Open MELABIPCAS opened 7 months ago
conf/train_asr_conformer_w2v2-base.yaml文件中做修改如下:
freeze_param: [ "frontend.upstream" ]
frontend: s3prl frontend_conf: frontend_conf: upstream: wav2vec2_local upstream_model_config: upstream_ckpt: ./chinese-wav2vec2-base.pt download_dir: ./hub multilayer_feature: true
preencoder: linear preencoder_conf: input_size: 768 # Note: If the upstream is changed, please change this value accordingly. output_size: 80 其中upstream_ckpt:中的选项后面 ./chinese-wav2vec2-base.pt是您们已经训练好的模型,不知道这样修改得对不对? 非常感谢!!!
+1,我也遇到了,请问解决了吗
conf/train_asr_conformer_w2v2-base.yaml文件中做修改如下:
network architecture
pretrained model related
freeze_param: [ "frontend.upstream" ]
frontend: s3prl frontend_conf: frontend_conf: upstream: wav2vec2_local upstream_model_config: upstream_ckpt: ./chinese-wav2vec2-base.pt download_dir: ./hub multilayer_feature: true
preencoder: linear preencoder_conf: input_size: 768 # Note: If the upstream is changed, please change this value accordingly. output_size: 80 其中upstream_ckpt:中的选项后面 ./chinese-wav2vec2-base.pt是您们已经训练好的模型,不知道这样修改得对不对? 非常感谢!!!