Upload folder using huggingface_hub

results_organized/bleu/st.csv

@@ -1,8 +1,9 @@
 Model,covost2_en_id_test,covost2_en_zh_test,covost2_en_ta_test,covost2_id_en_test,covost2_zh_en_test,covost2_ta_en_test
-whisper_large_v3,,,,,,
+whisper_large_v3,1.600581653970121,0.16408986541757878,0.02107778621423822,46.01512198258627,14.673689493155793,2.451098639578599
 gemini-1.5-flash,,,,,,
-Qwen-Audio-Chat,4.102230932924371,15.330641138043728,,0.45648619714728844,9.898238298955656,0.01699144301093184
+Qwen-Audio-Chat,4.102230932924371,15.330641138043728,0.03451483807236294,0.45648619714728844,9.898238298955656,0.01699144301093184
 SALMONN_7B,14.102682915273142,33.88941292215531,0.00046745670226766583,26.89649039333571,5.296039450108202,0.3649023706010388
+old_models,,,,,,
 Qwen2-Audio-7B-Instruct,16.325186897428104,25.765420247070075,0.03245972071872916,6.326113431899141,16.466557744958333,0.04425838146050298
 WavLLM_fairseq,13.841886973016162,31.96381187282953,0.0033159224040994286,5.933522277713613,2.368659001743569,0.1695522548322915
 MERaLiON-AudioLLM-Whisper-SEA-LION,37.60224687716629,43.941098854450516,14.407399367512914,44.43289180618449,18.76473995941838,5.023057608950299

results_organized/llama3_70b_judge/accent_recognition.csv

@@ -1,10 +1,11 @@
 Model,voxceleb_accent_test,imda_ar_sentence,imda_ar_dialogue
 whisper_large_v3,,,
 gemini-1.5-flash,,,
-Qwen-Audio-Chat,48.05088223225277,,
-SALMONN_7B,34.222404595814524,,
-Qwen2-Audio-7B-Instruct,29.187525646286417,,
+Qwen-Audio-Chat,48.05088223225277,3.933333333333333,0.6666666666666667
+SALMONN_7B,34.222404595814524,2.5166666666666666,0.06666666666666667
+old_models,,,
+Qwen2-Audio-7B-Instruct,29.187525646286417,2.55,0.9666666666666667
 WavLLM_fairseq,,,
-MERaLiON-AudioLLM-Whisper-SEA-LION,47.01682396389003,7.966666666666667,77.83333333333333
+MERaLiON-AudioLLM-Whisper-SEA-LION,47.01682396389003,7.816666666666666,77.83333333333333
 cascade_whisper_large_v3_llama_3_8b_instruct,39.32704144439885,,
 cascade_whisper_large_v2_gemma2_9b_cpt_sea_lionv3_instruct,24.640951990151827,,

results_organized/llama3_70b_judge/audio_captioning.csv

@@ -3,6 +3,7 @@ whisper_large_v3,,
 gemini-1.5-flash,,
 Qwen-Audio-Chat,47.04090909090909,32.9364161849711
 SALMONN_7B,37.445454545454545,23.76878612716763
+old_models,,
 Qwen2-Audio-7B-Instruct,40.77727272727273,33.78034682080925
 WavLLM_fairseq,5.5,6.901734104046243
 MERaLiON-AudioLLM-Whisper-SEA-LION,38.00454545454545,33.97687861271676

results_organized/llama3_70b_judge/audio_scene_question_answering.csv

@@ -3,6 +3,7 @@ whisper_large_v3,,,
 gemini-1.5-flash,,,
 Qwen-Audio-Chat,61.934856587263,50.22364217252396,42.69736842105263
 SALMONN_7B,57.75401069518716,50.287539936102235,47.30263157894737
+old_models,,,
 Qwen2-Audio-7B-Instruct,50.919591292758774,45.75079872204473,44.473684210526315
 WavLLM_fairseq,43.01199466903598,29.840255591054312,26.25
 MERaLiON-AudioLLM-Whisper-SEA-LION,63.15021876519203,49.77635782747604,46.31578947368421

results_organized/llama3_70b_judge/emotion_recognition.csv

@@ -1,9 +1,10 @@
 Model,iemocap_emotion_test,meld_sentiment_test,meld_emotion_test
 whisper_large_v3,,,
 gemini-1.5-flash,,,
-Qwen-Audio-Chat,,,
-SALMONN_7B,,,
-Qwen2-Audio-7B-Instruct,,,
+Qwen-Audio-Chat,29.382470119521916,44.90421455938697,50.72796934865901
+SALMONN_7B,23.804780876494025,41.7624521072797,30.536398467432953
+old_models,,,
+Qwen2-Audio-7B-Instruct,53.98406374501992,53.9463601532567,41.60919540229885
 WavLLM_fairseq,,,
 MERaLiON-AudioLLM-Whisper-SEA-LION,48.505976095617534,46.206896551724135,36.36015325670498
 cascade_whisper_large_v3_llama_3_8b_instruct,,,

results_organized/llama3_70b_judge/gender_recognition.csv

@@ -1,10 +1,11 @@
 Model,voxceleb_gender_test,iemocap_gender_test,imda_gr_sentence,imda_gr_dialogue
 whisper_large_v3,,,,
 gemini-1.5-flash,,,,
-Qwen-Audio-Chat,,,,
-SALMONN_7B,,,,
-Qwen2-Audio-7B-Instruct,,,,
+Qwen-Audio-Chat,70.5990972507181,50.0996015936255,57.550000000000004,37.2
+SALMONN_7B,88.79770209273697,81.31474103585658,59.766666666666666,42.733333333333334
+old_models,,,,
+Qwen2-Audio-7B-Instruct,99.1177677472302,92.80876494023903,68.38333333333333,61.56666666666667
 WavLLM_fairseq,,,,
-MERaLiON-AudioLLM-Whisper-SEA-LION,99.75379565038982,93.48605577689243,66.36666666666666,93.76666666666667
+MERaLiON-AudioLLM-Whisper-SEA-LION,99.75379565038982,93.48605577689243,66.13333333333333,93.76666666666667
 cascade_whisper_large_v3_llama_3_8b_instruct,,,,
-cascade_whisper_large_v2_gemma2_9b_cpt_sea_lionv3_instruct,,,,
+cascade_whisper_large_v2_gemma2_9b_cpt_sea_lionv3_instruct,,15.737051792828685,,

results_organized/llama3_70b_judge/music_understanding.csv

@@ -1,9 +1,10 @@
 Model,muchomusic_test
 whisper_large_v3,
 gemini-1.5-flash,
-Qwen-Audio-Chat,
-SALMONN_7B,
-Qwen2-Audio-7B-Instruct,
+Qwen-Audio-Chat,59.0564448188711
+SALMONN_7B,50.88458298230834
+old_models,
+Qwen2-Audio-7B-Instruct,71.60909856781802
 WavLLM_fairseq,
 MERaLiON-AudioLLM-Whisper-SEA-LION,57.7927548441449
 cascade_whisper_large_v3_llama_3_8b_instruct,

results_organized/llama3_70b_judge/sds_singlish.csv

@@ -3,6 +3,7 @@ whisper_large_v3,,,,
 gemini-1.5-flash,,,,
 Qwen-Audio-Chat,16.4,16.0,28.2,40.4
 SALMONN_7B,9.0,7.0,17.2,24.2
+old_models,,,,
 Qwen2-Audio-7B-Instruct,33.8,24.8,40.4,46.2
 WavLLM_fairseq,31.6,31.6,45.199999999999996,49.400000000000006
 MERaLiON-AudioLLM-Whisper-SEA-LION,48.4,46.4,57.0,62.599999999999994

results_organized/llama3_70b_judge/speech_instruction.csv

@@ -3,6 +3,7 @@ whisper_large_v3,,
 gemini-1.5-flash,,
 Qwen-Audio-Chat,10.600000000000001,9.8
 SALMONN_7B,15.8,17.2
+old_models,,
 Qwen2-Audio-7B-Instruct,44.800000000000004,52.599999999999994
 WavLLM_fairseq,19.2,21.6
 MERaLiON-AudioLLM-Whisper-SEA-LION,65.6,74.80000000000001

results_organized/llama3_70b_judge/sqa_english.csv

@@ -1,9 +1,10 @@
 Model,slue_p2_sqa5_test,public_sg_speech_qa_test,spoken_squad_test,cn_college_listen_mcq_test,dream_tts_mcq_test
 whisper_large_v3,,,,,
 gemini-1.5-flash,,,,,
-Qwen-Audio-Chat,79.36274509803921,63.16860465116279,64.8327415436367,,
-SALMONN_7B,83.48039215686273,59.24418604651163,66.39506634273968,,
-Qwen2-Audio-7B-Instruct,80.04901960784315,58.31395348837209,64.86264249672958,,
+Qwen-Audio-Chat,79.36274509803921,63.16860465116279,64.8327415436367,63.232056362835756,59.749085206481965
+SALMONN_7B,83.48039215686273,59.24418604651163,66.39506634273968,50.99075297225891,56.455828541557764
+old_models,,,,,
+Qwen2-Audio-7B-Instruct,80.04901960784315,58.31395348837209,64.86264249672958,74.7247908410392,66.49242028227914
 WavLLM_fairseq,83.92156862745098,58.54651162790698,77.64903756307233,,
 MERaLiON-AudioLLM-Whisper-SEA-LION,86.76470588235293,59.7093023255814,73.66473556344609,88.50726552179657,84.31782540512285
 cascade_whisper_large_v3_llama_3_8b_instruct,82.99019607843137,64.94186046511628,83.81984675761541,,

results_organized/llama3_70b_judge/sqa_singlish.csv

@@ -3,6 +3,7 @@ whisper_large_v3,,,,
 gemini-1.5-flash,,,,
 Qwen-Audio-Chat,32.2,37.8,47.800000000000004,51.4
 SALMONN_7B,40.599999999999994,36.6,44.6,46.8
+old_models,,,,
 Qwen2-Audio-7B-Instruct,42.0,39.6,51.6,53.6
 WavLLM_fairseq,45.199999999999996,46.6,50.8,62.199999999999996
 MERaLiON-AudioLLM-Whisper-SEA-LION,51.4,53.2,64.80000000000001,67.2

results_organized/llama3_70b_judge/under_development_llama3_70b_judge.csv

@@ -1,8 +1,9 @@
 Model,ytb_sqa_batch1,ytb_sds_batch1,ytb_pqa_batch1
 whisper_large_v3,,,46.5934065934066
-gemini-1.5-flash,,,
+gemini-1.5-flash,68.31041257367387,,49.908424908424905
 Qwen-Audio-Chat,53.49705304518664,33.44307270233196,37.30769230769231
-SALMONN_7B,,,
+SALMONN_7B,48.46758349705304,21.39917695473251,31.117216117216117
+old_models,,,
 Qwen2-Audio-7B-Instruct,53.87033398821218,45.89849108367627,37.1978021978022
 WavLLM_fairseq,,,
 MERaLiON-AudioLLM-Whisper-SEA-LION,58.192534381139495,45.0480109739369,40.860805860805854

results_organized/meteor/audio_captioning.csv

@@ -3,6 +3,7 @@ whisper_large_v3,,
 gemini-1.5-flash,,
 Qwen-Audio-Chat,0.27553015076950976,0.2355106805560457
 SALMONN_7B,0.20994052484339956,0.17175112770658157
+old_models,,
 Qwen2-Audio-7B-Instruct,0.19891712076314283,0.21342294856199182
 WavLLM_fairseq,0.041732965094428545,0.06399522524688675
 MERaLiON-AudioLLM-Whisper-SEA-LION,0.24920047034353812,0.3175511907248581

results_organized/wer/asr_english.csv

@@ -3,6 +3,7 @@ whisper_large_v3,0.01878749009695552,0.03660128246354058,0.10001863741235596,0.1
 gemini-1.5-flash,,,,,,,,,
 Qwen-Audio-Chat,0.020258799562379748,0.043467569561352074,0.11272421128398918,0.31419144746723354,0.13018910022587737,0.2655529121410546,0.3664994875132684,0.04052375714133636,0.2911540507002305
 SALMONN_7B,0.10270871845172973,0.09671439650443565,0.3062255383962828,0.23699946689025367,0.10765150204693537,0.2577708974886327,0.3597423676988383,0.0459884319222171,0.14231519234178336
+old_models,,,,,,,,,
 Qwen2-Audio-7B-Instruct,0.035141660693401744,0.060415760304159495,0.11438872500819404,0.2165498391593041,0.11723812890302816,0.18872219319407232,0.23542555661330924,0.06114048472375004,0.08739585179932637
 WavLLM_fairseq,0.02103218017882069,0.04798834811886432,0.14533325621300636,0.3792176325635977,0.15491778414546403,0.6447482518259942,0.6671766188447099,0.06621482559171073,0.4536784258110264
 MERaLiON-AudioLLM-Whisper-SEA-LION,0.022918474365262006,0.041576030415949455,0.07811646454714301,0.21050407754683692,0.14457154747310655,0.13488732754499672,0.1652245056860175,0.07884745040985061,0.10228682857649353

results_organized/wer/asr_mandarin.csv

@@ -3,6 +3,7 @@ whisper_large_v3,0.12359684029221357
 gemini-1.5-flash,
 Qwen-Audio-Chat,0.9469917443725129
 SALMONN_7B,0.8259290055631446
+old_models,
 Qwen2-Audio-7B-Instruct,0.09260359129694522
 WavLLM_fairseq,0.7054601967888183
 MERaLiON-AudioLLM-Whisper-SEA-LION,0.13165449110094832

results_organized/wer/asr_singlish.csv

@@ -1,8 +1,9 @@
 Model,imda_part1_asr_test,imda_part2_asr_test,imda_part3_30s_asr_test,imda_part4_30s_asr_test,imda_part5_30s_asr_test,imda_part6_30s_asr_test
-whisper_large_v3,,,,,,
+whisper_large_v3,0.06844171360300393,0.3171008846684522,0.27026366524560785,0.4618189591218298,0.2143555471246589,0.1698509342851144
 gemini-1.5-flash,,,,,,
 Qwen-Audio-Chat,0.10550313315290274,0.45479263046830615,0.6412550574306894,1.173131813552289,0.3016882870525747,0.31394240863063033
 SALMONN_7B,0.0925804013361617,0.42346400454508565,0.6569229098215983,0.7593582215292535,0.34868891450584405,0.24872817713464365
+old_models,,,,,,
 Qwen2-Audio-7B-Instruct,0.07197717796796138,0.1905689473257041,0.35076166942732234,0.5613424034000176,0.27856006770658537,0.2245352799625317
 WavLLM_fairseq,0.10077292565771828,0.4463923382842302,0.7540934640345399,1.143645714142011,0.39796588405247263,0.42541061709652933
 MERaLiON-AudioLLM-Whisper-SEA-LION,0.042254894789457,0.048088629169710254,0.2919053954978684,0.36589168730903193,0.17694182194919086,0.11292172031202054

results_organized/wer/under_development_wer.csv

@@ -2,7 +2,8 @@ Model,cna_test,idpc_test,parliament_test,ukusnews_test,mediacorp_test,idpc_short
 whisper_large_v3,0.13841717398269784,0.19880239520958085,0.0753619074652285,0.07135564378899603,0.12054884024828487,0.1662526275558953,0.05543951935226013,0.06168908700151238,0.11715763436024286,0.12226319428439733,0.17210509244242622,0.7225930420711975,0.5377268970583734
 gemini-1.5-flash,,,,,,,,,,0.1089344703080587,,,
 Qwen-Audio-Chat,0.19753284203780838,0.7710863986313088,0.26279685873781816,0.3158631121194933,0.4498529892192094,0.6008025988916491,0.09347360821020603,0.10399586086125925,0.2548909377108163,0.2297764461857571,0.4315277327278625,0.8783373786407767,1.05567969634822
-SALMONN_7B,0.15395706504325538,0.4550898203592814,0.3010928186204939,0.18918510115333712,0.32089186540346293,0.26313777947639977,0.08676929424202573,0.09042426172092653,0.1751742747919946,,0.3238620391393664,1.2721817691477886,1.0189782362484312
+SALMONN_7B,0.15395706504325538,0.4550898203592814,0.3010928186204939,0.18918510115333712,0.32089186540346293,0.26313777947639977,0.08676929424202573,0.09042426172092653,0.1751742747919946,0.21487285856956287,0.3238620391393664,1.2721817691477886,1.0189782362484312
+old_models,,,,,,,,,,,,,
 Qwen2-Audio-7B-Instruct,0.2067713339741536,0.19093242087254064,0.23270886555019396,0.13843826810361126,0.18694870957203527,0.21326199120963119,0.08416492612361723,0.1194380323171217,0.17180121430177647,0.16843358684796805,0.2080008649583739,0.5522518878101402,0.5486546879304539
 WavLLM_fairseq,0.29856883477517887,0.7686911890504705,0.5216434856656259,0.5911892607298166,0.3595230316889905,0.36728454041658704,0.09512390087929656,0.2066783411605508,,,,1.2913969795037756,1.2204842511249197
 MERaLiON-AudioLLM-Whisper-SEA-LION,0.15924383210509452,0.30008554319931563,0.058922319992430694,0.12554358101720553,0.170859196341065,0.24918784635964075,0.056935097083623425,0.10144869855926132,0.13301101866426804,0.12222478297610817,0.25844956211482323,0.388282092772384,0.35550521901496834