Update README.md

README.md

@@ -49,7 +49,7 @@ The below is the overal performance:
 
 The details of evaluation are as follows:
 <br>
-<b>(Note: The evaluation results we present are based on the initial model of the original training framework. However, the scores showed a decline after converting the framework to Huggingface. Here, we present our complete evaluation results, starting with the initial model results from the paper.)</b>
+<b>(Note: The evaluation results we present are based on the initial model of the original training framework. However, the scores showed some fluctuations after converting the framework to Huggingface. Here, we present our complete evaluation results, starting with the initial model results from the paper.)</b>
 
 <table><thead><tr><th rowspan="2">Task</th><th rowspan="2">Dataset</th><th rowspan="2">Model</th><th colspan="2">Performance</th></tr><tr><th>Metrics</th><th>Results</th></tr></thead><tbody><tr><td rowspan="15">ASR</td><td rowspan="7"><b>Librispeech</b><br>dev-clean | dev-other | <br>test-clean | test-other</td><td>SpeechT5</td><td rowspan="7">WER </td><td>2.1 | 5.5 | 2.4 | 5.8</td></tr><tr><td>SpeechNet</td><td>- | - | 30.7 | -</td></tr><tr><td>SLM-FT</td><td>- | - | 2.6 | 5.0</td></tr><tr><td>SALMONN</td><td>- | - | 2.1 | 4.9</td></tr><tr><td>SpeechVerse</td><td>- | - | 2.1 | 4.4</td></tr><tr><td>Qwen-Audio</td><td>1.8 | 4.0 | 2.0 | 4.2</td></tr><tr><td>Qwen2-Audio</td><td><b>1.3 | 3.4 | 1.6 | 3.6</b></td></tr><tr><td rowspan="2"><b>Common Voice 15</b> <br>en | zh | yue | fr</td><td>Whisper-large-v3</td><td rowspan="2">WER </td><td>9.3 | 12.8 | 10.9 | 10.8</td></tr><tr><td>Qwen2-Audio</td><td><b>8.6 | 6.9 | 5.9 | 9.6</b></td></tr>
 <tr><td rowspan="2"><b>Fleurs</b> <br>zh</td><td>Whisper-large-v3</td><td rowspan="2">WER </td><td>7.7</td></tr><tr><td>Qwen2-Audio</td><td><b>7.5</b></td></tr><tr><td rowspan="4"><b>Aishell2</b> <br>Mic | iOS | Android</td><td>MMSpeech-base</td><td rowspan="4">WER </td><td>4.5 | 3.9 | 4.0</td></tr><tr><td>Paraformer-large</td><td>- | <b>2.9</b> | -</td></tr><tr><td>Qwen-Audio</td><td>3.3 | 3.1 | 3.3</td></tr><tr><td>Qwen2-Audio</td><td><b>3.0</b> | 3.0 | <b>2.9</b></td></tr><tr><td rowspan="8">S2TT</td><td rowspan="5"><b>CoVoST2</b> <br>en-de | de-en | <br>en-zh | zh-en</td><td>SALMONN</td><td rowspan="5">BLEU </td><td>18.6 | - | 33.1 | -</td></tr><tr><td>SpeechLLaMA</td><td>- | 27.1 | - | 12.3</td></tr><tr><td>BLSP</td><td>14.1 | - | - | -</td></tr><tr><td>Qwen-Audio</td><td>25.1 | 33.9 | 41.5 | 15.7</td></tr><tr><td>Qwen2-Audio</td><td><b>29.9 | 35.2 | 45.2 | 24.4</b></td></tr>