Upload folder using huggingface_hub

app/content.py

@@ -69,6 +69,13 @@ displayname2datasetname = {
     'YouTube ASR: Chinese with English Prompt': 'ytb_asr_batch3_chinese',
     'YouTube ASR: Chinese with Chinese Prompt': 'ytb_asr_batch3_zh_zh_prompt',
     
+    'YouTube SQA: Malay': 'ytb_sqa_batch3_malay',
+    'YouTube SQA: Chinese': 'ytb_sqa_batch3_chinese',
+    'YouTube SQA: Tamil': 'ytb_sqa_batch3_tamil',
+    
+    'YouTube SDS: Malay': 'ytb_sds_batch3_malay',
+    'YouTube SDS: Chinese': 'ytb_sds_batch3_chinese',
+    'YouTube SDS: Tamil': 'ytb_sds_batch3_tamil',
     
     'SEAME-Dev-Mandarin'   : 'seame_dev_man',
     'SEAME-Dev-Singlish'   : 'seame_dev_sge',
@@ -158,21 +165,44 @@ dataset_diaplay_information = {
     
     'YouTube ASR: Chinese with Chinese Prompt': 'YouTube Evaluation Dataset for ASR Task: <br> This dataset contains Chinese and some Chinese-English codeswitch audio clips, featuring with Chinese prompts. <br> It includes approximately 3.32 hours of audio, with individual clips ranging from 17 seconds to 1966 seconds in length.',
     
+    'YouTube ASR: Tamil with Tamil Prompt': 'YouTube Evaluation Dataset for ASR Task: <br> This dataset contains Tamil and some Tamil-English codeswitch audio clips, featuring with Tamil prompts. <br> It includes approximately 2.44 hours of audio, with individual clips ranging from 30 seconds to 324 seconds in length.',
+    
+    'YouTube ASR: Tamil with English Prompt': 'YouTube Evaluation Dataset for ASR Task: <br> This dataset contains Tamil and some Tamil-English codeswitch audio clips, featuring with English prompts. <br> It includes approximately 2.44 hours of audio, with individual clips ranging from 30 seconds to 324 seconds in length.',
+    
+    'YouTube ASR Translation: Malay2English': 'YouTube Evaluation Dataset for ASR Task: <br> The audio of dataset is same as <i>YouTube ASR: Malay<i>',
+    
+    # 'YouTube ASR Translation: Chinese2English': 'YouTube Evaluation Dataset for ASR Task: <br> The audio of dataset is same as <i>YouTube ASR: Chinese<i>',
+    
+    # 'YouTube ASR Translation: Tamil2English': 'YouTube Evaluation Dataset for ASR Task: <br> The audio of dataset is same as <i>YouTube ASR: Tamil<i>',
+    
+    
+    
     'SEAME-Dev-Mandarin'   : 'Under Development',
     'SEAME-Dev-Singlish'   : 'Under Development',
 
     'YouTube SQA: English with Singapore Content': 'YouTube Evaluation Dataset for Speech-QA Task: <br> This dataset contains English and Singlish audio clips, featuring Singapore-related content. <br> It includes approximately 7.6 hours of audio, with individual clips ranging from 8 seconds to 32 seconds in length.',
     
+    'YouTube SQA: Malay': 'YouTube Evaluation Dataset for Speech-QA Task: <br> The auido of this dataset is same as <i>YouTube ASR: Malay<i>, it contains Malay and some Malay-English codeswitch audio clips, featuring with English prompts. <br> It includes approximately 2.55 hours of audio, with indicidual clips ranging form 30 seconds to 95 seconds in length.',
+    
+    'YouTube SQA: Chinese': 'YouTube Evaluation Dataset for Speech-QA Task: <br> The auido of this dataset is same as <i>YouTube ASR: Chinese<i>',
+    
+    'YouTube SQA: Tamil': 'YouTube Evaluation Dataset for Speech-QA Task: <br> The auido of this dataset is same as <i>YouTube ASR: Tamil<i>',
+    
     'YouTube SDS: English with Singapore Content': 'YouTube Evaluation Dataset for Summary Task: <br> This dataset contains English and Singlish audio clips, featuring Singapore-related content. <br> It includes approximately 5.4 hours of audio, with individual clips ranging from 8 seconds to 32 seconds in length.',
     
+    'YouTube SDS: Malay': 'YouTube Evaluation Dataset for Speech-QA Task: <br> The auido of this dataset is same as <i>YouTube ASR: Malay<i>, it contains Malay and some Malay-English codeswitch audio clips, featuring with English prompts. <br> It includes approximately 2.55 hours of audio, with indicidual clips ranging form 30 seconds to 95 seconds in length.',
+    
+    'YouTube SDS: Chinese': 'YouTube Evaluation Dataset for Speech-QA Task: <br> The auido of this dataset is same as <i>YouTube ASR: Chinese<i>',
+    
+    'YouTube SDS: Tamil': 'YouTube Evaluation Dataset for Speech-QA Task: <br> The auido of this dataset is same as <i>YouTube ASR: Tamil<i>',
+    
+    
     'YouTube PQA: English with Singapore Content': 'YouTube Evaluation Dataset for Paralinguistics QA Task: <br> This dataset contains English and Singlish audio clips, featuring Singapore-related content. <br> It includes approximately 41.4 hours of audio, with individual clips ranging from 41 seconds to 83 seconds in length.',
     
     
                 }
 
 
-
-
 metrics_info = {
     'wer'                    : 'Word Error Rate (WER) - The Lower, the better.',
     'llama3_70b_judge_binary': 'Model-as-a-Judge Peformance. Using LLAMA-3-70B. Scale from 0-100. The higher, the better.',

app/pages.py

@@ -522,11 +522,6 @@ def music_understanding():
 
 
 
-
-
-
-
-
 def under_development():
     st.title("Task: Under Development")