-
Notifications
You must be signed in to change notification settings - Fork 2.1k
Pull requests: fishaudio/fish-speech
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Fix #1120: Noise issue after LoRA merge and inference
#1160
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #852: Low Model Accuracy After Extended Training with Unicode-Base
#1159
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #819: Why is the streaming parameter of the TTS interface set to T
#1158
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #1118: Crash on startup: AttributeError: torchaudio.list_audio_back
#1157
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #1133: how to change the return audio length in streaming mode
#1156
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #834: When the 'compile' option is enabled, using uvicorn to start
#1155
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #1142: [BUG] api server doesn’t start with workers > 1 in docker (u
#1153
opened Feb 21, 2026 by
danielalanbates
Loading…
Fix #1143: [BUG] only one worker runs with workers > 1 due to CUDA fork
#1152
opened Feb 21, 2026 by
danielalanbates
Loading…
fix: use batch-safe indexing in top-p filtering to support batched logits
#1151
opened Feb 11, 2026 by
Mr-Neutr0n
Loading…
1 of 2 tasks
fix: apply temperature scaling before top-p filtering in logits_to_probs
#1150
opened Feb 11, 2026 by
Mr-Neutr0n
Loading…
3 tasks
fix: add torchaudio 2.9 compatibility for list_audio_backends removal
#1148
opened Feb 7, 2026 by
Mr-Neutr0n
Loading…
fix(server): make uvicorn workers>1 work (factory + spawn for CUDA)
#1141
opened Jan 8, 2026 by
vladlearns
•
Draft
ProTip!
Mix and match filters to narrow down what you’re looking for.