FUNASR训练
blame | 历史 | 补丁 | 提交 | 提交对比 | ignore whitespace
游雁
2024-06-05 46568227c9ff0a1f6ebf9e398dfdb6ad51cce4c5 
 funasr/models/llm_asr/adaptor.py


@@ -84,7 +84,7 @@


        self.blocks = nn.ModuleList(


            [


                EncoderLayer(


                    output_size,


                    llm_dim,


                    MultiHeadedAttention(


                        kwargs.get("attention_heads", 8),


                        llm_dim,