ASRT_SpeechRecognition
ASRT_SpeechRecognition copied to clipboard
nl8590687
A Deep-Learning-Based Chinese Speech Recognition System 基于深度学习的中文语音识别系统
Hello, Do you think this project is suitable for the code-switching topic, especially in Mandarin-English?
模型在训练时，保存一次模型之后会出现如下错误。本人电脑显卡为GeForce 1060，batch_size设置为8，请问是因为显卡内存不够的原因吗？ 2020-06-08 21:06:33.504024: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1618] Found device 0 with properties: name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate(GHz): 1.7845 pciBusID: 0000:01:00.0 2020-06-08 21:06:33.504241: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened...
You said this repo has at least one model which is created with attention. But I can't find any. could you edit that? Or show me which model is attention...
我用的是v0.6.0版本，windows系统，单独跑test.py是没有问题的，参考asrserver.py，我重新基于Flask写了一个简单的api服务，如下： import json from flask import Flask, request, jsonify from SpeechModel251 import ModelSpeech from LanguageModel2 import ModelLanguage from keras import backend as K from general_function.file_wav import * datapath = "./"...
input_length.append(data_input.shape[0] // 8 + data_input.shape[0] % 8)
感谢作者的分享。 用了data_thchs30和ST-CMDS-20170001_1-OS数据，只有CPU，一天24h只能训练4500step，训练到3万次效果还不是很好（和作者的62.5万次还有差距），2.4万次的时候loss=25(供大家参考)。 预测拼接结果还是有一些是正确的，问题出在拼接转汉字，语音模型能用transformer替代希望能有提高。
Hi, I used my collected data to train the model, but unfortunately, there is an error during the training process. Here is the log information.  Can u give me...
I tried to use TF2.0 but there're several functions which are deprecated. I made changes to the code, to make it compatible with TF2.0. If you're thinking of testing or...
