Repository Issue Activity (beta)

QwenAudio/CosyVoice

Current issue state, recent activity, and per-issue timelines from the indexed issue data.

Open Issues
39
New in 7 Days
1
Closed in 7 Days
17
Average Open Age
292 days
Stale 30+ Days
0
Stale 90+ Days
0
Last 2 Weeks
DateOpenedClosedCommentsEventsOpen Backlog
2026-08-2600000
2026-08-2503481
2026-08-2404480
2026-08-2302372
2026-08-2203360
2026-08-2102482
2026-08-2003482
2026-08-1910366
2026-08-1832264
2026-08-1702251
2026-08-1601351
2026-08-1502240
2026-08-1401362
2026-08-13027104
This Week

Opened: 0

Closed: 17

Comments: 22

Events: 45

Top Labels
stale (65)
Issue Explorer
IssueAuthorStateLabelsCommentsReactionsUpdated

#1242 cosyvoice2 LLM训练WARNING get infinite grad_norm, check your code/data if it appears frequently

Opened 1 year ago
CrazyThrusdayV50
closed - completed
stale
302 days ago

#1243 unable to start webui server

Opened 1 year ago
bli1348
closed - completed
stale
302 days ago

#1244 可以训练新的语言吗?

Opened 1 year ago
Yarken00
open
stale
302 days ago

#1903 Cosyvoice3在使用自然语言指令时会把指令读出来

Opened 3 months ago
trimonster233
closed - completed
stale
502 days ago

#1245 模型调用是线程安全的吗? cosyvoice = CosyVoice2(model_dir=model_path, load_jit=False, load_trt=False, fp16=False)

Opened 1 year ago
kandada
closed - completed
stale
303 days ago

#1246 推理硬件选择及模型量化?

Opened 1 year ago
Gusha-nye
closed - completed
stale
303 days ago

#1247 实时输出24000的PCM音频流,流式播放需要8000的采样率,请问这个怎么解决?

Opened 1 year ago
BoragoCode
closed - completed
stale
303 days ago

#1304 无法复现作者zeroshot推理的相似度

Opened 1 year ago
boji123
closed - completed
stale
1903 days ago

#1248 音频重建相关问题

Opened 1 year ago
VJJJJJJ1
closed - completed
stale
304 days ago

#1249 单词读错

Opened 1 year ago
xipingL
open
No labels
304 days ago

#1250 如何使用inference_vc完美的复刻音频的情绪和语气?

Opened 1 year ago
wei-rh
open
stale
304 days ago

#1310 使用自己训练的模型打开jit,trt之后,发现没有原始模型加速效果明显

Opened 1 year ago
brooks348
closed - completed
stale
414 days ago

#1251 汉语多音字读的不准

Opened 1 year ago
crp2022
closed - completed
stale
305 days ago

#1253 CosyVoice2-0.5B遇到的问题

Opened 1 year ago
pawacat
closed - completed
stale
405 days ago

#1254 英文单词读错误,或者直接就给跳过去了

Opened 1 year ago
lsh-boop
closed - completed
stale
405 days ago

#1255 cosyvoice2训练问题

Opened 1 year ago
zhusy09
open
stale
1206 days ago

#1259 Decoder.__init__() got an unexpected keyword argument 'causal'")

Opened 1 year ago
jetsonearth
closed - completed
stale
606 days ago

#1361 流式flow+TensorRT推理,偶现句子变成静音段。

Opened 1 year ago
TythonLee
open
stale
606 days ago

#1912 Inference requires training-only dependencies (pyworld, librosa) due to YAML config referencing dataset.processor

Opened 2 months ago
Iceberg-Yang
closed - completed
stale
306 days ago

#1260 把“第三块饼”读成“第三[break]块饼”

Opened 1 year ago
dsun29
closed - completed
stale
307 days ago

#1262 The inferrence speed is too slow

Opened 1 year ago
OutisLi
open
No labels
317 days ago

#1321 提高flow模型训练速度和GPU浮点利用率的方法收集

Opened 1 year ago
JohnHerry
closed - completed
stale
1107 days ago

#1323 cosyvoice3

Opened 1 year ago
brooks348
closed - completed
stale
501217 days ago

#1936 把权重从系统内存拷贝到显存OOM问题

Opened 7 days ago
QUBOOR-CREATE
open
No labels
107 days ago

#1626 使用grpo进行训练时,默认参数actor_rollout_ref.actor.use_kl_loss=False是不是应该改为true?

Opened 10 months ago
huangjun0413
closed - completed
stale
407 days ago

Rows per page:

1–25 of 97