yl4579/styletts issues and pull requests

#81 - Unable to Access Pretrained Hifi-GAN and StyleTTS Model Links

Issue - State: open - Opened by nqwu 26 days ago

#80 - Questions regarding the pretrained text aligner

Issue - State: open - Opened by gnitoah about 1 month ago

#79 - Finetune multispeaker model

Issue - State: open - Opened by Atharva1006 about 2 months ago

#78 - Train both stages consecutively

Issue - State: open - Opened by Atharva1006 3 months ago

#77 - Is there a first-stage pretrained model?

Issue - State: open - Opened by cnint0627 5 months ago

#76 - can't import mask_from_lens from monotonic_align

Issue - State: open - Opened by cnint0627 5 months ago - 3 comments

#75 - amazing work,can it support that generate the phoneme delayed time sequence?

Issue - State: open - Opened by CasonTsai 6 months ago

#74 - MelSpectrogram() and unspecified sampling rate

Issue - State: open - Opened by dsplog 9 months ago - 1 comment

#73 - Voice Quality issue using Librispeech

Issue - State: open - Opened by Anshu-Kumar-1 12 months ago

#72 - Marathi Support ?

Issue - State: open - Opened by raushanagrawal about 1 year ago

#71 - Is the uv detector trained in the pretrained pitch detector?

Issue - State: open - Opened by auspicious3000 about 1 year ago

#70 - Has anyone had this problem when converting to onnx?

Issue - State: open - Opened by bobo-paopao about 1 year ago

#69 - S2S

Issue - State: closed - Opened by pawanhv about 1 year ago - 1 comment

#68 - StyleTTS 2

Issue - State: closed - Opened by fakerybakery about 1 year ago - 1 comment

#67 - continue correct number of epoch from pretrained

Pull Request - State: open - Opened by magicse about 1 year ago

#66 - Update train_second.py

Pull Request - State: closed - Opened by magicse about 1 year ago

#65 - question

Issue - State: closed - Opened by magicse about 1 year ago - 2 comments

#64 - Pre-training model sound quality issues

Issue - State: closed - Opened by gachaun about 1 year ago - 5 comments

#63 - train on ESD

Issue - State: closed - Opened by bobo-paopao over 1 year ago - 1 comment

#62 - Inference exact time for each word

Issue - State: closed - Opened by enla51 over 1 year ago - 5 comments

#61 - Question: Fine tuning LibriTTS with StyleTTS

Issue - State: closed - Opened by Yahya-khodr over 1 year ago - 5 comments

#60 - The pronunciations of single words or short words is poor?

Issue - State: closed - Opened by GuangChen2016 over 1 year ago - 1 comment

#59 - Training Model with new Dataset

Issue - State: closed - Opened by Yahya-khodr over 1 year ago

#58 - Training the model

Issue - State: closed - Opened by Yahya-khodr over 1 year ago - 1 comment

#57 - what is the mean=-4 and std=4 meannig?

Issue - State: closed - Opened by skysbird over 1 year ago - 1 comment

#56 - First stage alignment training failed when TMA_CEloss=True

Issue - State: closed - Opened by auspicious3000 over 1 year ago - 2 comments

#55 - style encoder inconsistency

Issue - State: closed - Opened by auspicious3000 over 1 year ago - 1 comment

#54 - batch size and number of epochs for large dataset

Issue - State: closed - Opened by auspicious3000 over 1 year ago - 1 comment

#53 - 多卡训练的问题

Issue - State: closed - Opened by bobo-paopao over 1 year ago - 5 comments

#52 - About train on Vietnamese Dataset

Issue - State: closed - Opened by christopherohit over 1 year ago - 1 comment

#51 - hi，如果使用v4000 单卡训练LibriTTS数据集，显卡内存只有16G， batch_size只能设置为8，训练时常大概为多长呢？谢谢

Issue - State: closed - Opened by bobo-paopao over 1 year ago - 1 comment

#50 - Need help for training

Issue - State: closed - Opened by nhanhttrong over 1 year ago - 5 comments

#49 - s2s loss 和mono loss在第一阶段的训练中一直为0

Issue - State: closed - Opened by bobo-paopao over 1 year ago - 1 comment

#48 - how to Inference

Issue - State: closed - Opened by luwentao1989 over 1 year ago - 1 comment

#47 - Source of Hifigan checkpoints?

Issue - State: closed - Opened by kmn1024 over 1 year ago - 1 comment

#46 - pretrained model of stage 1

Issue - State: closed - Opened by nhanhttrong over 1 year ago

#45 - Can you offer the loss log of style tts

Issue - State: closed - Opened by hdmjdp over 1 year ago - 1 comment

#44 - asr phone dict different from this

Issue - State: closed - Opened by hdmjdp over 1 year ago - 1 comment

#43 - about ASR model

Issue - State: closed - Opened by hdmjdp over 1 year ago - 1 comment

#42 - Would you recommend changing the code in the inference notebook when running a PL-BERT finetuned model?

Issue - State: closed - Opened by ghost over 1 year ago - 1 comment

#41 - error while running train_second.py (caused by size mismatch)

Issue - State: closed - Opened by ghost over 1 year ago - 2 comments

#40 - Code for Emotional Speech Synthesis

Issue - State: closed - Opened by satani99 over 1 year ago - 1 comment

#39 - Questions about the Evaluations

Issue - State: closed - Opened by Zhongxu-Wang over 1 year ago - 1 comment

#38 - turns out your code doesn't join the read wav paths from train_list.txt file with the dataset path (the location of train_list.txt)

Issue - State: closed - Opened by ghost over 1 year ago

#37 - running train_first.py raises error

Issue - State: closed - Opened by ghost over 1 year ago - 3 comments

#36 - Any-to-any and emotion examples

Issue - State: closed - Opened by sleimanitani over 1 year ago - 6 comments

#35 - Phoneme sequence padding

Issue - State: closed - Opened by WorkingJack over 1 year ago - 3 comments

#34 - Fixing librosa compatibility issue and UTF-8 issue

Pull Request - State: closed - Opened by Artyom17 over 1 year ago

#33 - What's r1_reg loss?

Issue - State: closed - Opened by splinter21 over 1 year ago - 1 comment

#33 - What's r1_reg loss?

Issue - State: closed - Opened by splinter21 over 1 year ago - 1 comment

#32 - Probleam about data processing

Issue - State: closed - Opened by Zhongxu-Wang over 1 year ago - 1 comment

#31 - crashes during training

Issue - State: closed - Opened by ppisljar over 1 year ago - 4 comments

#30 - Why don't use "attention_weight" in train_first.py ?

Issue - State: closed - Opened by dy2009 over 1 year ago - 1 comment

#29 - Why I can't use your mel to train HiFi-Gan Vocoder ?

Issue - State: closed - Opened by dy2009 over 1 year ago - 1 comment

#28 - f0_extractor ?

Issue - State: closed - Opened by dy2009 over 1 year ago

#27 - monotonic error

Issue - State: closed - Opened by zdj97 over 1 year ago - 1 comment

#26 - s2s loss become nan

Issue - State: closed - Opened by startreker-shzy over 1 year ago - 2 comments

#25 - Trying to have the inference demo work - but no success

Issue - State: closed - Opened by olup over 1 year ago - 3 comments

#24 - train from scratch of another phoneme set

Issue - State: closed - Opened by zdj97 over 1 year ago - 5 comments

#23 - Require Imports??

Issue - State: closed - Opened by ziyaad30 over 1 year ago - 1 comment

#22 - Pause between sentences?

Issue - State: closed - Opened by ziyaad30 almost 2 years ago - 5 comments

#21 - Fix Requirements, Add Colab

Pull Request - State: open - Opened by nivibilla almost 2 years ago - 10 comments

#20 - Training on custom data

Issue - State: closed - Opened by sulaiman-kagumire almost 2 years ago - 1 comment

#19 - Finetuning

Issue - State: closed - Opened by nivibilla almost 2 years ago - 1 comment

#18 - train_second step not fix decoder?

Issue - State: closed - Opened by liuhuang31 almost 2 years ago - 1 comment

#17 - finetune the vocoder

Issue - State: closed - Opened by sinhprous1 almost 2 years ago - 5 comments

#16 - Sounds weird when the input is short

Issue - State: closed - Opened by harryfyodor almost 2 years ago - 1 comment

#15 - Multilingual Training [Question]

Issue - State: closed - Opened by lexkoro almost 2 years ago - 2 comments

#14 - Adding BigVGAN as Vocoder

Issue - State: closed - Opened by crypticsymmetry almost 2 years ago - 18 comments

#13 - Using Single Gpu Pretrained model on Multi-Gpu training

Issue - State: closed - Opened by crypticsymmetry almost 2 years ago - 7 comments

#12 - Process for setting up other datasets

Issue - State: closed - Opened by crypticsymmetry almost 2 years ago - 1 comment

#11 - Training the model end2end

Issue - State: closed - Opened by erogol about 2 years ago - 7 comments
Labels: discussion

#10 - mandrain support?

Issue - State: open - Opened by lucasjinreal about 2 years ago - 102 comments
Labels: help wanted

GitHub / yl4579/styletts issues and pull requests