Search results

RoPE uses wrong dimension as a sequence length in Vocos implementation

Status: Open.

#36 In zhenye234/X-Codec-2.0;

· Uncomfy opened on Aug 29, 2025

How to train the model with multi nodes?

Status: Open.

#35 In zhenye234/X-Codec-2.0;

· TZYSJTU opened on Jul 30, 2025

How to resume ckpt for training ?

Status: Open.

#34 In zhenye234/X-Codec-2.0;

· TZYSJTU opened on Jul 30, 2025

Make <code>requirements.txt</code> less strict to make <code>xcodec2</code> module usable

Status: Open.

#33 In zhenye234/X-Codec-2.0;

· 3manifold opened on Jul 23, 2025

Reproducibility of X-codec1 metrics in Table 1

Status: Open.

#32 In zhenye234/X-Codec-2.0;

· francissfy opened on Jul 22, 2025

[Q] Training Details

Status: Open.

#30 In zhenye234/X-Codec-2.0;

· gevmin94 opened on May 26, 2025

X-codec-2.0 supports streaming decode

Status: Open.

#28 In zhenye234/X-Codec-2.0;

· saifulislam79 opened on Apr 13, 2025

torch &amp; numpy version pins in the pypi package

Status: Open.

#27 In zhenye234/X-Codec-2.0;

· AbrahamSanders opened on Apr 7, 2025

About the Selection of Acoustic Decoder

Status: Open.

#26 In zhenye234/X-Codec-2.0;

· VJJJJJJ1 opened on Apr 3, 2025

Using <code>transformers==4.49.0</code> leads to incorrect results

Status: Open.

#24 In zhenye234/X-Codec-2.0;

· oisinturbitt opened on Mar 6, 2025

NameError: name 'Wav2Vec2BertModel' is not defined

Status: Open.

#23 In zhenye234/X-Codec-2.0;

· tianyashu2020 opened on Mar 6, 2025

model_spk = init_model('wavlm_large','/aifs4su/data/zheny/models_fd_ckpt/wavlm_large_finetune.pth')

Status: Open.

#22 In zhenye234/X-Codec-2.0;

· tianyashu2020 opened on Mar 6, 2025