-
4989e0a7e8
Fix audio quality: use original orpheus_tts convert_to_audio decoder
main
Alex
2026-04-14 01:07:29 -05:00
-
57a2e24101
Fix SNAC decoding: correct token offset + device attribute
Alex
2026-04-13 16:41:14 -05:00
-
16aa526656
Fix SNAC code offset: subtract per-layer offset (position*4096)
Alex
2026-04-13 16:04:54 -05:00
-
d650fd06b9
OrpheusTail v2: transformers streaming engine (replaces vLLM)
Alex
2026-04-13 08:38:30 -05:00
-
cfc9b1a5a0
Revert to sync LLM + sentence-level streaming
Alex
2026-04-12 23:45:11 -05:00
-
25ed6625aa
True streaming TTS: AsyncLLMEngine + incremental token decoding
Alex
2026-04-12 23:36:24 -05:00
-
14af1d0600
token limit and chunking
Alex
2026-02-06 10:07:05 -06:00
-
75a5fc0a95
Fix streaming endpoint max_tokens limit - Day 72
vixy
2026-01-12 16:56:43 -06:00
-
0fa4042025
Increase max_tokens from 1200 to 4000 - Day 72
vixy
2026-01-12 16:41:01 -06:00
-
96cd33732d
Fix audio assembly - chunks are already bytes from SNAC decoder
vixy
2026-01-11 19:47:19 -06:00
-
fe43eda6bd
Fix token extraction - use regex to find custom_token patterns
vixy
2026-01-11 19:33:31 -06:00
-
af35dc46d5
Use sync vllm.LLM instead of AsyncLLMEngine to avoid event loop conflicts
vixy
2026-01-11 18:58:12 -06:00
-
0b88188907
Debug: add verbose logging to generate_speech_sync
vixy
2026-01-11 18:44:07 -06:00
-
4eab3ccc01
Fix: wrap sync generator in executor, not async for
vixy
2026-01-11 18:32:06 -06:00
-
4d11334f33
Fix async iteration over vLLM generator - use async for instead of sync for
vixy
2026-01-11 18:18:37 -06:00
-
a164bed590
Fix _map_model_params call signature
vixy
2026-01-11 17:59:49 -06:00
-
d0d7633a00
Monkey-patch OrpheusModel to support max_model_len on Jetson
vixy
2026-01-11 17:52:33 -06:00
-
0e43b76204
Use GitHub orpheus-tts (supports max_model_len) to fix OOM on Jetson
vixy
2026-01-11 17:39:55 -06:00
-
86cf77d2d9
Add HuggingFace token for gated model access
vixy
2026-01-11 17:29:30 -06:00
-
ec965580ae
Try medium-3b model name for PyPI package
vixy
2026-01-11 17:23:49 -06:00
-
8cc9154080
Fix: remove unsupported max_model_len param for PyPI package
vixy
2026-01-11 17:17:48 -06:00
-
5d69182bdf
Fix: use regular PyPI for orpheus-speech on Jetson
vixy
2026-01-11 17:11:58 -06:00
-
28d6df98b8
Use dustynv/vllm base image for Jetson CUDA support
vixy
2026-01-11 16:15:12 -06:00
-
453271e49a
Add .gitignore
vixy
2026-01-11 15:51:34 -06:00
-
ed579a77ee
Initial commit: OrpheusTail TTS service
vixy
2026-01-11 15:51:08 -06:00