kazeia

Commit Graph

Author	SHA1	Message	Date
Kazeia Team	a688edc9ec	Reduce talker KV_LEN 100→64: saves 148ms (RTF 1.31) KV window of 64 sufficient for ~70 token generation (10 prefill + 58 gen). 36% less KV memcpy per talker step (28L × 2 × 64×8×128 vs 100×8×128). Generation: 3795ms → 3647ms, total: 6438ms → 6093ms Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-09 12:47:30 +02:00
Kazeia Team	389ffa7c61	Initial commit: Kazeia TTS pipeline on NPU via ExecuTorch Full Qwen3-TTS-0.6B pipeline running on Snapdragon 8 Elite NPU: - Talker (28L) and Code Predictor (5L) as .pte on QNN HTP fp16 - JNI integration, no root required - Validated audio quality: RTF 3.9 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-09 08:42:11 +02:00

Author

SHA1

Message

Date

Kazeia Team

a688edc9ec

Reduce talker KV_LEN 100→64: saves 148ms (RTF 1.31)

KV window of 64 sufficient for ~70 token generation (10 prefill + 58 gen).
36% less KV memcpy per talker step (28L × 2 × 64×8×128 vs 100×8×128).

Generation: 3795ms → 3647ms, total: 6438ms → 6093ms

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-09 12:47:30 +02:00

Kazeia Team

389ffa7c61

Initial commit: Kazeia TTS pipeline on NPU via ExecuTorch

Full Qwen3-TTS-0.6B pipeline running on Snapdragon 8 Elite NPU:
  - Talker (28L) and Code Predictor (5L) as .pte on QNN HTP fp16
  - JNI integration, no root required
  - Validated audio quality: RTF 3.9

  Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-09 08:42:11 +02:00

2 Commits