 0.4444444444444444, "wip": 0.5555555555555556, "word_hits": 10, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 8, "reference_words": 10, "char_hits": 49, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 43, "reference_chars": 49, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_275174-common_voice_en_275175", "reference": "The requirements are rather high for this job.", "transcript": "The requirements are rather high for this job. What's up, Inky and the Bartons?", "duration_s": 4.48, "metrics": { "wer": 0.875, "cer": 0.7111111111111111, "mer": 0.4666666666666667, "wil": 0.4666666666666667, "wip": 0.5333333333333333, "word_hits": 8, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 7, "reference_words": 8, "char_hits": 45, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 32, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19770519-common_voice_en_19770516", "reference": "His interests include classical music and literature.", "transcript": "His interests include classical music and literature. Winkel leads a lake in the temporal.", "duration_s": 4.0, "metrics": { "wer": 1.0, "cer": 0.6923076923076923, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 7, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 7, "reference_words": 7, "char_hits": 52, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 36, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_21621452-common_voice_en_21621455", "reference": "It is named after Edward Singleton Holden.", "transcript": "It is named after Edward Singleton Holden. It is named after Edward Singleton Holden.", "duration_s": 6.4, "metrics": { "wer": 1.0, "cer": 1.024390243902439, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 7, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 7, "reference_words": 7, "char_hits": 41, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 42, "reference_chars": 41, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_31892400-common_voice_en_31892402", "reference": "Each Area is further divided into groups.", "transcript": "Each area is further divided into groups. Each area is further divided into groups.", "duration_s": 5.92, "metrics": { "wer": 1.0, "cer": 1.025, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 7, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 7, "reference_words": 7, "char_hits": 40, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 41, "reference_chars": 40, "exact_match_rate": 0.0 } } ] }, { "model": "csm", "checkpoint": "sesame/csm-1b", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.026710206815707946, "cer": 0.014878295542462656, "corpus_metrics_match": true, "word_substitutions": 174, "word_deletions": 33, "word_insertions": 112, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 880, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.64, 3.04, 3.92, 4.88, 6.8, 24.96 ], "high_clipping_recordings": 1, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.26959247648902823, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.024163568773234202, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.026970954356846474, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.026835631755497576, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "their", "transcript": "the", "count": 2 }, { "reference": "underground", "transcript": "boondoground", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "infrastructure", "transcript": "infrastructures", "count": 1 }, { "reference": "wincott", "transcript": "wincart", "count": 1 }, { "reference": "s", "transcript": "arp", "count": 1 }, { "reference": "roaring", "transcript": "boreiring", "count": 1 }, { "reference": "beard", "transcript": "woodbeard", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "favored", "transcript": "favor", "count": 1 } ], "common_deletions": [ [ "and", 2 ], [ "a", 2 ], [ "mart", 1 ], [ "not", 1 ], [ "run", 1 ], [ "park", 1 ], [ "have", 1 ], [ "computer", 1 ] ], "common_insertions": [ [ "the", 5 ], [ "one", 4 ], [ "that", 3 ], [ "so", 2 ], [ "and", 2 ], [ ".", 1 ], [ "muv", 1 ], [ "but", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_21059484-common_voice_en_21059485", "reference": "As he stated; 'I have lost two sisters and you offer me twenty servants.", "transcript": "As he stated, spalsibles conducted, shells notified.", "duration_s": 3.28, "metrics": { "wer": 0.7857142857142857, "cer": 0.6031746031746031, "mer": 0.7857142857142857, "wil": 0.9081632653061225, "wip": 0.09183673469387754, "word_hits": 3, "word_substitutions": 4, "word_deletions": 7, "word_insertions": 0, "reference_words": 14, "char_hits": 27, "char_substitutions": 20, "char_deletions": 16, "char_insertions": 2, "reference_chars": 63, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20933101-common_voice_en_20933104", "reference": "The film was loosely based on \"The Dragon Knight\" by Gordon R. Dickson.", "transcript": "The film was loosely based on Dever, Jang, and Dekarren, Dummy, Nut, the Gout, Abut, by Gordon R. Dixon.", "duration_s": 11.84, "metrics": { "wer": 0.6923076923076923, "cer": 0.6119402985074627, "mer": 0.47368421052631576, "wil": 0.5951417004048583, "wip": 0.4048582995951417, "word_hits": 10, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 6, "reference_words": 13, "char_hits": 56, "char_substitutions": 9, "char_deletions": 2, "char_insertions": 30, "reference_chars": 67, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_9559-common_voice_en_9562", "reference": "This wasn't a strange place; it was a new one.", "transcript": "This wasn't a strange place. Sure. Spurs it. Does me wonder", "duration_s": 24.96, "metrics": { "wer": 0.5454545454545454, "cer": 0.4222222222222222, "mer": 0.46153846153846156, "wil": 0.6287878787878788, "wip": 0.3712121212121212, "word_hits": 7, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 2, "reference_words": 11, "char_hits": 38, "char_substitutions": 7, "char_deletions": 0, "char_insertions": 12, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_21504874-common_voice_en_21504873", "reference": "Maze systems provide users many choices; however, some users may find navigation difficult.", "transcript": "These systems provide users many choices. PaaS use are never. Some users may find navigation difficult.", "duration_s": 5.92, "metrics": { "wer": 0.38461538461538464, "cer": 0.20454545454545456, "mer": 0.3125, "wil": 0.4182692307692307, "wip": 0.5817307692307693, "word_hits": 11, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 3, "reference_words": 13, "char_hits": 82, "char_substitutions": 6, "char_deletions": 0, "char_insertions": 12, "reference_chars": 88, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_32377071-common_voice_en_32377070", "reference": "The name means \"common land\" or \"land shared between local inhabitants\".", "transcript": "The name means one palm in land or one land shared between local inhabitants.", "duration_s": 6.88, "metrics": { "wer": 0.36363636363636365, "cer": 0.19402985074626866, "mer": 0.2857142857142857, "wil": 0.35064935064935066, "wip": 0.6493506493506493, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 3, "reference_words": 11, "char_hits": 63, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 9, "reference_chars": 67, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19916471-common_voice_en_19916473", "reference": "\u2018Coming, Sir,\u2019 replied Sam, running upstairs.", "transcript": "Is it he coming, Solao replied Sam, running upstairs.", "duration_s": 3.76, "metrics": { "wer": 0.6666666666666666, "cer": 0.3333333333333333, "mer": 0.4444444444444444, "wil": 0.537037037037037, "wip": 0.462962962962963, "word_hits": 5, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 3, "reference_words": 6, "char_hits": 37, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 11, "reference_chars": 39, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18754308-common_voice_en_18754311", "reference": "Dopamine neurons have axons that run the entire length of the pathway.", "transcript": "Don't be mean neurons have axons that run the entire length of the pathway.", "duration_s": 4.4, "metrics": { "wer": 0.3333333333333333, "cer": 0.15942028985507245, "mer": 0.26666666666666666, "wil": 0.32777777777777783, "wip": 0.6722222222222222, "word_hits": 11, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 3, "reference_words": 12, "char_hits": 65, "char_substitutions": 3, "char_deletions": 1, "char_insertions": 7, "reference_chars": 69, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_28003405-common_voice_en_28003408", "reference": "Both terms can be literally translated to mean \"inner city\" and \"city centre\".", "transcript": "Both terms can be literally translated to mean double-end city and dub-in-city center.", "duration_s": 5.04, "metrics": { "wer": 0.3076923076923077, "cer": 0.2191780821917808, "mer": 0.25, "wil": 0.3076923076923077, "wip": 0.6923076923076923, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 3, "reference_words": 13, "char_hits": 69, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 12, "reference_chars": 73, "exact_match_rate": 0.0 } } ] }, { "model": "parlertts", "checkpoint": "parler-tts/parler-tts-mini-v1", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.026542744704010718, "cer": 0.0144617032672737, "corpus_metrics_match": true, "word_substitutions": 227, "word_deletions": 43, "word_insertions": 47, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 879, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.9287981859410431, 2.972154195011338, 3.6397278911564626, 4.3769614512471655, 5.38702947845805, 10.193560090702947 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.24605678233438485, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.026022304832713755, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.025242047026279392, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.030935519940365263, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.021739130434782608, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 2 }, { "reference": "he", "transcript": "you", "count": 2 }, { "reference": "bread", "transcript": "rudd", "count": 2 }, { "reference": "mabuchi", "transcript": "mabushi", "count": 2 }, { "reference": "and", "transcript": "in", "count": 2 }, { "reference": "a", "transcript": "his", "count": 2 }, { "reference": "franzen", "transcript": "frunzen", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "infrastructure", "transcript": "antiretes", "count": 1 }, { "reference": "s pheromone trail", "transcript": "tartus from montreal", "count": 1 }, { "reference": "internet", "transcript": "any", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 } ], "common_deletions": [ [ "is", 5 ], [ "a", 4 ], [ "and", 3 ], [ "mart", 1 ], [ "run", 1 ], [ "lash", 1 ], [ "dickson", 1 ], [ "by his barber", 1 ] ], "common_insertions": [ [ "the", 3 ], [ "a", 2 ], [ "in", 2 ], [ "an", 1 ], [ "by", 1 ], [ "sigh", 1 ], [ "anti", 1 ], [ "of course", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't right, it's positive, you kept fire negative.", "duration_s": 2.8792743764172335, "metrics": { "wer": 1.0, "cer": 0.5714285714285714, "mer": 0.6363636363636364, "wil": 0.7922077922077922, "wip": 0.20779220779220778, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 4, "reference_words": 7, "char_hits": 35, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 20, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23809069-common_voice_en_23809074", "reference": "The \"Review\" somehow limped on, buoyed up by a narrow but devoted subscription base.", "transcript": "The Revue somehow limped on Boyd up by an air of a devoted subscription base.", "duration_s": 5.073560090702948, "metrics": { "wer": 0.42857142857142855, "cer": 0.15, "mer": 0.4, "wil": 0.6142857142857143, "wip": 0.38571428571428573, "word_hits": 9, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 1, "reference_words": 14, "char_hits": 69, "char_substitutions": 6, "char_deletions": 5, "char_insertions": 1, "reference_chars": 80, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_32298822-common_voice_en_32298846", "reference": "Sensitivity analysis is part of recent checklists or guidelines for environmental modelling.", "transcript": "Sensitivity analysis is part of recent", "duration_s": 2.5309750566893423, "metrics": { "wer": 0.5, "cer": 0.5777777777777777, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 6, "word_substitutions": 0, "word_deletions": 6, "word_insertions": 0, "reference_words": 12, "char_hits": 38, "char_substitutions": 0, "char_deletions": 52, "char_insertions": 0, "reference_chars": 90, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_29264353-common_voice_en_29264354", "reference": "Fredrik Johansson is featured in the documentary film \"Beyond the Game\".", "transcript": "Frederick Johansson is featured in the documentary filmed by Anne Begain.", "duration_s": 3.924172335600907, "metrics": { "wer": 0.45454545454545453, "cer": 0.2028985507246377, "mer": 0.45454545454545453, "wil": 0.7024793388429753, "wip": 0.29752066115702475, "word_hits": 6, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 0, "reference_words": 11, "char_hits": 60, "char_substitutions": 7, "char_deletions": 2, "char_insertions": 5, "reference_chars": 69, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_652024-common_voice_en_652023", "reference": "Add autry inman to my Hanging Out and Relaxing playlist.", "transcript": "Add autry in man to my hanging out and relaxing through the war.", "duration_s": 3.7384126984126986, "metrics": { "wer": 0.5, "cer": 0.2727272727272727, "mer": 0.38461538461538464, "wil": 0.5076923076923077, "wip": 0.49230769230769234, "word_hits": 8, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 3, "reference_words": 10, "char_hits": 48, "char_substitutions": 7, "char_deletions": 0, "char_insertions": 8, "reference_chars": 55, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_21653057-common_voice_en_21653072", "reference": "The station is owned by the Hearst Television subsidiary of the Hearst Corporation.", "transcript": "The station is owned by the Hairst Casidiary of the Hairst Corporation.", "duration_s": 4.0751020408163265, "metrics": { "wer": 0.3076923076923077, "cer": 0.21951219512195122, "mer": 0.3076923076923077, "wil": 0.48076923076923084, "wip": 0.5192307692307692, "word_hits": 9, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 13, "char_hits": 64, "char_substitutions": 6, "char_deletions": 12, "char_insertions": 0, "reference_chars": 82, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18529345-common_voice_en_18529361", "reference": "Dynamic Link Libraries are reusable modules of windows software applications.", "transcript": "Dynamic Link Libraries are huge windows of Windows of Windows applications.", "duration_s": 4.423401360544218, "metrics": { "wer": 0.4, "cer": 0.2631578947368421, "mer": 0.36363636363636365, "wil": 0.5545454545454546, "wip": 0.44545454545454544, "word_hits": 7, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 10, "char_hits": 60, "char_substitutions": 10, "char_deletions": 6, "char_insertions": 4, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_123125-common_voice_en_123126", "reference": "After all, who doesn\u2019t want to overcome new challenges and achieve great heights?", "transcript": "After all, who doesn't I not want to overcome new challenges and achieve great heights?", "duration_s": 5.073560090702948, "metrics": { "wer": 0.2857142857142857, "cer": 0.08860759493670886, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 12, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 2, "reference_words": 14, "char_hits": 79, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 7, "reference_chars": 79, "exact_match_rate": 0.0 } } ] }, { "model": "melotts", "checkpoint": "EN", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.02637528259231349, "cer": 0.01065285960840326, "corpus_metrics_match": true, "word_substitutions": 268, "word_deletions": 26, "word_insertions": 21, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 845, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.6253968253968254, 3.84, 4.574331065759637, 5.328979591836735, 6.2693877551020405, 8.301133786848073 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.18095238095238095, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.027881040892193308, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.026625172890733056, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.025717480432351846, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "a", "transcript": "the", "count": 3 }, { "reference": "this", "transcript": "the", "count": 3 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "in", "transcript": "and", "count": 2 }, { "reference": "he", "transcript": "you", "count": 2 }, { "reference": "it", "transcript": "i", "count": 2 }, { "reference": "off", "transcript": "of", "count": 2 }, { "reference": "up", "transcript": "out", "count": 2 }, { "reference": "mabuchi", "transcript": "mapuche", "count": 2 }, { "reference": "arranged", "transcript": "ranged", "count": 2 }, { "reference": "have", "transcript": "had", "count": 2 }, { "reference": "parfrey", "transcript": "paffrey", "count": 1 } ], "common_deletions": [ [ "is", 3 ], [ "a", 2 ], [ "mart", 1 ], [ "run", 1 ], [ "lash", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "was", 1 ] ], "common_insertions": [ [ "is", 2 ], [ "and", 2 ], [ "all", 1 ], [ "rheumatism", 1 ], [ "a", 1 ], [ "lemon", 1 ], [ "queen", 1 ], [ "main", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19691845-common_voice_en_19691884", "reference": "There is a growing body of case law on Bayh-Dole.", "transcript": "There is a growing body of kiss love on Beidou.", "duration_s": 3.680362811791383, "metrics": { "wer": 0.36363636363636365, "cer": 0.25, "mer": 0.36363636363636365, "wil": 0.5545454545454546, "wip": 0.44545454545454544, "word_hits": 7, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 37, "char_substitutions": 8, "char_deletions": 3, "char_insertions": 1, "reference_chars": 48, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18962809-common_voice_en_18962813", "reference": "Examples of contract brewers include Epic Brewing Company, Funk Estate and Yeastie Boys.", "transcript": "Examples of contract brewers include Epic Brewing Company, Funke State, and Yes Deep Boys.", "duration_s": 6.176507936507937, "metrics": { "wer": 0.3076923076923077, "cer": 0.08139534883720931, "mer": 0.2857142857142857, "wil": 0.4505494505494505, "wip": 0.5494505494505495, "word_hits": 10, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 82, "char_substitutions": 2, "char_deletions": 2, "char_insertions": 3, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_597578-common_voice_en_597583", "reference": "Wilma\u2019s gone cat sitting again for the weekend if you want to come over.", "transcript": "We'll must come cat sitting again for the weekend if you want to come over.", "duration_s": 4.748480725623582, "metrics": { "wer": 0.26666666666666666, "cer": 0.14084507042253522, "mer": 0.25, "wil": 0.3999999999999999, "wip": 0.6000000000000001, "word_hits": 12, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 15, "char_hits": 66, "char_substitutions": 5, "char_deletions": 0, "char_insertions": 5, "reference_chars": 71, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_586741-common_voice_en_586737", "reference": "Please search The Best of Steely Dan: Then and Now saga.", "transcript": "Please search the best of Steely Danban and Naosaga.", "duration_s": 4.1099319727891155, "metrics": { "wer": 0.36363636363636365, "cer": 0.12962962962962962, "mer": 0.36363636363636365, "wil": 0.505050505050505, "wip": 0.494949494949495, "word_hits": 7, "word_substitutions": 2, "word_deletions": 2, "word_insertions": 0, "reference_words": 11, "char_hits": 48, "char_substitutions": 2, "char_deletions": 4, "char_insertions": 1, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_9559-common_voice_en_9560", "reference": "He didn't consider mending the hole \u2014 the stones could fall through any time they wanted.", "transcript": "They didn't consider meaning the hole the stones could fall through anytime they wanted.", "duration_s": 5.2128798185941045, "metrics": { "wer": 0.25, "cer": 0.05747126436781609, "mer": 0.25, "wil": 0.3999999999999999, "wip": 0.6000000000000001, "word_hits": 12, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 16, "char_hits": 85, "char_substitutions": 0, "char_deletions": 2, "char_insertions": 3, "reference_chars": 87, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19867710-common_voice_en_19867712", "reference": "Whilst at college Dave, met singer and songwriter Adrian Snell.", "transcript": "Well, Step College Dave, Met singer and songwriter Adrian Snow.", "duration_s": 4.504671201814059, "metrics": { "wer": 0.3, "cer": 0.14754098360655737, "mer": 0.3, "wil": 0.51, "wip": 0.48999999999999994, "word_hits": 7, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 6, "char_deletions": 2, "char_insertions": 1, "reference_chars": 61, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive. You're negative.", "duration_s": 2.8908843537414968, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19724242-common_voice_en_19724243", "reference": "The brewery is best known for its Harpoon India Pale Ale.", "transcript": "The brewery is best known for its harpoon Enneapaleo.", "duration_s": 4.284081632653061, "metrics": { "wer": 0.2727272727272727, "cer": 0.14285714285714285, "mer": 0.2727272727272727, "wil": 0.3535353535353536, "wip": 0.6464646464646464, "word_hits": 8, "word_substitutions": 1, "word_deletions": 2, "word_insertions": 0, "reference_words": 11, "char_hits": 48, "char_substitutions": 4, "char_deletions": 4, "char_insertions": 0, "reference_chars": 56, "exact_match_rate": 0.0 } } ] }, { "model": "gptsovits", "checkpoint": "lj1995/GPT-SoVITS", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.02637528259231349, "cer": 0.008703802892340653, "corpus_metrics_match": true, "word_substitutions": 232, "word_deletions": 33, "word_insertions": 50, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 895, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.8, 2.68, 3.36, 4.04, 5.04, 7.28 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.24761904761904763, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.05204460966542751, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.025357307514983864, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.024972046216921357, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "and", "transcript": "in", "count": 5 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "no one", "transcript": "when to", "count": 2 }, { "reference": "returned", "transcript": "return", "count": 2 }, { "reference": "and out", "transcript": "in doubt", "count": 2 }, { "reference": "force times", "transcript": "stein is", "count": 2 }, { "reference": "parfrey", "transcript": "pearfree", "count": 1 }, { "reference": "tutting", "transcript": "aang", "count": 1 }, { "reference": "a", "transcript": "the", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "infrastructure", "transcript": "infrastructures", "count": 1 }, { "reference": "wincott", "transcript": "winkett", "count": 1 } ], "common_deletions": [ [ "and", 4 ], [ "are", 2 ], [ "is", 2 ], [ "mart", 1 ], [ "similar", 1 ], [ "as", 1 ], [ "run in", 1 ], [ "lash", 1 ] ], "common_insertions": [ [ "for", 3 ], [ "we", 2 ], [ "know", 2 ], [ "engine", 2 ], [ "2 d", 1 ], [ "norway", 1 ], [ "bad", 1 ], [ "fill", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_18962809-common_voice_en_18962813", "reference": "Examples of contract brewers include Epic Brewing Company, Funk Estate and Yeastie Boys.", "transcript": "Examples of contract brewers include Depik Brewing Company, Funcus Dayton Yeast Eye eBoys.", "duration_s": 5.52, "metrics": { "wer": 0.46153846153846156, "cer": 0.18604651162790697, "mer": 0.46153846153846156, "wil": 0.7100591715976332, "wip": 0.28994082840236685, "word_hits": 7, "word_substitutions": 6, "word_deletions": 0, "word_insertions": 0, "reference_words": 13, "char_hits": 75, "char_substitutions": 8, "char_deletions": 3, "char_insertions": 5, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22713724-common_voice_en_22713725", "reference": "Since work done is force times distance, mechanical power is force times speed.", "transcript": "Since work done is for Stein's distance, mechanical power is for Stein's speed.", "duration_s": 4.16, "metrics": { "wer": 0.46153846153846156, "cer": 0.18181818181818182, "mer": 0.4, "wil": 0.5846153846153846, "wip": 0.41538461538461535, "word_hits": 9, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 2, "reference_words": 13, "char_hits": 65, "char_substitutions": 12, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_597578-common_voice_en_597583", "reference": "Wilma\u2019s gone cat sitting again for the weekend if you want to come over.", "transcript": "We'll meet us on cat sitting again for the weekend if you want to come over.", "duration_s": 3.6, "metrics": { "wer": 0.3333333333333333, "cer": 0.15492957746478872, "mer": 0.29411764705882354, "wil": 0.4352941176470587, "wip": 0.5647058823529413, "word_hits": 12, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 2, "reference_words": 15, "char_hits": 68, "char_substitutions": 1, "char_deletions": 2, "char_insertions": 8, "reference_chars": 71, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_566283-common_voice_en_566281", "reference": "A thin man lies against the side of the street with his shirt and a shoe off and bags nearby", "transcript": "A thin man lies against the side of the street with his shirt and a shoe off in bags near bygs. Near bygs.", "duration_s": 5.56, "metrics": { "wer": 0.25, "cer": 0.15217391304347827, "mer": 0.21739130434782608, "wil": 0.29565217391304344, "wip": 0.7043478260869566, "word_hits": 18, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 3, "reference_words": 20, "char_hits": 90, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 12, "reference_chars": 92, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_28894436-common_voice_en_28894442", "reference": "Orithyia was later made into the goddess of cold mountain winds.", "transcript": "or H-Y-I-A was later made into the goddess of cold mountain winds.", "duration_s": 4.12, "metrics": { "wer": 0.45454545454545453, "cer": 0.07936507936507936, "mer": 0.3333333333333333, "wil": 0.39393939393939403, "wip": 0.606060606060606, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 4, "reference_words": 11, "char_hits": 60, "char_substitutions": 3, "char_deletions": 0, "char_insertions": 2, "reference_chars": 63, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18105426-common_voice_en_18105428", "reference": "Will no one go?", "transcript": "We'll know when to go.", "duration_s": 0.88, "metrics": { "wer": 1.0, "cer": 0.7857142857142857, "mer": 0.6666666666666666, "wil": 0.8333333333333334, "wip": 0.16666666666666666, "word_hits": 2, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 2, "reference_words": 4, "char_hits": 12, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 9, "reference_chars": 14, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18994734-common_voice_en_18994738", "reference": "When played, failure cards are set aside and not returned to the deck.", "transcript": "When played, fill your cards or set aside and not return to the deck.", "duration_s": 3.92, "metrics": { "wer": 0.3076923076923077, "cer": 0.14705882352941177, "mer": 0.2857142857142857, "wil": 0.4505494505494505, "wip": 0.5494505494505495, "word_hits": 10, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 62, "char_substitutions": 1, "char_deletions": 5, "char_insertions": 4, "reference_chars": 68, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_17429166-common_voice_en_17429168", "reference": "Will no one go?", "transcript": "We'll know when to go.", "duration_s": 0.88, "metrics": { "wer": 1.0, "cer": 0.7857142857142857, "mer": 0.6666666666666666, "wil": 0.8333333333333334, "wip": 0.16666666666666666, "word_hits": 2, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 2, "reference_words": 4, "char_hits": 12, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 9, "reference_chars": 14, "exact_match_rate": 0.0 } } ] }, { "model": "zonos", "checkpoint": "Zyphra/Zonos-v0.1-transformer", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.02302604035836892, "cer": 0.0096411355115158, "corpus_metrics_match": true, "word_substitutions": 178, "word_deletions": 58, "word_insertions": 39, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 873, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.9404081632653061, 3.1579138321995464, 3.959002267573696, 4.760090702947846, 5.828208616780046, 8.219863945578231 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.18545454545454546, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.031598513011152414, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.023397879207007838, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.020499440924338426, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "it", "transcript": "that", "count": 3 }, { "reference": "the", "transcript": "this", "count": 3 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "her", "transcript": "our", "count": 2 }, { "reference": "farm", "transcript": "form", "count": 2 }, { "reference": "real", "transcript": "briel", "count": 2 }, { "reference": "the", "transcript": "a", "count": 2 }, { "reference": "movement", "transcript": "movements", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "hypergamy", "transcript": "hypergome", "count": 1 }, { "reference": "southwest", "transcript": "west", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 } ], "common_deletions": [ [ "the", 11 ], [ "it", 8 ], [ "a", 7 ], [ "is", 2 ], [ "she", 2 ], [ "and", 2 ], [ "he", 2 ], [ "mart", 1 ] ], "common_insertions": [ [ "but", 8 ], [ "and", 3 ], [ "out", 1 ], [ "south", 1 ], [ "he said", 1 ], [ "a", 1 ], [ "queen", 1 ], [ "war", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_593372-common_voice_en_593370", "reference": "I had a memo from Art today.", "transcript": "I had a memo for Mark to do.", "duration_s": 2.1130158730158732, "metrics": { "wer": 0.5714285714285714, "cer": 0.25925925925925924, "mer": 0.5, "wil": 0.7142857142857143, "wip": 0.2857142857142857, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 7, "char_hits": 22, "char_substitutions": 3, "char_deletions": 2, "char_insertions": 2, "reference_chars": 27, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_15734837-common_voice_en_15734838", "reference": "Tyler, Lucy, Michelle, we're going to space!", "transcript": "Lila, Lucy, Michelle, we go into space.", "duration_s": 2.75156462585034, "metrics": { "wer": 0.5, "cer": 0.2682926829268293, "mer": 0.5, "wil": 0.7142857142857143, "wip": 0.2857142857142857, "word_hits": 4, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 30, "char_substitutions": 5, "char_deletions": 6, "char_insertions": 0, "reference_chars": 41, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25821292-common_voice_en_25821294", "reference": "In the final reunion film, Bionic Ever After?", "transcript": "in the final reunion film by Annika Berafter.", "duration_s": 3.2856235827664397, "metrics": { "wer": 0.375, "cer": 0.20930232558139536, "mer": 0.375, "wil": 0.609375, "wip": 0.390625, "word_hits": 5, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 8, "char_hits": 37, "char_substitutions": 4, "char_deletions": 2, "char_insertions": 3, "reference_chars": 43, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_28003405-common_voice_en_28003408", "reference": "Both terms can be literally translated to mean \"inner city\" and \"city centre\".", "transcript": "Both terms can be literally translated to mean in a city centre.", "duration_s": 5.050340136054421, "metrics": { "wer": 0.23076923076923078, "cer": 0.136986301369863, "mer": 0.23076923076923078, "wil": 0.3589743589743589, "wip": 0.6410256410256411, "word_hits": 10, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 13, "char_hits": 63, "char_substitutions": 0, "char_deletions": 10, "char_insertions": 0, "reference_chars": 73, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_487077-common_voice_en_487079", "reference": "I feel sick, what should I do?", "transcript": "Fold the two. I feel sick. What should I do?", "duration_s": 2.2058956916099772, "metrics": { "wer": 0.42857142857142855, "cer": 0.39285714285714285, "mer": 0.3, "wil": 0.30000000000000004, "wip": 0.7, "word_hits": 7, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 3, "reference_words": 7, "char_hits": 28, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 11, "reference_chars": 28, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "you aren't positive, you're negative.", "duration_s": 2.2291156462585033, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19992384-common_voice_en_19992617", "reference": "He then moved to Belgian side Lierse, but has since retired.", "transcript": "Lepithyn moved to Belgian side Leas but has since retired.", "duration_s": 3.7500226757369615, "metrics": { "wer": 0.2727272727272727, "cer": 0.1206896551724138, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 52, "char_substitutions": 4, "char_deletions": 2, "char_insertions": 1, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20938865-common_voice_en_20938866", "reference": "She repealed all her brother's religious laws and imprisoned the country's leading Protestant clerics.", "transcript": "grew pure to all her brother's religious laws and imprisoned the country's leading Protestant clerics.", "duration_s": 5.921088435374149, "metrics": { "wer": 0.1875, "cer": 0.0970873786407767, "mer": 0.17647058823529413, "wil": 0.27941176470588236, "wip": 0.7205882352941176, "word_hits": 14, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 16, "char_hits": 94, "char_substitutions": 8, "char_deletions": 1, "char_insertions": 1, "reference_chars": 103, "exact_match_rate": 0.0 } } ] }, { "model": "orpheustts", "checkpoint": "canopylabs/orpheus-3b-0.1-ft", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.020011722347818808, "cer": 0.007692078795453193, "corpus_metrics_match": true, "word_substitutions": 183, "word_deletions": 33, "word_insertions": 23, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 911, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.9626666666666666, 3.925333333333333, 4.608, 5.376, 6.485333333333333, 9.130666666666666 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.2510460251046025, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.020446096654275093, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.019594283079760258, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.02161759224748416, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "the", "transcript": "a", "count": 4 }, { "reference": "a", "transcript": "the", "count": 2 }, { "reference": "thieves", "transcript": "bees", "count": 2 }, { "reference": "tutting", "transcript": "tarim", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "hypergamy", "transcript": "hykergani", "count": 1 }, { "reference": "s", "transcript": "is", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "we", "transcript": "were", "count": 1 }, { "reference": "lleida", "transcript": "leida", "count": 1 }, { "reference": "cribb", "transcript": "crude", "count": 1 }, { "reference": "cylinder", "transcript": "cylinders", "count": 1 } ], "common_deletions": [ [ "is", 4 ], [ "mart", 1 ], [ "are", 1 ], [ "run", 1 ], [ "s", 1 ], [ "computer", 1 ], [ "dole", 1 ], [ "gates", 1 ] ], "common_insertions": [ [ "thank you", 3 ], [ "tida", 1 ], [ "one", 1 ], [ "the", 1 ], [ "queen", 1 ], [ "chen", 1 ], [ "governor", 1 ], [ "lierre", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_21059484-common_voice_en_21059485", "reference": "As he stated; 'I have lost two sisters and you offer me twenty servants.", "transcript": "As he stated,", "duration_s": 6.741333333333333, "metrics": { "wer": 0.7857142857142857, "cer": 0.8095238095238095, "mer": 0.7857142857142857, "wil": 0.7857142857142857, "wip": 0.21428571428571427, "word_hits": 3, "word_substitutions": 0, "word_deletions": 11, "word_insertions": 0, "reference_words": 14, "char_hits": 12, "char_substitutions": 0, "char_deletions": 51, "char_insertions": 0, "reference_chars": 63, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18944281-common_voice_en_18944280", "reference": "Recordings of Greene are scarce in the compact disc era.", "transcript": "Recordings of green are scarce in a compact disparate.", "duration_s": 3.925333333333333, "metrics": { "wer": 0.4, "cer": 0.16363636363636364, "mer": 0.4, "wil": 0.6000000000000001, "wip": 0.39999999999999997, "word_hits": 6, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 48, "char_substitutions": 3, "char_deletions": 4, "char_insertions": 2, "reference_chars": 55, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20791578-common_voice_en_20791579", "reference": "During that cruise Princess Cruises marked the first ever call in Ilhabela, Brazil.", "transcript": "During that cruise, Princess Cruz is marked the first ever call in Ilha Lepela, Brazil.", "duration_s": 5.12, "metrics": { "wer": 0.3076923076923077, "cer": 0.08860759493670886, "mer": 0.26666666666666666, "wil": 0.3794871794871796, "wip": 0.6205128205128204, "word_hits": 11, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 2, "reference_words": 13, "char_hits": 75, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 3, "reference_chars": 79, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_24988318-common_voice_en_24988314", "reference": "The Canton consists of a mayor and five council members.", "transcript": "Bacchant insists of a mayor and five council members.", "duration_s": 4.181333333333333, "metrics": { "wer": 0.3, "cer": 0.17307692307692307, "mer": 0.3, "wil": 0.4555555555555556, "wip": 0.5444444444444444, "word_hits": 7, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 43, "char_substitutions": 6, "char_deletions": 3, "char_insertions": 0, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18962809-common_voice_en_18962813", "reference": "Examples of contract brewers include Epic Brewing Company, Funk Estate and Yeastie Boys.", "transcript": "Examples of contract brewers include Epic Brewing Company, Funka State, and Yeasty Boys.", "duration_s": 6.314666666666667, "metrics": { "wer": 0.23076923076923078, "cer": 0.046511627906976744, "mer": 0.23076923076923078, "wil": 0.40828402366863903, "wip": 0.591715976331361, "word_hits": 10, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 13, "char_hits": 83, "char_substitutions": 1, "char_deletions": 2, "char_insertions": 1, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 3.2426666666666666, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20461046-common_voice_en_20461051", "reference": "A list of the mythical founders of Germanic peoples follows.", "transcript": "A list of the mythical founders of Germanic people's hollows.", "duration_s": 5.12, "metrics": { "wer": 0.3, "cer": 0.05084745762711865, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 10, "char_hits": 58, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 59, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_29890305-common_voice_en_29890301", "reference": "His father used to play Dholak at Radio Pakistan Hyderabad.", "transcript": "His father used to play dole-lack at Radio Pakistan Haderad.", "duration_s": 4.010666666666666, "metrics": { "wer": 0.3, "cer": 0.13793103448275862, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 10, "char_hits": 54, "char_substitutions": 1, "char_deletions": 3, "char_insertions": 4, "reference_chars": 58, "exact_match_rate": 0.0 } } ] }, { "model": "inflecttts", "checkpoint": "owensong/Inflect-Micro-v2", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.017583521728208994, "cer": 0.0058918050348152114, "corpus_metrics_match": true, "word_substitutions": 178, "word_deletions": 18, "word_insertions": 14, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 932, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.9386666666666666, 3.037333333333333, 3.7226666666666666, 4.437333333333333, 5.382933333333332, 7.509333333333333 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.2714285714285714, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.01858736059479554, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.01751959428307976, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.017890421170331718, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "area", "transcript": "aria", "count": 5 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "his", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "hypergamy", "transcript": "gemi", "count": 1 }, { "reference": "there", "transcript": "though", "count": 1 }, { "reference": "whilst", "transcript": "wollstatt", "count": 1 }, { "reference": "lleida", "transcript": "leyda", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "he", "transcript": "ihasz", "count": 1 }, { "reference": "authored", "transcript": "offered", "count": 1 }, { "reference": "overhaul", "transcript": "overall", "count": 1 } ], "common_deletions": [ [ "mart", 1 ], [ "at", 1 ], [ "run", 1 ], [ "has", 1 ], [ "pond", 1 ], [ "skunk is", 1 ], [ "dole", 1 ], [ "gates", 1 ] ], "common_insertions": [ [ "khyper", 1 ], [ "and", 1 ], [ "queen", 1 ], [ "maine", 1 ], [ "it", 1 ], [ "governor", 1 ], [ "mound", 1 ], [ "life", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_597578-common_voice_en_597583", "reference": "Wilma\u2019s gone cat sitting again for the weekend if you want to come over.", "transcript": "Well must concat sitting again for the week and if you want to come over.", "duration_s": 3.8613333333333335, "metrics": { "wer": 0.4, "cer": 0.15492957746478872, "mer": 0.375, "wil": 0.5555555555555556, "wip": 0.4444444444444444, "word_hits": 10, "word_substitutions": 4, "word_deletions": 1, "word_insertions": 1, "reference_words": 15, "char_hits": 63, "char_substitutions": 6, "char_deletions": 2, "char_insertions": 3, "reference_chars": 71, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18962809-common_voice_en_18962813", "reference": "Examples of contract brewers include Epic Brewing Company, Funk Estate and Yeastie Boys.", "transcript": "Examples of contract brewers include Epic Brewing Company, Funka State, and Yeasty Boise.", "duration_s": 5.504, "metrics": { "wer": 0.3076923076923077, "cer": 0.06976744186046512, "mer": 0.3076923076923077, "wil": 0.5207100591715976, "wip": 0.4792899408284023, "word_hits": 9, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 0, "reference_words": 13, "char_hits": 82, "char_substitutions": 2, "char_deletions": 2, "char_insertions": 2, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20933101-common_voice_en_20933102", "reference": "To have darshan one has to swim to the temple.", "transcript": "The Hamdarshan won't has to swim to the temple.", "duration_s": 2.592, "metrics": { "wer": 0.4, "cer": 0.26666666666666666, "mer": 0.4, "wil": 0.64, "wip": 0.36, "word_hits": 6, "word_substitutions": 4, "word_deletions": 0, "word_insertions": 0, "reference_words": 10, "char_hits": 39, "char_substitutions": 4, "char_deletions": 2, "char_insertions": 6, "reference_chars": 45, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20030172-common_voice_en_20030174", "reference": "Various ecclesiastical areas, too small for a diocese, are termed prefects.", "transcript": "Various ecclesiastical arias, two small fura diocese, are termed prefects.", "duration_s": 5.066666666666666, "metrics": { "wer": 0.36363636363636365, "cer": 0.08333333333333333, "mer": 0.36363636363636365, "wil": 0.5545454545454546, "wip": 0.44545454545454544, "word_hits": 7, "word_substitutions": 3, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 66, "char_substitutions": 3, "char_deletions": 3, "char_insertions": 0, "reference_chars": 72, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25190404-common_voice_en_25190411", "reference": "He has also authored a number of books on the subject.", "transcript": "Ihasz also offered a number of books on the subject.", "duration_s": 3.0613333333333332, "metrics": { "wer": 0.2727272727272727, "cer": 0.16981132075471697, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 45, "char_substitutions": 5, "char_deletions": 3, "char_insertions": 1, "reference_chars": 53, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19622969-common_voice_en_19622971", "reference": "A skunk's spray is powerful enough to ward off bears and other potential attackers.", "transcript": "Eskunk spray is powerful enough to ward off bears and other potential attackers.", "duration_s": 4.682666666666667, "metrics": { "wer": 0.2, "cer": 0.060240963855421686, "mer": 0.2, "wil": 0.2615384615384615, "wip": 0.7384615384615385, "word_hits": 12, "word_substitutions": 1, "word_deletions": 2, "word_insertions": 0, "reference_words": 15, "char_hits": 78, "char_substitutions": 1, "char_deletions": 4, "char_insertions": 0, "reference_chars": 83, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18699078-common_voice_en_18699081", "reference": "The national monument surrounds the Castle Mountain Mine Area.", "transcript": "The national monument surrounds the castle mound and mine aria.", "duration_s": 3.562666666666667, "metrics": { "wer": 0.3333333333333333, "cer": 0.08196721311475409, "mer": 0.3, "wil": 0.4555555555555556, "wip": 0.5444444444444444, "word_hits": 7, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 9, "char_hits": 57, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 1, "reference_chars": 61, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.048, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } } ] }, { "model": "higgstts", "checkpoint": "bosonai/higgs-tts-2-3b-base", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.017416059616511763, "cer": 0.006710111289650658, "corpus_metrics_match": true, "word_substitutions": 150, "word_deletions": 34, "word_insertions": 24, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 935, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.76, 3.08, 3.88, 4.76, 6.28, 8.92 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.27884615384615385, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.013011152416356878, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.017750115260488704, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.017517704062616473, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "the", "transcript": "a", "count": 4 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "franzen", "transcript": "franson", "count": 1 }, { "reference": "a", "transcript": "poi", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "snell", "transcript": "snow", "count": 1 }, { "reference": "lleida", "transcript": "leda", "count": 1 }, { "reference": "had", "transcript": "would", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "her", "transcript": "our", "count": 1 }, { "reference": "octavian", "transcript": "tavian", "count": 1 } ], "common_deletions": [ [ "a", 6 ], [ "it", 2 ], [ "will", 2 ], [ "toy", 1 ], [ "mart", 1 ], [ "run", 1 ], [ "styles", 1 ], [ "lash", 1 ] ], "common_insertions": [ [ "would", 2 ], [ "the", 2 ], [ "queen", 1 ], [ "principle", 1 ], [ "maine", 1 ], [ "or", 1 ], [ "governor", 1 ], [ "hard", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_603760-common_voice_en_603759", "reference": "He remembered something his grandfather had once told him: that butterflies were a good omen.", "transcript": "he remembered something his grandfather had once told him. That aisle which you coin over your lap", "duration_s": 5.36, "metrics": { "wer": 0.4666666666666667, "cer": 0.3076923076923077, "mer": 0.4117647058823529, "wil": 0.607843137254902, "wip": 0.39215686274509803, "word_hits": 10, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 2, "reference_words": 15, "char_hits": 70, "char_substitutions": 20, "char_deletions": 1, "char_insertions": 7, "reference_chars": 91, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_21319256-common_voice_en_21319257", "reference": "Louis' famous response: He can run, but he can't hide.", "transcript": "Louis' famous response. He can run.", "duration_s": 4.28, "metrics": { "wer": 0.45454545454545453, "cer": 0.38461538461538464, "mer": 0.45454545454545453, "wil": 0.4545454545454546, "wip": 0.5454545454545454, "word_hits": 6, "word_substitutions": 0, "word_deletions": 5, "word_insertions": 0, "reference_words": 11, "char_hits": 32, "char_substitutions": 0, "char_deletions": 20, "char_insertions": 0, "reference_chars": 52, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23726733-common_voice_en_23726734", "reference": "She was also a \"full partner\" in his espionage.", "transcript": "She was also alive in this espionage.", "duration_s": 3.2, "metrics": { "wer": 0.4444444444444444, "cer": 0.2727272727272727, "mer": 0.4444444444444444, "wil": 0.6031746031746031, "wip": 0.39682539682539686, "word_hits": 5, "word_substitutions": 2, "word_deletions": 2, "word_insertions": 0, "reference_words": 9, "char_hits": 33, "char_substitutions": 2, "char_deletions": 9, "char_insertions": 1, "reference_chars": 44, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_554379-common_voice_en_554380", "reference": "I got drunk yesterday and thought I built a super computer.", "transcript": "I got drunk yesterday and thought I'd build a supercomputer.", "duration_s": 3.6, "metrics": { "wer": 0.36363636363636365, "cer": 0.13793103448275862, "mer": 0.3333333333333333, "wil": 0.47107438016528924, "wip": 0.5289256198347108, "word_hits": 8, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 1, "reference_words": 11, "char_hits": 56, "char_substitutions": 1, "char_deletions": 1, "char_insertions": 6, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_2586258-common_voice_en_2586260", "reference": "Oh, what are you going to when he arrives?", "transcript": "or where you're going to when he arrives.", "duration_s": 2.04, "metrics": { "wer": 0.4444444444444444, "cer": 0.2564102564102564, "mer": 0.4, "wil": 0.5555555555555556, "wip": 0.4444444444444444, "word_hits": 6, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 1, "reference_words": 9, "char_hits": 34, "char_substitutions": 2, "char_deletions": 3, "char_insertions": 5, "reference_chars": 39, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19977527-common_voice_en_19977526", "reference": "That same year, a civilian executive version, Disponent, saw the light of day.", "transcript": "That same year, a civilian executive version of the despot then saw the light of day.", "duration_s": 4.08, "metrics": { "wer": 0.3076923076923077, "cer": 0.17567567567567569, "mer": 0.25, "wil": 0.3076923076923077, "wip": 0.6923076923076923, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 3, "reference_words": 13, "char_hits": 71, "char_substitutions": 2, "char_deletions": 1, "char_insertions": 10, "reference_chars": 74, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 1.6, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19992384-common_voice_en_19992617", "reference": "He then moved to Belgian side Lierse, but has since retired.", "transcript": "He then moved to Belgianside, Leas, but has since retired.", "duration_s": 3.68, "metrics": { "wer": 0.2727272727272727, "cer": 0.06896551724137931, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 54, "char_substitutions": 1, "char_deletions": 3, "char_insertions": 0, "reference_chars": 58, "exact_match_rate": 0.0 } } ] }, { "model": "neutts", "checkpoint": "neuphonic/neutts-2e", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.014736665829356108, "cer": 0.006204249241206927, "corpus_metrics_match": true, "word_substitutions": 116, "word_deletions": 24, "word_insertions": 36, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 1, "exact_transcripts": 967, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.08, 3.375, 4.08, 4.78, 5.732999999999997, 11.52 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.39204545454545453, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.016728624535315983, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.012102351313969572, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.023108460678345134, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "jolliest", "transcript": "joliest", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roaring", "transcript": "warring", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "colonel", "transcript": "colin", "count": 1 }, { "reference": "lleida", "transcript": "leida", "count": 1 }, { "reference": "don", "transcript": "donlash", "count": 1 }, { "reference": "iro", "transcript": "euro", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "stridulation", "transcript": "straggulation", "count": 1 }, { "reference": "tao", "transcript": "dao", "count": 1 } ], "common_deletions": [ [ "banana bread", 2 ], [ "mart", 1 ], [ "lash", 1 ], [ "his", 1 ], [ "is", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "irk", 1 ] ], "common_insertions": [ [ "is", 1 ], [ "and becomes with may challenge and proper teen 2 bacon eye teens for b a t e.2 with the commons cons fops it is fair", 1 ], [ "maine", 1 ], [ "for black readers", 1 ], [ "governor", 1 ], [ "hobbit", 1 ], [ "hang", 1 ], [ "de", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queens College, Cambridge where he read English with French. And, becomes with May Challenge and Proper Teen 2 Bacon Eye Teens for B.A.T.E.2 with the Commons Cons Fops It's Fair.", "duration_s": 11.52, "metrics": { "wer": 1.9230769230769231, "cer": 1.5194805194805194, "mer": 0.6578947368421053, "wil": 0.6578947368421053, "wip": 0.34210526315789475, "word_hits": 13, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 25, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 117, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_24814747-common_voice_en_24814749", "reference": "Ross also wrote a novelization of the theatrical film \"House of Dark Shadows\".", "transcript": "Ross also wrote a novelization of the theatrical film.", "duration_s": 3.52, "metrics": { "wer": 0.3076923076923077, "cer": 0.29333333333333333, "mer": 0.3076923076923077, "wil": 0.3076923076923077, "wip": 0.6923076923076923, "word_hits": 9, "word_substitutions": 0, "word_deletions": 4, "word_insertions": 0, "reference_words": 13, "char_hits": 53, "char_substitutions": 0, "char_deletions": 22, "char_insertions": 0, "reference_chars": 75, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_30619661-common_voice_en_30619662", "reference": "Campbell helped establish branch libraries for Black readers.", "transcript": "Campbell helped establish branch libraries for Black readers. For Black readers.", "duration_s": 5.62, "metrics": { "wer": 0.375, "cer": 0.3, "mer": 0.2727272727272727, "wil": 0.2727272727272727, "wip": 0.7272727272727273, "word_hits": 8, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 3, "reference_words": 8, "char_hits": 60, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 18, "reference_chars": 60, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.48, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18786931-common_voice_en_18786930", "reference": "Immediately following the demerger, Williams Holdings launched a takeover bid for Racal.", "transcript": "Immediately following the de-murger, Williams Holdings launched a takeover bid for Racel.", "duration_s": 5.04, "metrics": { "wer": 0.25, "cer": 0.03488372093023256, "mer": 0.23076923076923078, "wil": 0.3589743589743589, "wip": 0.6410256410256411, "word_hits": 10, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 84, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 1, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19211875-common_voice_en_19211870", "reference": "The airport is served by Calm Air and Perimeter Aviation with flights to Winnipeg.", "transcript": "The airport is served by Comair and Perenor Aviation with flights to Winnipeg.", "duration_s": 5.16, "metrics": { "wer": 0.21428571428571427, "cer": 0.08641975308641975, "mer": 0.21428571428571427, "wil": 0.3351648351648352, "wip": 0.6648351648351648, "word_hits": 11, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 14, "char_hits": 74, "char_substitutions": 3, "char_deletions": 4, "char_insertions": 0, "reference_chars": 81, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 5.66, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_27794266-common_voice_en_27794264", "reference": "Don Lash Park in Auburn is named for him.", "transcript": "Donlash Park in Auburn is named for him.", "duration_s": 3.4, "metrics": { "wer": 0.2222222222222222, "cer": 0.025, "mer": 0.2222222222222222, "wil": 0.3194444444444444, "wip": 0.6805555555555556, "word_hits": 7, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 9, "char_hits": 39, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 40, "exact_match_rate": 0.0 } } ] }, { "model": "qwen3tts", "checkpoint": "Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.013815624215021352, "cer": 0.0043890971850264835, "corpus_metrics_match": true, "word_substitutions": 137, "word_deletions": 17, "word_insertions": 11, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 961, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.12, 3.82, 4.88, 6.0, 7.84, 12.16 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.28484848484848485, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.016728624535315983, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.013485477178423237, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.014535967200894522, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "this", "transcript": "the", "count": 4 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "farm", "transcript": "firm", "count": 2 }, { "reference": "thieves", "transcript": "themes", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "critics", "transcript": "kukiks", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "we", "transcript": "but", "count": 1 }, { "reference": "lleida", "transcript": "leda", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "norwegia", "transcript": "norvegia", "count": 1 } ], "common_deletions": [ [ "mart", 1 ], [ "are", 1 ], [ "run", 1 ], [ "is", 1 ], [ "computer", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "operate", 1 ] ], "common_insertions": [ [ ".", 1 ], [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "to", 1 ], [ "sub", 1 ], [ "farsi", 1 ], [ "out", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_615162-common_voice_en_615165", "reference": "In the past I've had trouble prioritizing stuff in this list.", "transcript": "In the past, I've had to blay way your to this listen.", "duration_s": 3.92, "metrics": { "wer": 0.5, "cer": 0.3870967741935484, "mer": 0.46153846153846156, "wil": 0.6858974358974359, "wip": 0.3141025641025641, "word_hits": 7, "word_substitutions": 5, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 41, "char_substitutions": 10, "char_deletions": 11, "char_insertions": 3, "reference_chars": 62, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20806032-common_voice_en_20806033", "reference": "These two agencies co-operate in aspects of law enforcement, including their joint marine unit.", "transcript": "These two agencies cooperate in aspects of law enforcement, including their joint marine units.", "duration_s": 7.28, "metrics": { "wer": 0.2, "cer": 0.02197802197802198, "mer": 0.2, "wil": 0.3142857142857143, "wip": 0.6857142857142857, "word_hits": 12, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 90, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 1, "reference_chars": 91, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive. You're negative.", "duration_s": 2.48, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_2586258-common_voice_en_2586259", "reference": "You can create bookmarks in you web browser to make browsing more convenient.", "transcript": "You can create bookmarks in uWebBrowser to make browsing more convenient.", "duration_s": 4.64, "metrics": { "wer": 0.23076923076923078, "cer": 0.05263157894736842, "mer": 0.23076923076923078, "wil": 0.30069930069930073, "wip": 0.6993006993006993, "word_hits": 10, "word_substitutions": 1, "word_deletions": 2, "word_insertions": 0, "reference_words": 13, "char_hits": 72, "char_substitutions": 0, "char_deletions": 4, "char_insertions": 0, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 5.76, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_26952530-common_voice_en_26952588", "reference": "Critics have taken Roney to task for sounding too similar to his idol.", "transcript": "Kukiks have taken Roni to task for sounding too similar to his idol.", "duration_s": 4.88, "metrics": { "wer": 0.15384615384615385, "cer": 0.10144927536231885, "mer": 0.15384615384615385, "wil": 0.28402366863905326, "wip": 0.7159763313609467, "word_hits": 11, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 13, "char_hits": 62, "char_substitutions": 5, "char_deletions": 2, "char_insertions": 0, "reference_chars": 69, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22972881-common_voice_en_22972884", "reference": "We're too much the ideal bad examples.", "transcript": "But too much the ideal bad examples.", "duration_s": 3.04, "metrics": { "wer": 0.25, "cer": 0.15789473684210525, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 32, "char_substitutions": 3, "char_deletions": 3, "char_insertions": 0, "reference_chars": 38, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 4.32, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } } ] }, { "model": "vibevoice", "checkpoint": "microsoft/VibeVoice-Realtime-0.5B", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.013145775768232438, "cer": 0.004151044456347081, "corpus_metrics_match": true, "word_substitutions": 122, "word_deletions": 20, "word_insertions": 15, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 960, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.3333333333333333, 3.7333333333333334, 4.533333333333333, 5.333333333333333, 6.533333333333333, 8.666666666666666 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.27388535031847133, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.013011152416356878, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.01256339326878746, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.01528140141632501, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "we", "transcript": "were", "count": 1 }, { "reference": "lleida", "transcript": "leyda", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "iro", "transcript": "eero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "restriction", "transcript": "restrictions", "count": 1 }, { "reference": "lancaster", "transcript": "lanxer", "count": 1 }, { "reference": "coral", "transcript": "choral", "count": 1 }, { "reference": "off", "transcript": "soft", "count": 1 } ], "common_deletions": [ [ "is", 3 ], [ "mart", 1 ], [ "are", 1 ], [ "run", 1 ], [ "computer", 1 ], [ "gates", 1 ], [ "would", 1 ], [ "operate", 1 ] ], "common_insertions": [ [ "the", 3 ], [ "queen", 1 ], [ "bo", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "function", 1 ], [ "near", 1 ], [ "t", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive. You're negative.", "duration_s": 2.533333333333333, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_2586258-common_voice_en_2586259", "reference": "You can create bookmarks in you web browser to make browsing more convenient.", "transcript": "You can create bookmarks in uWebBrowser to make browsing more convenient.", "duration_s": 5.466666666666667, "metrics": { "wer": 0.23076923076923078, "cer": 0.05263157894736842, "mer": 0.23076923076923078, "wil": 0.30069930069930073, "wip": 0.6993006993006993, "word_hits": 10, "word_substitutions": 1, "word_deletions": 2, "word_insertions": 0, "reference_words": 13, "char_hits": 72, "char_substitutions": 0, "char_deletions": 4, "char_insertions": 0, "reference_chars": 76, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_566283-common_voice_en_566281", "reference": "A thin man lies against the side of the street with his shirt and a shoe off and bags nearby", "transcript": "A thin man lies against the side of the street, with his shirt and his shoe off, and bags near by.", "duration_s": 5.866666666666666, "metrics": { "wer": 0.15, "cer": 0.043478260869565216, "mer": 0.14285714285714285, "wil": 0.22857142857142865, "wip": 0.7714285714285714, "word_hits": 18, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 20, "char_hits": 91, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 3, "reference_chars": 92, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 6.133333333333334, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22972881-common_voice_en_22972884", "reference": "We're too much the ideal bad examples.", "transcript": "were too much the ideal bad examples.", "duration_s": 2.8, "metrics": { "wer": 0.25, "cer": 0.05263157894736842, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 36, "char_substitutions": 0, "char_deletions": 2, "char_insertions": 0, "reference_chars": 38, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 4.0, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge, where he read English with French.", "duration_s": 5.066666666666666, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_103675-common_voice_en_103677", "reference": "The boy knew the desert sensed his fear.", "transcript": "The Bo'oi knew the desert sensed his fear.", "duration_s": 3.2, "metrics": { "wer": 0.25, "cer": 0.07692307692307693, "mer": 0.2222222222222222, "wil": 0.3194444444444444, "wip": 0.6805555555555556, "word_hits": 7, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 8, "char_hits": 38, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 39, "exact_match_rate": 0.0 } } ] }, { "model": "outetts", "checkpoint": "OuteAI/Llama-OuteTTS-1.0-1B", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.012392196265594908, "cer": 0.00437421888948402, "corpus_metrics_match": true, "word_substitutions": 105, "word_deletions": 28, "word_insertions": 15, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 972, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.7996666666666667, 3.973, 4.693, 5.439666666666667, 6.4883333333333315, 8.253 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.33783783783783783, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.00929368029739777, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.011871830336560626, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.014908684308609765, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "lleida", "transcript": "leder", "count": 1 }, { "reference": "bartley", "transcript": "barclay", "count": 1 }, { "reference": "child", "transcript": "children", "count": 1 }, { "reference": "him", "transcript": "into", "count": 1 }, { "reference": "iro", "transcript": "eero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "coral", "transcript": "choral", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 }, { "reference": "stamps", "transcript": "stamp", "count": 1 }, { "reference": "for", "transcript": "to", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "missus", 2 ], [ "mart", 1 ], [ "will", 1 ], [ "flags in certain", 1 ], [ "to", 1 ], [ "in", 1 ], [ "computer", 1 ] ], "common_insertions": [ [ "ah", 2 ], [ "k", 1 ], [ "a", 1 ], [ ".", 1 ], [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "or", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_21589881-common_voice_en_21589886", "reference": "The language from which the lexicon is derived is called the \"lexifier\".", "transcript": "The language from which the lexifier", "duration_s": 3.3596666666666666, "metrics": { "wer": 0.5, "cer": 0.4782608695652174, "mer": 0.5, "wil": 0.5, "wip": 0.5, "word_hits": 6, "word_substitutions": 0, "word_deletions": 6, "word_insertions": 0, "reference_words": 12, "char_hits": 36, "char_substitutions": 0, "char_deletions": 33, "char_insertions": 0, "reference_chars": 69, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20531306-common_voice_en_20531308", "reference": "His alleged code name in Soviet intelligence and in the Venona files is \"Ted\".", "transcript": "His alleged codename in Soviet intelligence and in the Venonophiles is Ted.", "duration_s": 5.799666666666667, "metrics": { "wer": 0.2857142857142857, "cer": 0.05333333333333334, "mer": 0.2857142857142857, "wil": 0.40476190476190477, "wip": 0.5952380952380952, "word_hits": 10, "word_substitutions": 2, "word_deletions": 2, "word_insertions": 0, "reference_words": 14, "char_hits": 71, "char_substitutions": 3, "char_deletions": 1, "char_insertions": 0, "reference_chars": 75, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20252760-common_voice_en_20252766", "reference": "There are three colour combinations that are used on several flags in certain regions.", "transcript": "There are three color combinations that are used on several regions.", "duration_s": 5.079666666666666, "metrics": { "wer": 0.21428571428571427, "cer": 0.2125, "mer": 0.21428571428571427, "wil": 0.2142857142857143, "wip": 0.7857142857142857, "word_hits": 11, "word_substitutions": 0, "word_deletions": 3, "word_insertions": 0, "reference_words": 14, "char_hits": 63, "char_substitutions": 0, "char_deletions": 17, "char_insertions": 0, "reference_chars": 80, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 3.333, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19730681-common_voice_en_19730680", "reference": "Underwater demolition specialists may still be referred to as underwater demolition teams.", "transcript": "Oh, well water demolition specialists may still be referred to as underwater demolition teams.", "duration_s": 7.119666666666666, "metrics": { "wer": 0.25, "cer": 0.06741573033707865, "mer": 0.21428571428571427, "wil": 0.27976190476190477, "wip": 0.7202380952380952, "word_hits": 11, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 2, "reference_words": 12, "char_hits": 85, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 2, "reference_chars": 89, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18786931-common_voice_en_18786930", "reference": "Immediately following the demerger, Williams Holdings launched a takeover bid for Racal.", "transcript": "Immediately following the de Mergill, Williams Holdings launched a takeover bid for Raquel.", "duration_s": 6.079666666666666, "metrics": { "wer": 0.25, "cer": 0.08139534883720931, "mer": 0.23076923076923078, "wil": 0.3589743589743589, "wip": 0.6410256410256411, "word_hits": 10, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 82, "char_substitutions": 4, "char_deletions": 0, "char_insertions": 3, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 6.053, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_38573-common_voice_en_38572", "reference": "The desert, with its endless monotony, put him to dreaming.", "transcript": "The desert, with its endless monotony, put into dreaming.", "duration_s": 4.759666666666667, "metrics": { "wer": 0.2, "cer": 0.05357142857142857, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 1, "char_deletions": 2, "char_insertions": 0, "reference_chars": 56, "exact_match_rate": 0.0 } } ] }, { "model": "chatterbox", "checkpoint": "ResembleAI/chatterbox", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.012308465209746294, "cer": 0.0035261560435636496, "corpus_metrics_match": true, "word_substitutions": 122, "word_deletions": 17, "word_insertions": 8, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 963, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.96, 2.64, 3.16, 3.76, 4.6, 5.8 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.2789115646258503, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.00929368029739777, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.012678653757491932, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.011926947446887812, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "the", "transcript": "a", "count": 3 }, { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "and", "transcript": "in", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "snell", "transcript": "snow", "count": 1 }, { "reference": "lleida", "transcript": "leida", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "make", "transcript": "makes", "count": 1 }, { "reference": "roads", "transcript": "road", "count": 1 }, { "reference": "iro", "transcript": "eero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 } ], "common_deletions": [ [ "is", 4 ], [ "mart", 1 ], [ "run", 1 ], [ "computer", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "operate", 1 ], [ "shot", 1 ] ], "common_insertions": [ [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "that", 1 ], [ "t", 1 ], [ "where", 1 ], [ "is", 1 ], [ "and", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 1.92, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 4.12, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 3.0, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge, where he read English with French.", "duration_s": 4.12, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "its most prominent representatives gathered in the Maine's Fine Arts Association.", "duration_s": 3.96, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 78, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_554379-common_voice_en_554380", "reference": "I got drunk yesterday and thought I built a super computer.", "transcript": "I got drunk yesterday and thought I built a supercomputer.", "duration_s": 2.72, "metrics": { "wer": 0.18181818181818182, "cer": 0.017241379310344827, "mer": 0.18181818181818182, "wil": 0.26363636363636356, "wip": 0.7363636363636364, "word_hits": 9, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 57, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19691845-common_voice_en_19691884", "reference": "There is a growing body of case law on Bayh-Dole.", "transcript": "There is a growing body of case law on Baidol.", "duration_s": 2.28, "metrics": { "wer": 0.18181818181818182, "cer": 0.08333333333333333, "mer": 0.18181818181818182, "wil": 0.26363636363636356, "wip": 0.7363636363636364, "word_hits": 9, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 44, "char_substitutions": 1, "char_deletions": 3, "char_insertions": 0, "reference_chars": 48, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20065698-common_voice_en_20065701", "reference": "Flood gates were erected and drainage systems improved.", "transcript": "Floodgates were erected and drainage systems improved.", "duration_s": 2.72, "metrics": { "wer": 0.25, "cer": 0.018518518518518517, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } } ] }, { "model": "mosstts", "checkpoint": "OpenMOSS-Team/MOSS-TTS-v1.5", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.012308465209746294, "cer": 0.00431470570731417, "corpus_metrics_match": true, "word_substitutions": 115, "word_deletions": 22, "word_insertions": 10, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 1, "exact_transcripts": 988, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.12, 4.32, 5.44, 6.24, 8.08, 113.12 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 1, "top20_fraction_of_word_errors": 0.40816326530612246, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.011152416356877323, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.012217611802674043, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.013045098770033545, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "ronnie", "count": 1 }, { "reference": "lleida", "transcript": "lida", "count": 1 }, { "reference": "as catch", "transcript": "us ketchkin", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "norwegia", "transcript": "norwidgia", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "ramblers", "transcript": "rambler", "count": 1 }, { "reference": "coral", "transcript": "choral", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 } ], "common_deletions": [ [ "is", 3 ], [ "a", 2 ], [ "mart", 1 ], [ "can", 1 ], [ "run", 1 ], [ "computer", 1 ], [ "law", 1 ], [ "dole", 1 ] ], "common_insertions": [ [ "queen", 1 ], [ "have", 1 ], [ "man", 1 ], [ "thank", 1 ], [ "governor", 1 ], [ "on", 1 ], [ "natar", 1 ], [ "to", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_21499767-common_voice_en_21499784", "reference": "The crown is conical, becoming a narrow rounded dome with age.", "transcript": "Thank you. Thank you. Thank you. Thank you.", "duration_s": 95.44, "metrics": { "wer": 1.0, "cer": 0.7666666666666667, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 8, "word_deletions": 3, "word_insertions": 0, "reference_words": 11, "char_hits": 14, "char_substitutions": 25, "char_deletions": 21, "char_insertions": 0, "reference_chars": 60, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20030172-common_voice_en_20030173", "reference": "The children of Gowanda had school that day.", "transcript": "Thank you. Thank you. Thank you. Thanks for watching!", "duration_s": 113.12, "metrics": { "wer": 1.125, "cer": 0.813953488372093, "mer": 1.0, "wil": 1.0, "wip": 0.0, "word_hits": 0, "word_substitutions": 8, "word_deletions": 0, "word_insertions": 1, "reference_words": 8, "char_hits": 17, "char_substitutions": 23, "char_deletions": 3, "char_insertions": 9, "reference_chars": 43, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19691845-common_voice_en_19691884", "reference": "There is a growing body of case law on Bayh-Dole.", "transcript": "There is a growing body of Keisla on Beidou.", "duration_s": 4.08, "metrics": { "wer": 0.36363636363636365, "cer": 0.25, "mer": 0.36363636363636365, "wil": 0.505050505050505, "wip": 0.494949494949495, "word_hits": 7, "word_substitutions": 2, "word_deletions": 2, "word_insertions": 0, "reference_words": 11, "char_hits": 37, "char_substitutions": 5, "char_deletions": 6, "char_insertions": 1, "reference_chars": 48, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18466098-common_voice_en_18466096", "reference": "Catch as catch can.", "transcript": "Catch us, Ketchkin.", "duration_s": 1.12, "metrics": { "wer": 0.75, "cer": 0.3333333333333333, "mer": 0.75, "wil": 0.9166666666666666, "wip": 0.08333333333333333, "word_hits": 1, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 4, "char_hits": 12, "char_substitutions": 5, "char_deletions": 1, "char_insertions": 0, "reference_chars": 18, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.08, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 6.88, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 6.56, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "he was educated at queen's college cambridge where he read english with french", "duration_s": 7.44, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } } ] }, { "model": "zonos2", "checkpoint": "Zyphra/ZONOS2", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.01180607887465461, "cer": 0.004210557638516931, "corpus_metrics_match": true, "word_substitutions": 105, "word_deletions": 28, "word_insertions": 8, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 983, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.6849886621315193, 2.7341496598639456, 3.517823129251701, 4.121541950113379, 5.23609977324263, 7.163356009070295 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3546099290780142, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.01486988847583643, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.012102351313969572, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.010436079016026835, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "sacking", "transcript": "sack", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "king", "transcript": "camp", "count": 1 }, { "reference": "s", "transcript": "is", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "we", "transcript": "were", "count": 1 }, { "reference": "lleida", "transcript": "leyda", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 }, { "reference": "fraser", "transcript": "frazier", "count": 1 } ], "common_deletions": [ [ "mart", 1 ], [ "are", 1 ], [ "s", 1 ], [ "is", 1 ], [ "computer", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "ernest prompted", 1 ] ], "common_insertions": [ [ "el", 1 ], [ "maine", 1 ], [ "louie", 1 ], [ "governor", 1 ], [ "rick", 1 ], [ "t", 1 ], [ "cove", 1 ], [ "is", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_28884498-common_voice_en_28884500", "reference": "It is shiny black with reddish legs, short elytra and no hind wings.", "transcript": "It is shiny black with reddish legs.", "duration_s": 2.36843537414966, "metrics": { "wer": 0.46153846153846156, "cer": 0.4696969696969697, "mer": 0.46153846153846156, "wil": 0.46153846153846156, "wip": 0.5384615384615384, "word_hits": 7, "word_substitutions": 0, "word_deletions": 6, "word_insertions": 0, "reference_words": 13, "char_hits": 35, "char_substitutions": 0, "char_deletions": 31, "char_insertions": 0, "reference_chars": 66, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20030172-common_voice_en_20030173", "reference": "The children of Gowanda had school that day.", "transcript": "The children of Gawanda had", "duration_s": 1.4628571428571429, "metrics": { "wer": 0.5, "cer": 0.3953488372093023, "mer": 0.5, "wil": 0.6, "wip": 0.4, "word_hits": 4, "word_substitutions": 1, "word_deletions": 3, "word_insertions": 0, "reference_words": 8, "char_hits": 26, "char_substitutions": 1, "char_deletions": 16, "char_insertions": 0, "reference_chars": 43, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20005340-common_voice_en_20005341", "reference": "Both resolutions failed in votes that split along the Governors' party lines.", "transcript": "Both resolutions failed in votes that split along the governor's party line.", "duration_s": 4.121541950113379, "metrics": { "wer": 0.25, "cer": 0.04, "mer": 0.23076923076923078, "wil": 0.3589743589743589, "wip": 0.6410256410256411, "word_hits": 10, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 74, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 2, "reference_chars": 75, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20806032-common_voice_en_20806033", "reference": "These two agencies co-operate in aspects of law enforcement, including their joint marine unit.", "transcript": "These two agencies cooperate in aspects of law enforcement, including their joint marine units.", "duration_s": 5.23609977324263, "metrics": { "wer": 0.2, "cer": 0.02197802197802198, "mer": 0.2, "wil": 0.3142857142857143, "wip": 0.6857142857142857, "word_hits": 12, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 90, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 1, "reference_chars": 91, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.0549659863945577, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18786931-common_voice_en_18786930", "reference": "Immediately following the demerger, Williams Holdings launched a takeover bid for Racal.", "transcript": "Immediately following the demurger, Williams Holdings launched a takeover bid for Rick Hall.", "duration_s": 4.818140589569161, "metrics": { "wer": 0.25, "cer": 0.06976744186046512, "mer": 0.23076923076923078, "wil": 0.3589743589743589, "wip": 0.6410256410256411, "word_hits": 10, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 84, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 4, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 4.760090702947846, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22972881-common_voice_en_22972884", "reference": "We're too much the ideal bad examples.", "transcript": "were too much the ideal bad examples.", "duration_s": 2.124625850340136, "metrics": { "wer": 0.25, "cer": 0.05263157894736842, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 36, "char_substitutions": 0, "char_deletions": 2, "char_insertions": 0, "reference_chars": 38, "exact_match_rate": 0.0 } } ] }, { "model": "xtts", "checkpoint": "artifacts/xtts2", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.011722347818805995, "cer": 0.003927870023210141, "corpus_metrics_match": true, "word_substitutions": 102, "word_deletions": 16, "word_insertions": 22, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 977, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.7893333333333333, 2.9546666666666663, 3.712, 4.544, 5.749333333333333, 7.701333333333333 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3142857142857143, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.011152416356877323, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.012102351313969572, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.01080879612374208, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "lleida", "transcript": "lida", "count": 1 }, { "reference": "review", "transcript": "reviews", "count": 1 }, { "reference": "don", "transcript": "donlash", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "ramblers", "transcript": "rambler", "count": 1 }, { "reference": "coral", "transcript": "choral", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 }, { "reference": "piece", "transcript": "pea", "count": 1 }, { "reference": "queens", "transcript": "is", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "to", 1 ], [ "mart", 1 ], [ "lash", 1 ], [ "computer", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "that", 1 ] ], "common_insertions": [ [ "they are around", 1 ], [ "tertian", 1 ], [ "queen", 1 ], [ "damaschino", 1 ], [ "maine", 1 ], [ "raised to", 1 ], [ "d t", 1 ], [ "governor", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_34980363-common_voice_en_34980360", "reference": "The township includes the unincorporated community of Middle Lancaster.", "transcript": "The township includes the unincorporated community of Middle Lancaster. They're around.", "duration_s": 5.664, "metrics": { "wer": 0.3333333333333333, "cer": 0.22857142857142856, "mer": 0.25, "wil": 0.25, "wip": 0.75, "word_hits": 9, "word_substitutions": 0, "word_deletions": 0, "word_insertions": 3, "reference_words": 9, "char_hits": 70, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 16, "reference_chars": 70, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20907200-common_voice_en_20907214", "reference": "Assume that a star radiates a continuous light spectrum up to X rays.", "transcript": "Assume that a star radiates a continuous light spectrum up to x raised to e.", "duration_s": 5.386666666666667, "metrics": { "wer": 0.23076923076923078, "cer": 0.11764705882352941, "mer": 0.2, "wil": 0.2615384615384615, "wip": 0.7384615384615385, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 2, "reference_words": 13, "char_hits": 67, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 7, "reference_chars": 68, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 1.9413333333333334, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19527999-common_voice_en_19528002", "reference": "Eventually the nomenclature was standardized to the conventions used today.", "transcript": "Eventually, the nomenclature was standardized to the convention's use today.", "duration_s": 4.64, "metrics": { "wer": 0.3, "cer": 0.04054054054054054, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 10, "char_hits": 73, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 2, "reference_chars": 74, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "he relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 5.333333333333333, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_27794266-common_voice_en_27794264", "reference": "Don Lash Park in Auburn is named for him.", "transcript": "Donlash Park in Auburn is named for him.", "duration_s": 2.6453333333333333, "metrics": { "wer": 0.2222222222222222, "cer": 0.025, "mer": 0.2222222222222222, "wil": 0.3194444444444444, "wip": 0.6805555555555556, "word_hits": 7, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 9, "char_hits": 39, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 40, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge where he read English with French.", "duration_s": 4.544, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_31959919-common_voice_en_31959927", "reference": "One of the most famous Greek painters during that time was Michael Damaskinos.", "transcript": "One of the most famous Greek painters during that time was Michael Damaschino Stott.", "duration_s": 4.362666666666667, "metrics": { "wer": 0.15384615384615385, "cer": 0.09090909090909091, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 76, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 6, "reference_chars": 77, "exact_match_rate": 0.0 } } ] }, { "model": "fishtts", "checkpoint": "fishaudio/s2-pro", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.011471154651260152, "cer": 0.0031690769505445457, "corpus_metrics_match": true, "word_substitutions": 114, "word_deletions": 15, "word_insertions": 8, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 976, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.9287981859410431, 3.250793650793651, 3.9938321995464854, 4.702040816326531, 5.851428571428571, 8.870022675736962 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.29927007299270075, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.0055762081784386614, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.010603964960811434, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.015654118524040254, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "franzen", "transcript": "franson", "count": 1 }, { "reference": "tutting", "transcript": "tooting", "count": 1 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "lleida", "transcript": "lida", "count": 1 }, { "reference": "middleton", "transcript": "milton", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "review", "transcript": "reviews", "count": 1 }, { "reference": "iro", "transcript": "ero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "mart", 1 ], [ "run", 1 ], [ "computer", 1 ], [ "gates", 1 ], [ "operate", 1 ], [ "shot", 1 ], [ "in", 1 ] ], "common_insertions": [ [ "queen", 1 ], [ "maine", 1 ], [ "luis", 1 ], [ "governor", 1 ], [ "netar", 1 ], [ "sean", 1 ], [ "t", 1 ], [ "is", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.786394557823129, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 4.73687074829932, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 3.7616326530612243, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_650392-common_voice_en_650389", "reference": "Soldier sitting on top of Jeep holding a machine gun and soldier sitting inside.", "transcript": "Soldiers sitting on top of jeep holding a machine gun, and soldiers sitting inside.", "duration_s": 4.876190476190477, "metrics": { "wer": 0.14285714285714285, "cer": 0.02531645569620253, "mer": 0.14285714285714285, "wil": 0.26530612244897966, "wip": 0.7346938775510203, "word_hits": 12, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 14, "char_hits": 79, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 79, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19622969-common_voice_en_19622971", "reference": "A skunk's spray is powerful enough to ward off bears and other potential attackers.", "transcript": "The skunk spray is powerful enough to ward off bears and other potential attackers.", "duration_s": 5.015510204081632, "metrics": { "wer": 0.13333333333333333, "cer": 0.07228915662650602, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 79, "char_substitutions": 1, "char_deletions": 3, "char_insertions": 2, "reference_chars": 83, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge, where he read English with French.", "duration_s": 4.6439909297052155, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "its most prominent representatives gathered in the Maine's Fine Arts Association.", "duration_s": 4.690430839002268, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 78, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_554379-common_voice_en_554380", "reference": "I got drunk yesterday and thought I built a super computer.", "transcript": "I got drunk yesterday and thought I built a supercomputer.", "duration_s": 3.250793650793651, "metrics": { "wer": 0.18181818181818182, "cer": 0.017241379310344827, "mer": 0.18181818181818182, "wil": 0.26363636363636356, "wip": 0.7363636363636364, "word_hits": 9, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 57, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 58, "exact_match_rate": 0.0 } } ] }, { "model": "echo", "checkpoint": "jordand/echo-tts-base", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.01121996148371431, "cer": 0.0034815211569362613, "corpus_metrics_match": true, "word_substitutions": 110, "word_deletions": 11, "word_insertions": 13, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 983, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.0216780045351475, 3.5294331065759637, 4.08671201814059, 5.2477097505668935, 6.870784580498847, 15.696689342403628 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.35074626865671643, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.020446096654275093, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.01094974642692485, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.010436079016026835, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "wish", "transcript": "which", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "it", "transcript": "a", "count": 1 }, { "reference": "lleida", "transcript": "lida", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "norwegia", "transcript": "norway", "count": 1 }, { "reference": "roads", "transcript": "road", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "criticism", "transcript": "criticisms", "count": 1 } ], "common_deletions": [ [ "mart", 1 ], [ "run", 1 ], [ "is", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "shot", 1 ], [ "jumper", 1 ], [ "strict", 1 ] ], "common_insertions": [ [ "a", 2 ], [ "is", 2 ], [ "but", 1 ], [ ".", 1 ], [ "main", 1 ], [ "the", 1 ], [ "ba", 1 ], [ "governor", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_564578-common_voice_en_564576", "reference": "Strike while the iron is hot Make hay while the sun shines", "transcript": "Fight while the night is dark How I fear the sun's shine", "duration_s": 15.696689342403628, "metrics": { "wer": 0.6666666666666666, "cer": 0.5, "mer": 0.6153846153846154, "wil": 0.8397435897435898, "wip": 0.16025641025641027, "word_hits": 5, "word_substitutions": 7, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 34, "char_substitutions": 18, "char_deletions": 6, "char_insertions": 5, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.461315192743764, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 5.479909297052155, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 3.6687528344671203, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_650392-common_voice_en_650389", "reference": "Soldier sitting on top of Jeep holding a machine gun and soldier sitting inside.", "transcript": "Soldiers sitting on top of Jeep holding a machine gun and soldiers sitting inside.", "duration_s": 5.2477097505668935, "metrics": { "wer": 0.14285714285714285, "cer": 0.02531645569620253, "mer": 0.14285714285714285, "wil": 0.26530612244897966, "wip": 0.7346938775510203, "word_hits": 12, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 14, "char_hits": 79, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 79, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_24985801-common_voice_en_24985802", "reference": "Errors from the model and adding accumulate, which eventually subtracted from the initial value.", "transcript": "errors from the model and adding a cumulate, which eventually subtracted from the initial value.", "duration_s": 5.294149659863946, "metrics": { "wer": 0.14285714285714285, "cer": 0.010638297872340425, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 14, "char_hits": 93, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 0, "reference_chars": 94, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_17854896-common_voice_en_17854894", "reference": "To wish one well To wish one ill.", "transcript": "To which one well, to which one ill", "duration_s": 7.151746031746032, "metrics": { "wer": 0.25, "cer": 0.125, "mer": 0.25, "wil": 0.4375, "wip": 0.5625, "word_hits": 6, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 8, "char_hits": 30, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 2, "reference_chars": 32, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "its most prominent representatives gathered in the main's fine arts association", "duration_s": 5.061950113378685, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 77, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 78, "exact_match_rate": 0.0 } } ] }, { "model": "styletts2", "checkpoint": "artifacts/styletts2/epochs_2nd_00020.pth", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.010885037260319853, "cer": 0.0029905374040349935, "corpus_metrics_match": true, "word_substitutions": 108, "word_deletions": 13, "word_insertions": 9, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 984, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.5229166666666667, 3.872916666666667, 4.635416666666666, 5.422916666666667, 6.447916666666667, 8.797916666666667 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3230769230769231, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.007434944237918215, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.011180267404333795, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.01080879612374208, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "wal mart", "transcript": "while marked", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "lleida", "transcript": "leyda", "count": 1 }, { "reference": "soyuz", "transcript": "a", "count": 1 }, { "reference": "unlike", "transcript": "like", "count": 1 }, { "reference": "norwegia", "transcript": "norvegia", "count": 1 }, { "reference": "roads", "transcript": "road", "count": 1 }, { "reference": "don", "transcript": "dunlash", "count": 1 }, { "reference": "iro", "transcript": "eero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 } ], "common_deletions": [ [ "re", 1 ], [ "lash", 1 ], [ "is", 1 ], [ "computer", 1 ], [ "gates", 1 ], [ "operate", 1 ], [ "jumper", 1 ], [ "web", 1 ] ], "common_insertions": [ [ "a", 1 ], [ "soy", 1 ], [ "and", 1 ], [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "limit", 1 ], [ "netor", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for a while, marked as a fitness trainer.", "duration_s": 6.347916666666666, "metrics": { "wer": 0.2, "cer": 0.09302325581395349, "mer": 0.1875, "wil": 0.2958333333333333, "wip": 0.7041666666666667, "word_hits": 13, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 15, "char_hits": 84, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 6, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.8229166666666665, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20095636-common_voice_en_20095638", "reference": "The automatic control system would be redesigned on future Soyuz spacecraft.", "transcript": "The automatic control system would be redesigned on future Soy, a spacecraft.", "duration_s": 5.872916666666667, "metrics": { "wer": 0.18181818181818182, "cer": 0.02666666666666667, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 73, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 0, "reference_chars": 75, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23650405-common_voice_en_23650404", "reference": "Unlike the sedans, the convertibles were assembled in Mexico.", "transcript": "And like the sedans, the convertibles were assembled in Mexico.", "duration_s": 4.697916666666667, "metrics": { "wer": 0.2222222222222222, "cer": 0.05084745762711865, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 9, "char_hits": 58, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 59, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_27794266-common_voice_en_27794264", "reference": "Don Lash Park in Auburn is named for him.", "transcript": "Dunlash Park in Auburn is named for him.", "duration_s": 3.4229166666666666, "metrics": { "wer": 0.2222222222222222, "cer": 0.05, "mer": 0.2222222222222222, "wil": 0.3194444444444444, "wip": 0.6805555555555556, "word_hits": 7, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 9, "char_hits": 38, "char_substitutions": 1, "char_deletions": 1, "char_insertions": 0, "reference_chars": 40, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_650392-common_voice_en_650389", "reference": "Soldier sitting on top of Jeep holding a machine gun and soldier sitting inside.", "transcript": "Soldiers sitting on top of jeep holding a machine gun and soldiers sitting inside.", "duration_s": 5.822916666666667, "metrics": { "wer": 0.14285714285714285, "cer": 0.02531645569620253, "mer": 0.14285714285714285, "wil": 0.26530612244897966, "wip": 0.7346938775510203, "word_hits": 12, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 14, "char_hits": 79, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 79, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge, where he read English with French.", "duration_s": 5.347916666666666, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "Its most prominent representatives gathered in the Maine's Fine Arts Association.", "duration_s": 5.397916666666666, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 78, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } } ] }, { "model": "f5tts", "checkpoint": "F5TTS_v1_Base", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.010550113036925395, "cer": 0.0031988335416294708, "corpus_metrics_match": true, "word_substitutions": 100, "word_deletions": 19, "word_insertions": 7, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 989, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.64, 2.752, 3.466666666666667, 4.117333333333334, 4.981333333333334, 7.370666666666667 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3492063492063492, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.020446096654275093, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.009451360073766712, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.012299664554603056, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "heather", "transcript": "hither", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "runny", "count": 1 }, { "reference": "lleida", "transcript": "leida", "count": 1 }, { "reference": "cribb", "transcript": "krebb", "count": 1 }, { "reference": "as", "transcript": "is", "count": 1 }, { "reference": "can", "transcript": "camp", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 } ], "common_deletions": [ [ "mart", 1 ], [ "run", 1 ], [ "is", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "for mature", 1 ], [ "it", 1 ], [ "operate", 1 ] ], "common_insertions": [ [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "esther", 1 ], [ "we", 1 ], [ "angela", 1 ], [ "t", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_597578-common_voice_en_597583", "reference": "Wilma\u2019s gone cat sitting again for the weekend if you want to come over.", "transcript": "We'll must gone cat sitting again for the weekend if you want to come over.", "duration_s": 4.010666666666666, "metrics": { "wer": 0.2, "cer": 0.11267605633802817, "mer": 0.1875, "wil": 0.2958333333333333, "wip": 0.7041666666666667, "word_hits": 13, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 15, "char_hits": 68, "char_substitutions": 3, "char_deletions": 0, "char_insertions": 5, "reference_chars": 71, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.208, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19992384-common_voice_en_19992617", "reference": "He then moved to Belgian side Lierse, but has since retired.", "transcript": "He then moved to Belgianside Leers, but has since retired.", "duration_s": 3.2426666666666666, "metrics": { "wer": 0.2727272727272727, "cer": 0.05172413793103448, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 55, "char_substitutions": 1, "char_deletions": 2, "char_insertions": 0, "reference_chars": 58, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20531306-common_voice_en_20531308", "reference": "His alleged code name in Soviet intelligence and in the Venona files is \"Ted\".", "transcript": "His alleged codename in Soviet intelligence and in the Venana files is, Ted.", "duration_s": 4.224, "metrics": { "wer": 0.21428571428571427, "cer": 0.02666666666666667, "mer": 0.21428571428571427, "wil": 0.3351648351648352, "wip": 0.6648351648351648, "word_hits": 11, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 14, "char_hits": 73, "char_substitutions": 1, "char_deletions": 1, "char_insertions": 0, "reference_chars": 75, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 4.8213333333333335, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18466098-common_voice_en_18466096", "reference": "Catch as catch can.", "transcript": "Catch's Catch Camp", "duration_s": 1.0133333333333334, "metrics": { "wer": 0.5, "cer": 0.16666666666666666, "mer": 0.5, "wil": 0.75, "wip": 0.25, "word_hits": 2, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 4, "char_hits": 16, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 1, "reference_chars": 18, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 3.082666666666667, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge where he read English with French.", "duration_s": 4.330666666666667, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } } ] }, { "model": "speecht5", "checkpoint": "microsoft/speecht5_tts", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.010131457757682325, "cer": 0.0029459025174076057, "corpus_metrics_match": true, "word_substitutions": 102, "word_deletions": 10, "word_insertions": 9, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 986, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 0.832, 2.944, 3.712, 4.576, 5.652799999999996, 7.68 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.32231404958677684, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.011152416356877323, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.009681881051175657, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.011554230339172567, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "wal", "transcript": "wall", "count": 1 }, { "reference": "s pheromone", "transcript": "as fairmont", "count": 1 }, { "reference": "roney", "transcript": "runny", "count": 1 }, { "reference": "personnel", "transcript": "personal", "count": 1 }, { "reference": "we", "transcript": "were", "count": 1 }, { "reference": "lleida", "transcript": "lieta", "count": 1 }, { "reference": "soccer", "transcript": "soxer", "count": 1 }, { "reference": "s", "transcript": "as", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "knots", "transcript": "nuts", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 } ], "common_deletions": [ [ "are", 1 ], [ "is", 1 ], [ "made", 1 ], [ "gates", 1 ], [ "a", 1 ], [ "operate", 1 ], [ "and", 1 ], [ "won", 1 ] ], "common_insertions": [ [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "it", 1 ], [ "t", 1 ], [ "does", 1 ], [ "for", 1 ], [ "der", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you are negative.", "duration_s": 2.432, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_123125-common_voice_en_123126", "reference": "After all, who doesn\u2019t want to overcome new challenges and achieve great heights?", "transcript": "after all who doesn't, to want to overcome new challenges and achieve great heights.", "duration_s": 4.672, "metrics": { "wer": 0.21428571428571427, "cer": 0.05063291139240506, "mer": 0.2, "wil": 0.3142857142857143, "wip": 0.6857142857142857, "word_hits": 12, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 14, "char_hits": 79, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 4, "reference_chars": 79, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_30471104-common_voice_en_30471105", "reference": "The ant colonies move around following the queen\u2019s pheromone trail.", "transcript": "The ant colonies move around following the queen, as Fairmont Trail.", "duration_s": 3.52, "metrics": { "wer": 0.18181818181818182, "cer": 0.09090909090909091, "mer": 0.18181818181818182, "wil": 0.3305785123966941, "wip": 0.6694214876033059, "word_hits": 9, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 11, "char_hits": 61, "char_substitutions": 4, "char_deletions": 1, "char_insertions": 1, "reference_chars": 66, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22972881-common_voice_en_22972884", "reference": "We're too much the ideal bad examples.", "transcript": "were too much the ideal bad examples.", "duration_s": 2.208, "metrics": { "wer": 0.25, "cer": 0.05263157894736842, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 36, "char_substitutions": 0, "char_deletions": 2, "char_insertions": 0, "reference_chars": 38, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge, where he read English with French.", "duration_s": 4.608, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "its most prominent representatives gathered in the Maine's Fine Arts Association.", "duration_s": 5.088, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 78, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20065698-common_voice_en_20065701", "reference": "Flood gates were erected and drainage systems improved.", "transcript": "Floodgates were erected and drainage systems improved.", "duration_s": 2.912, "metrics": { "wer": 0.25, "cer": 0.018518518518518517, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20005340-common_voice_en_20005341", "reference": "Both resolutions failed in votes that split along the Governors' party lines.", "transcript": "Both resolutions failed in votes that split along the governor's party lines.", "duration_s": 4.576, "metrics": { "wer": 0.16666666666666666, "cer": 0.02666666666666667, "mer": 0.15384615384615385, "wil": 0.22435897435897445, "wip": 0.7756410256410255, "word_hits": 11, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 75, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 75, "exact_match_rate": 0.0 } } ] }, { "model": "omnivoice", "checkpoint": "k2-fsa/OmniVoice", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.009880264590136481, "cer": 0.002648336606558353, "corpus_metrics_match": true, "word_substitutions": 90, "word_deletions": 21, "word_insertions": 7, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 996, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.6, 3.32, 4.08, 4.84, 5.88, 8.24 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3813559322033898, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.011152416356877323, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.009797141539880129, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.01006336190831159, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "we", "transcript": "were", "count": 1 }, { "reference": "lleida", "transcript": "leyda", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "ramblers", "transcript": "rambler", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 }, { "reference": "and", "transcript": "in", "count": 1 }, { "reference": "queens", "transcript": "is", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "mart", 1 ], [ "are", 1 ], [ "run", 1 ], [ "computer", 1 ], [ "gates", 1 ], [ "or", 1 ], [ "operate", 1 ] ], "common_insertions": [ [ "queen", 1 ], [ "maine", 1 ], [ "louie", 1 ], [ "governor", 1 ], [ "t", 1 ], [ "hindu", 1 ], [ "is", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_20030172-common_voice_en_20030174", "reference": "Various ecclesiastical areas, too small for a diocese, are termed prefects.", "transcript": "Various ecclesiastical areas , are termed prefects", "duration_s": 4.84, "metrics": { "wer": 0.45454545454545453, "cer": 0.3333333333333333, "mer": 0.45454545454545453, "wil": 0.4545454545454546, "wip": 0.5454545454545454, "word_hits": 6, "word_substitutions": 0, "word_deletions": 5, "word_insertions": 0, "reference_words": 11, "char_hits": 48, "char_substitutions": 0, "char_deletions": 24, "char_insertions": 0, "reference_chars": 72, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.44, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_586741-common_voice_en_586737", "reference": "Please search The Best of Steely Dan: Then and Now saga.", "transcript": "Please search The Best of Steely Dan, then in NowSaga.", "duration_s": 3.52, "metrics": { "wer": 0.2727272727272727, "cer": 0.05555555555555555, "mer": 0.2727272727272727, "wil": 0.4181818181818181, "wip": 0.5818181818181819, "word_hits": 8, "word_substitutions": 2, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 51, "char_substitutions": 1, "char_deletions": 2, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 5.6, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_22972881-common_voice_en_22972884", "reference": "We're too much the ideal bad examples.", "transcript": "were too much the ideal bad examples.", "duration_s": 2.48, "metrics": { "wer": 0.25, "cer": 0.05263157894736842, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 36, "char_substitutions": 0, "char_deletions": 2, "char_insertions": 0, "reference_chars": 38, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 3.68, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge, where he read English with French.", "duration_s": 5.08, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "its most prominent representatives gathered in the Maine's Fine Arts Association.", "duration_s": 5.2, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 78, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } } ] }, { "model": "supertonic", "checkpoint": "Supertone/supertonic-3", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.009796533534287867, "cer": 0.002663214902100815, "corpus_metrics_match": true, "word_substitutions": 97, "word_deletions": 11, "word_insertions": 9, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 990, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.3566213151927438, 3.5367176870748303, 4.371972789115647, 5.19936507936508, 6.275939909297052, 8.340544217687075 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3333333333333333, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.011152416356877323, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.008529276164130935, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.013417815877748788, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.021739130434782608, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "soldier", "transcript": "soldiers", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "beard", "transcript": "bird", "count": 1 }, { "reference": "roney", "transcript": "runny", "count": 1 }, { "reference": "lleida", "transcript": "leida", "count": 1 }, { "reference": "as", "transcript": "is", "count": 1 }, { "reference": "can", "transcript": "cam", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "roads", "transcript": "road", "count": 1 }, { "reference": "iro", "transcript": "iroh", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "mart", 1 ], [ "run", 1 ], [ "computer", 1 ], [ "the", 1 ], [ "operate", 1 ], [ "and", 1 ], [ "name", 1 ] ], "common_insertions": [ [ "an", 1 ], [ "queen", 1 ], [ "maine", 1 ], [ "governor", 1 ], [ "angela", 1 ], [ "them", 1 ], [ "t", 1 ], [ "ula", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.3352607709750566, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37457199-common_voice_en_37457214", "reference": "Stevenson is also the director of music ministries at Angeles Mesa Presbyterian Church.", "transcript": "Stevenson is also the director of music ministries at Angela's Mezza Presbyterian Church.", "duration_s": 6.045396825396826, "metrics": { "wer": 0.23076923076923078, "cer": 0.05813953488372093, "mer": 0.21428571428571427, "wil": 0.3351648351648352, "wip": 0.6648351648351648, "word_hits": 11, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 84, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 3, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 6.204829931972789, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_18466098-common_voice_en_18466096", "reference": "Catch as catch can.", "transcript": "Catch's Catch Cam", "duration_s": 1.7158956916099772, "metrics": { "wer": 0.5, "cer": 0.1111111111111111, "mer": 0.5, "wil": 0.75, "wip": 0.25, "word_hits": 2, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 4, "char_hits": 16, "char_substitutions": 2, "char_deletions": 0, "char_insertions": 0, "reference_chars": 18, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 4.094331065759637, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_650392-common_voice_en_650389", "reference": "Soldier sitting on top of Jeep holding a machine gun and soldier sitting inside.", "transcript": "Soldiers sitting on top of jeep holding a machine gun and soldiers sitting inside.", "duration_s": 5.370022675736961, "metrics": { "wer": 0.14285714285714285, "cer": 0.02531645569620253, "mer": 0.14285714285714285, "wil": 0.26530612244897966, "wip": 0.7346938775510203, "word_hits": 12, "word_substitutions": 2, "word_deletions": 0, "word_insertions": 0, "reference_words": 14, "char_hits": 79, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 79, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_23962172-common_voice_en_23962173", "reference": "He was educated at Queens' College, Cambridge where he read English with French.", "transcript": "He was educated at Queen's College, Cambridge where he read English with French.", "duration_s": 5.0377551020408164, "metrics": { "wer": 0.15384615384615385, "cer": 0.025974025974025976, "mer": 0.14285714285714285, "wil": 0.20879120879120883, "wip": 0.7912087912087912, "word_hits": 12, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 13, "char_hits": 77, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 77, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "Its most prominent representatives gathered in the Maine's Fine Arts Association.", "duration_s": 5.458231292517007, "metrics": { "wer": 0.18181818181818182, "cer": 0.038461538461538464, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 78, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } } ] }, { "model": "kokoro", "checkpoint": "hexgrad/Kokoro-82M", "samples": 1088, "unique_ids_and_reference_texts_match": true, "all_per_sample_metrics_match": true, "wer": 0.009629071422590639, "cer": 0.0024400404689638756, "corpus_metrics_match": true, "word_substitutions": 94, "word_deletions": 15, "word_insertions": 6, "reference_words": 11943, "empty_transcripts": 0, "wer_over_100_percent": 0, "exact_transcripts": 994, "duplicate_audio_hashes": 0, "duration_percentiles_s": [ 1.525, 3.4, 4.225, 5.025, 6.091249999999996, 8.3 ], "high_clipping_recordings": 0, "quiet_recordings_below_minus_50_dbfs": 0, "top20_fraction_of_word_errors": 0.3565217391304348, "length_buckets": [ { "min_chars": 0, "max_chars_exclusive": 40, "samples": 89, "wer": 0.00929368029739777, "empty_transcripts": 0 }, { "min_chars": 40, "max_chars_exclusive": 80, "samples": 805, "wer": 0.009797141539880129, "empty_transcripts": 0 }, { "min_chars": 80, "max_chars_exclusive": 120, "samples": 192, "wer": 0.009317927692881103, "empty_transcripts": 0 }, { "min_chars": 120, "max_chars_exclusive": 10000, "samples": 2, "wer": 0.0, "empty_transcripts": 0 } ], "common_substitutions": [ { "reference": "s", "transcript": "is", "count": 2 }, { "reference": "wal", "transcript": "walmart", "count": 1 }, { "reference": "roney", "transcript": "roni", "count": 1 }, { "reference": "lleida", "transcript": "leyda", "count": 1 }, { "reference": "as", "transcript": "is", "count": 1 }, { "reference": "re", "transcript": "rerun", "count": 1 }, { "reference": "roads", "transcript": "road", "count": 1 }, { "reference": "iro", "transcript": "eero", "count": 1 }, { "reference": "ogilvie", "transcript": "ogilvy", "count": 1 }, { "reference": "tulu", "transcript": "chilu", "count": 1 }, { "reference": "dickson", "transcript": "dixon", "count": 1 }, { "reference": "ilopango", "transcript": "ilapango", "count": 1 } ], "common_deletions": [ [ "is", 2 ], [ "mart", 1 ], [ "run", 1 ], [ "dole", 1 ], [ "gates", 1 ], [ "operate", 1 ], [ "are", 1 ], [ "shot", 1 ] ], "common_insertions": [ [ "monet", 1 ], [ "governor", 1 ], [ "nedar", 1 ], [ "t", 1 ], [ "is", 1 ], [ "lowe", 1 ] ], "worst_examples": [ { "id": "seed_common_voice_en_19778240-common_voice_en_19778244", "reference": "She married David Wallett, and moved to Los Angeles to try acting.", "transcript": "She married David Wallet and moved to Lowe's Angels to try acting.", "duration_s": 4.625, "metrics": { "wer": 0.3333333333333333, "cer": 0.09375, "mer": 0.3076923076923077, "wil": 0.48076923076923084, "wip": 0.5192307692307692, "word_hits": 9, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 62, "char_substitutions": 0, "char_deletions": 2, "char_insertions": 4, "reference_chars": 64, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19284142-common_voice_en_19284143", "reference": "You aren\u2019t positive, you\u2019re negative.", "transcript": "You aren't positive, you're negative.", "duration_s": 2.325, "metrics": { "wer": 0.42857142857142855, "cer": 0.08571428571428572, "mer": 0.42857142857142855, "wil": 0.6734693877551021, "wip": 0.32653061224489793, "word_hits": 4, "word_substitutions": 3, "word_deletions": 0, "word_insertions": 0, "reference_words": 7, "char_hits": 34, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 2, "reference_chars": 35, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_25347734-common_voice_en_25347736", "reference": "He relocated to Bentonville, Arkansas, where he worked for Wal-Mart as a fitness trainer.", "transcript": "He relocated to Bentonville, Arkansas, where he worked for Walmart as a fitness trainer.", "duration_s": 5.75, "metrics": { "wer": 0.13333333333333333, "cer": 0.011627906976744186, "mer": 0.13333333333333333, "wil": 0.1952380952380952, "wip": 0.8047619047619048, "word_hits": 13, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 15, "char_hits": 85, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 86, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_37368335-common_voice_en_37368333", "reference": "The first season continues to re-run in many territories.", "transcript": "The first season continues to rerun in many territories.", "duration_s": 4.05, "metrics": { "wer": 0.2, "cer": 0.018518518518518517, "mer": 0.2, "wil": 0.28888888888888886, "wip": 0.7111111111111111, "word_hits": 8, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 10, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_33966405-common_voice_en_33966447", "reference": "Its most prominent representatives gathered in the Manes Fine Arts Association.", "transcript": "Its most prominent representatives gathered in the Monet's Fine Arts Association.", "duration_s": 5.5, "metrics": { "wer": 0.18181818181818182, "cer": 0.05128205128205128, "mer": 0.16666666666666666, "wil": 0.24242424242424243, "wip": 0.7575757575757576, "word_hits": 10, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 11, "char_hits": 77, "char_substitutions": 1, "char_deletions": 0, "char_insertions": 3, "reference_chars": 78, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_19691845-common_voice_en_19691884", "reference": "There is a growing body of case law on Bayh-Dole.", "transcript": "There's a growing body of case law on Baydol.", "duration_s": 3.1, "metrics": { "wer": 0.18181818181818182, "cer": 0.0625, "mer": 0.18181818181818182, "wil": 0.26363636363636356, "wip": 0.7363636363636364, "word_hits": 9, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 11, "char_hits": 45, "char_substitutions": 0, "char_deletions": 3, "char_insertions": 0, "reference_chars": 48, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20065698-common_voice_en_20065701", "reference": "Flood gates were erected and drainage systems improved.", "transcript": "Floodgates were erected and drainage systems improved.", "duration_s": 3.8, "metrics": { "wer": 0.25, "cer": 0.018518518518518517, "mer": 0.25, "wil": 0.3571428571428572, "wip": 0.6428571428571428, "word_hits": 6, "word_substitutions": 1, "word_deletions": 1, "word_insertions": 0, "reference_words": 8, "char_hits": 53, "char_substitutions": 0, "char_deletions": 1, "char_insertions": 0, "reference_chars": 54, "exact_match_rate": 0.0 } }, { "id": "seed_common_voice_en_20005340-common_voice_en_20005341", "reference": "Both resolutions failed in votes that split along the Governors' party lines.", "transcript": "Both resolutions failed in votes that split along the governor's party lines.", "duration_s": 4.85, "metrics": { "wer": 0.16666666666666666, "cer": 0.02666666666666667, "mer": 0.15384615384615385, "wil": 0.22435897435897445, "wip": 0.7756410256410255, "word_hits": 11, "word_substitutions": 1, "word_deletions": 0, "word_insertions": 1, "reference_words": 12, "char_hits": 75, "char_substitutions": 0, "char_deletions": 0, "char_insertions": 2, "reference_chars": 75, "exact_match_rate": 0.0 } } ] } ] }