 Ethanfel
|
3cfbd03da3
|
Run tests with repository on Python path
|
2026-09-06 23:01:27 +02:00 |
|
|
Ethanfel
|
68239634ea
|
Harden SAM-Audio loading and inference
|
2026-09-06 22:59:57 +02:00 |
|
|
Ethanfel
|
dca2f22766
|
Use canonical T5 repository for Xet downloads
|
2026-09-06 22:08:44 +02:00 |
|
|
Ethanfel
|
9ca7d45794
|
Add native ComfyUI video separation node
|
2026-08-23 17:48:45 +02:00 |
|
|
Ethanfel
|
e74fa69e2d
|
Use unique Comfy Registry node ID
|
2026-08-23 17:28:33 +02:00 |
|
|
Ethanfel
|
901525e23c
|
Add Comfy Registry publishing workflow
|
2026-08-23 17:25:51 +02:00 |
|
|
Ethanfel
|
79e2425116
|
Integrate model lifecycle and lazy span loading
|
2026-08-23 17:08:31 +02:00 |
|
|
Ethanfel
|
602a3d1eec
|
Handle strided tensors in Kitchen attention
|
2026-08-23 16:48:33 +02:00 |
|
|
Ethanfel
|
373cc07c7a
|
Ensure SAM-Audio checkpoint weights are loaded
|
2026-08-23 16:20:10 +02:00 |
|
|
Ethanfel
|
18843d0ae4
|
Reuse cached support model snapshots
|
2026-08-23 16:13:59 +02:00 |
|
|
Ethanfel
|
5fc75a99b1
|
Add overlapping chunk processing
|
2026-08-23 16:10:29 +02:00 |
|
|
Ethanfel
|
50ffce7c88
|
Fix current Hub loading and demo sampler values
|
2026-08-23 15:50:30 +02:00 |
|
|
Ethanfel
|
e6fe77dbbe
|
Avoid TorchCodec for tensor-native inference
|
2026-08-23 15:41:34 +02:00 |
|
|
Ethanfel
|
1f27b8d316
|
Add Comfy Kitchen attention backend
|
2026-08-23 15:36:32 +02:00 |
|
|
Ethanfel
|
fe72f2ce74
|
Add verified model mirror fallback
|
2026-08-23 15:21:26 +02:00 |
|
|
Ethanfel
|
53d0ef6e16
|
Initial ComfyUI SAM-Audio integration
|
2026-08-23 15:01:20 +02:00 |
|