feat(audio8_tts): Falcon-H1 0.1B (Mamba2 + attention) support and codec/AR performance - #444

Open

gqf2008 wants to merge 55 commits into

0xShug0:mainfrom

gqf2008:feat/audio8-tts-falcon-h1-01b

Commits on Sep 4, 2026

feat(audio8_tts): Falcon-H1 0.1B (Mamba2 + attention) stateful forward port
gqf2008
committed
993a9a2
chore(ggml-metal): renormalize ggml-metal.metal CRLF -> LF
gqf2008
committed
0f8b614
fix: Falcon-H1 state corruption and numerical stability
gqf2008
committed
720c3b5
docs: record Falcon-H1 0.1B port status and open issues
gqf2008
committed
332a777
docs(audio8_tts): mark Falcon-H1 open issues resolved with root causes
gqf2008
committed
01e3e0d
perf(audio8_tts): run fast AR graph on a dedicated CPU backend
gqf2008
committed
e44fc03
feat(audio8_tts): time codec decode graph build vs compute
gqf2008
committed
e81db68
perf(conv): Metal fast path for stride-1 conv1d on time-fast layouts
gqf2008
committed
24ec127
perf(audio8_tts): run Falcon-H1 per-token step graph on the CPU backend
gqf2008
committed
7757f0e
perf(audio8_tts): keep Falcon KV cache in padded host buffers, write slots in-graph
gqf2008
committed
f281804
perf(audio8_tts): reuse the Falcon step graph per KV capacity bucket
gqf2008
committed
c5b3186
perf(audio8_tts): accumulate conv1d per-tap GEMMs in-place via ggml_mul_mat_acc
gqf2008
committed
24a245b
perf: fused snake1d activation op (GGML_OP_SNAKE_1D) for audio8_tts codec
gqf2008
committed
c08b443
perf(audio8_tts): zero-copy Falcon step state and constants on CPU backend
gqf2008
committed
855b7ef
perf(audio8_tts): run decoder blocks in channel-fast layout on Metal
gqf2008
committed
b2b9708
fix(ggml-metal): ssm_scan reduction reads garbage when sgptg < NW
gqf2008
committed
71325ea
fix(audio8_tts): Falcon-H1 0.1B logits parity and long-run stability
gqf2008
committed
fcba3a4
Fix soprano_tts maintainer information in README (#451)
drzsdrtfg
authored
3198855
Add BreezeTTS and CosyVoice3 model wiring
0xShug0
committed
ba32fa0
Normalize CosyVoice3 and Breeze GGUF package names
0xShug0
committed
2139887
Add CosyVoice3 and Breeze UI support
0xShug0
committed
2d8fea1
Move Breeze voice design to design route
0xShug0
committed
b34c583
perf(breeze_tts): pre-allocate depth projection & generation staging buffers (#348)

FraserHum
authored and
0xShug0
committed
96c90bc
Add BreezeTTS streaming mode
0xShug0
committed
7c9d1a1
Breeze-TTS 2 performance: weight packing + fused bf16 rounding (#393)

IIIIIllllIIIIIlllll
authored and
0xShug0
committed
ff38f15
Breeze encoder chunked vram (#431)

IIIIIllllIIIIIlllll
authored and
0xShug0
committed
8a9cb75
Merge pull request #452 from 0xShug0/dev
0xShug0
authored
11e19c1
Update release notes and contributor acknowledgments
0xShug0
committed
99d07ed
Update native UI model catalog
0xShug0
committed
edc9712
Fix CosyVoice3 Metal flow layout (#455)
0xShug0
authored
a2a4d4f
ggml: additive INT8/ternary types, fused CPU ops, and kernels for VibeASR (1/2) (#447)
XsquirrelC
authored
62a9441
server: add /v1/audio/transcriptions/details for transcript detail fields (#401)

authored
801c6a2
[Feature] ModelScope download (#450)
IIIIIllllIIIIIlllll
authored
70f64fe
sanotts: sanoTTS voice family community model (seven voices, 294k-2.27M params, en/vi/id, GGUF FP32) (#449)
voidash
authored
cd98afb
higgs_audio_tts: give the codec decoder right-hand context at the end of the stream (#454)

niksedk
and
claude
authored
0befc6f
Add sanoTTS to native WebUI
0xShug0
committed
f4f8405

Commits on Sep 5, 2026

Add PocketTTS streaming support (#458)
0xShug0
authored
a8fccb4
community_models: VibeASR INT8/ternary ASR pipeline (2/2) (#448)
XsquirrelC
authored
8bcaa94
Add 'Ctrl' variant to breeze_tts entry (#463)
Th-Underscore
authored
5e64ee6
gguf: let a finished GGUF be re-converted (#464)

CaptainArni
and
claude
authored
49b0322
Add Vulkan Docker image support (#466)
0xShug0
authored
ca804b5
Add CLI and server version output (#469)
0xShug0
authored
c41ffef

Commits on Sep 6, 2026

Commits on Sep 7, 2026

Use ViewKV flash attention for Qwen prefix prefill (#477)
0xShug0
authored
c55826b
Add VoxCPM2 stateless chunk strategy (#478)
0xShug0
authored
19c44f0
Buffer chunked HTTP uploads (#479)
0xShug0
authored
9c2f238
fix(build): guard CUDA driver probe out of HIP builds (#480)
IIIIIllllIIIIIlllll
authored
baf0bb8
Make music generation UI friendlier
0xShug0
committed
e445bee
Add Vevo2 source audio chunking (#482)
0xShug0
authored
a10738a
Merge commit 'fcba3a4' into feat/audio8-tts-falcon-h1-01b
gqf2008
committed
5d0a801