whisper-large-v3-turbo微调模型

对whisper-large-v3-turbo开源模型用粤语和四川话数据集进行微调,得到微调后的模型。

微调后测试cer

带符号 去符号
微调前 0.8835504720183913 0.831893711149071
微调后 0.16332876251694223 0.14408498883946694

微调所用数据集

ChenChenyu/VoiceDataSet

Downloads last month
24
Safetensors
Model size
809M params
Tensor type
F32
·
Inference API
Unable to determine this model's library. Check the docs .

Model tree for ChenChenyu/whisper-large-v3-turbo-finetuned

Finetuned
(125)
this model

Dataset used to train ChenChenyu/whisper-large-v3-turbo-finetuned