mlboydaisuke commited on
Commit
6d64def
·
verified ·
1 Parent(s): d7805b0

litertlm_manifest.json: VibeVoice-ASR-BitNet.litertlm's sha256, size and sections after the move to external weights; min runtime 0.19.0 and the Android GPU known issue (LiteRT-LM#3986)

Browse files
Files changed (1) hide show
  1. litertlm_manifest.json +19 -7
litertlm_manifest.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "manifest_schema": "0.1.3",
3
  "repo": "litert-community/VibeVoice-ASR-BitNet",
4
- "generated": "2026-10-07",
5
  "generator": "make_manifest.py",
6
  "model": {
7
  "display_name": "VibeVoice-ASR-BitNet (speech-to-text LLM, 24 kHz audio in)",
@@ -22,8 +22,8 @@
22
  "variants": [
23
  {
24
  "file": "VibeVoice-ASR-BitNet.litertlm",
25
- "sha256": "5ca907b0343d3e6bd9ec3dbf8aecbcc99b733633ef007a78a7e0f3502010af1b",
26
- "size_bytes": 1983019248,
27
  "sections": [
28
  {
29
  "type": "LlmMetadataProto",
@@ -35,18 +35,28 @@
35
  },
36
  {
37
  "type": "TFLiteModel",
38
- "size_bytes": 237024176,
39
  "model_type": "tf_lite_embedder"
40
  },
41
  {
42
  "type": "TFLiteModel",
43
- "size_bytes": 792689088,
44
  "model_type": "tf_lite_prefill_decode"
45
  },
46
  {
47
  "type": "TFLiteModel",
48
  "size_bytes": 951122160,
49
  "model_type": "tf_lite_audio_encoder_hw"
 
 
 
 
 
 
 
 
 
 
50
  }
51
  ],
52
  "quantization": "LM: per-tensor ternary weights (VibeASR.cpp absmean, applied at conversion) stored as int4 blockwise-128 min-max (exact for a ternary tensor), int8 embedding; audio encoder: int8 dynamic on linears, fp32 convs, 30 s (225-frame) window",
@@ -55,7 +65,7 @@
55
  "gpu"
56
  ],
57
  "default_backend": "cpu",
58
- "min_runtime_version": "0.16.1",
59
  "recommended": [
60
  {
61
  "platform": "macos",
@@ -140,9 +150,11 @@
140
  }
141
  ],
142
  "known_issues": [
 
143
  "Audio encoder does not run on the GPU delegates: on Metal/WebGPU the 30 s window dispatches 90001+ workgroups per stem kernel (limit 65535); on Adreno (OpenCL, all 1857 nodes delegated) the fp16 path returns an empty transcript; keep audio_backend=cpu",
144
  "Audio placeholder tokens <|object_ref_start|>/<|object_ref_end|>/<|box_start|> are added to the bundled tokenizer.json (ids 151646/151647/151648, the ids the model was trained with); the upstream tokenizer.json has no string for them"
145
- ]
 
146
  }
147
  ]
148
  }
 
1
  {
2
  "manifest_schema": "0.1.3",
3
  "repo": "litert-community/VibeVoice-ASR-BitNet",
4
+ "generated": "2026-11-09",
5
  "generator": "make_manifest.py",
6
  "model": {
7
  "display_name": "VibeVoice-ASR-BitNet (speech-to-text LLM, 24 kHz audio in)",
 
22
  "variants": [
23
  {
24
  "file": "VibeVoice-ASR-BitNet.litertlm",
25
+ "sha256": "498ad85c609fb3821989db5f3767a07e38b82d5554bf26f73e091d939d9c1b0b",
26
+ "size_bytes": 1983053824,
27
  "sections": [
28
  {
29
  "type": "LlmMetadataProto",
 
35
  },
36
  {
37
  "type": "TFLiteModel",
38
+ "size_bytes": 3650664,
39
  "model_type": "tf_lite_embedder"
40
  },
41
  {
42
  "type": "TFLiteModel",
43
+ "size_bytes": 20937360,
44
  "model_type": "tf_lite_prefill_decode"
45
  },
46
  {
47
  "type": "TFLiteModel",
48
  "size_bytes": 951122160,
49
  "model_type": "tf_lite_audio_encoder_hw"
50
+ },
51
+ {
52
+ "type": "TFLiteWeights",
53
+ "size_bytes": 233373696,
54
+ "model_type": "tf_lite_embedder"
55
+ },
56
+ {
57
+ "type": "TFLiteWeights",
58
+ "size_bytes": 771784704,
59
+ "model_type": "tf_lite_prefill_decode"
60
  }
61
  ],
62
  "quantization": "LM: per-tensor ternary weights (VibeASR.cpp absmean, applied at conversion) stored as int4 blockwise-128 min-max (exact for a ternary tensor), int8 embedding; audio encoder: int8 dynamic on linears, fp32 convs, 30 s (225-frame) window",
 
65
  "gpu"
66
  ],
67
  "default_backend": "cpu",
68
+ "min_runtime_version": "0.19.0",
69
  "recommended": [
70
  {
71
  "platform": "macos",
 
150
  }
151
  ],
152
  "known_issues": [
153
+ "Android GPU needs LiteRT-LM 0.19.0 or later: since 2026-11-09 this file's weights sit in TFLiteWeights sections at the end of the file (LiteRT's external weights), and the GPU engine of the v0.18.0 and v0.18.1 Android builds crashes (SIGSEGV) on such a file (https://github.com/google-ai-edge/LiteRT-LM/issues/3986); the earlier file with the weights inside the graph stays at revision 193b2d579a786206286ac36e74b6ea513562bdf5 (https://huggingface.co/litert-community/VibeVoice-ASR-BitNet/tree/193b2d579a786206286ac36e74b6ea513562bdf5)",
154
  "Audio encoder does not run on the GPU delegates: on Metal/WebGPU the 30 s window dispatches 90001+ workgroups per stem kernel (limit 65535); on Adreno (OpenCL, all 1857 nodes delegated) the fp16 path returns an empty transcript; keep audio_backend=cpu",
155
  "Audio placeholder tokens <|object_ref_start|>/<|object_ref_end|>/<|box_start|> are added to the bundled tokenizer.json (ids 151646/151647/151648, the ids the model was trained with); the upstream tokenizer.json has no string for them"
156
+ ],
157
+ "context_length": 2048
158
  }
159
  ]
160
  }