@omote/core 0.5.7 → 0.6.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -35,11 +35,7 @@ The most common use case: feed TTS audio chunks and get back 52 ARKit blendshape
35
35
  import { FullFacePipeline, createA2E } from '@omote/core';
36
36
 
37
37
  // 1. Create A2E backend (auto-detects GPU vs CPU)
38
- const lam = createA2E({
39
- gpuModelUrl: '/models/lam-wav2vec2.onnx',
40
- cpuModelUrl: '/models/wav2arkit_cpu.onnx',
41
- mode: 'auto',
42
- });
38
+ const lam = createA2E(); // auto-detects GPU vs CPU, fetches from HF CDN (192MB fp16)
43
39
  await lam.load();
44
40
 
45
41
  // 2. Create pipeline with expression profile
@@ -72,12 +68,7 @@ Auto-detects platform: Chrome/Edge/Android use WebGPU, Safari/iOS use WASM CPU f
72
68
  ```typescript
73
69
  import { createA2E } from '@omote/core';
74
70
 
75
- const a2e = createA2E({
76
- gpuModelUrl: '/models/lam-wav2vec2.onnx', // 384MB, WebGPU
77
- cpuModelUrl: '/models/wav2arkit_cpu.onnx', // 404MB, WASM
78
- mode: 'auto', // 'auto' | 'gpu' | 'cpu'
79
- fallbackOnError: true, // GPU failure → auto-switch to CPU
80
- });
71
+ const a2e = createA2E(); // auto-detects: GPU (192MB fp16) or CPU (404MB WASM)
81
72
  await a2e.load();
82
73
 
83
74
  const { blendshapes } = await a2e.infer(audioSamples); // Float32Array (16kHz)
@@ -89,7 +80,7 @@ const { blendshapes } = await a2e.infer(audioSamples); // Float32Array (16kHz)
89
80
  ```typescript
90
81
  import { Wav2Vec2Inference, LAM_BLENDSHAPES } from '@omote/core';
91
82
 
92
- const lam = new Wav2Vec2Inference({ modelUrl: '/models/lam-wav2vec2.onnx' });
83
+ const lam = new Wav2Vec2Inference({ modelUrl: '/models/model_fp16.onnx' });
93
84
  await lam.load();
94
85
 
95
86
  const { blendshapes } = await lam.infer(audioSamples);
@@ -317,7 +308,7 @@ Place models in your public assets directory:
317
308
 
318
309
  ```
319
310
  public/models/
320
- lam-wav2vec2.onnx # A2E lip sync — WebGPU (384MB)
311
+ model_fp16.onnx # A2E lip sync — WebGPU (192MB fp16, from omote-ai/lam-a2e)
321
312
  wav2arkit_cpu.onnx # A2E lip sync — WASM fallback (1.86MB graph)
322
313
  wav2arkit_cpu.onnx.data # A2E lip sync — WASM fallback (402MB weights)
323
314
  sensevoice/model.int8.onnx # SenseVoice ASR (239MB)