qaihm-bot commited on
Commit
f664f4f
·
verified ·
1 Parent(s): 9f79afe

See https://github.com/qualcomm/ai-hub-models/releases/v0.48.0 for changelog.

Files changed (1) hide show
  1. README.md +136 -138
README.md CHANGED
@@ -15,7 +15,7 @@ pipeline_tag: automatic-speech-recognition
15
  Zipformer streaming ASR (Automatic Speech Recognition) model is a state-of-the-art system designed for transcribing spoken language into written text streamingly. This model is based on the transformer architecture and has been optimized for edge inference by replacing linear layers with convolutional (conv) layers. It exhibits robust performance in realistic, noisy environments, making it highly reliable for real-world applications. Specifically, it excels in long-form transcription, capable of accurately transcribing audios. Time to the first token is the encoder's latency, while time to each additional token is joiner's latency, where we assume a max decoded length specified below.
16
 
17
  This is based on the implementation of Zipformer found [here](https://github.com/k2-fsa/icefall).
18
- This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/zipformer) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
 
20
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
 
@@ -28,55 +28,53 @@ Below are pre-exported model assets ready for deployment.
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
- | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_x_elite.zip)
32
- | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_8gen3.zip)
33
- | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_qcs8550_proxy.zip)
34
- | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
35
- | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_gen5.zip)
36
- | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_qcs9075.zip)
37
- | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_x2_elite.zip)
38
- | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_snapdragon_x_elite.zip)
39
- | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_snapdragon_8gen3.zip)
40
- | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS8550 (Proxy) | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_qcs8550_proxy.zip)
41
- | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_snapdragon_8_elite_for_galaxy.zip)
42
- | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_snapdragon_8_elite_gen5.zip)
43
- | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X2 Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_snapdragon_x2_elite.zip)
44
- | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS9075 | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-precompiled_qnn_onnx-w16a16-qualcomm_qcs9075.zip)
45
- | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_x_elite.zip)
46
- | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_8gen3.zip)
47
- | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_qcs8275_proxy.zip)
48
- | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_qcs8550_proxy.zip)
49
- | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_sa8775p.zip)
50
- | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
51
- | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_8_elite_gen5.zip)
52
- | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_sa7255p.zip)
53
- | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_sa8295p.zip)
54
- | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_qcs9075.zip)
55
- | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-float-qualcomm_qcs8450_proxy.zip)
56
- | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X Elite | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_snapdragon_x_elite.zip)
57
- | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_snapdragon_8gen3.zip)
58
- | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8275 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_qcs8275_proxy.zip)
59
- | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8550 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_qcs8550_proxy.zip)
60
- | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA8775P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_sa8775p.zip)
61
- | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_snapdragon_8_elite_for_galaxy.zip)
62
- | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_snapdragon_8_elite_gen5.zip)
63
- | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X2 Elite | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_snapdragon_x2_elite.zip)
64
- | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA7255P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_sa7255p.zip)
65
- | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS9075 | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.47.0/zipformer-qnn_context_binary-w16a16-qualcomm_qcs9075.zip)
66
 
67
  For more device-specific assets and performance metrics, visit **[Zipformer on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/zipformer)**.
68
 
69
 
70
  ### Option 2: Export with Custom Configurations
71
 
72
- Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/zipformer) Python library to compile and export the model with your own:
73
  - Custom weights (e.g., fine-tuned checkpoints)
74
  - Custom input shapes
75
  - Target device and runtime configurations
76
 
77
  This option is ideal if you need to customize the model beyond the default configuration provided here.
78
 
79
- See our repository for [Zipformer on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/zipformer) for usage instructions.
80
 
81
  ## Model Details
82
 
@@ -96,111 +94,111 @@ See our repository for [Zipformer on GitHub](https://github.com/quic/ai-hub-mode
96
  ## Performance Summary
97
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
98
  |---|---|---|---|---|---|---
99
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 0.16 ms | 7 - 7 MB | NPU
100
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.059 ms | 0 - 7 MB | NPU
101
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.067 ms | 0 - 8 MB | NPU
102
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 0.137 ms | 0 - 3 MB | NPU
103
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.051 ms | 0 - 7 MB | NPU
104
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.053 ms | 0 - 10 MB | NPU
105
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 0.042 ms | 7 - 7 MB | NPU
106
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X Elite | 0.161 ms | 7 - 7 MB | NPU
107
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Gen 3 Mobile | 0.067 ms | 0 - 6 MB | NPU
108
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS8550 (Proxy) | 0.078 ms | 0 - 1 MB | NPU
109
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS9075 | 0.19 ms | 0 - 3 MB | NPU
110
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.05 ms | 0 - 7 MB | NPU
111
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.05 ms | 0 - 10 MB | NPU
112
- | ZipformerDecoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X2 Elite | 0.039 ms | 7 - 7 MB | NPU
113
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 0.151 ms | 0 - 0 MB | NPU
114
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 0.063 ms | 0 - 7 MB | NPU
115
  | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 0.319 ms | 0 - 8 MB | NPU
116
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 0.077 ms | 0 - 1 MB | NPU
117
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 0.208 ms | 0 - 5 MB | NPU
118
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 0.12 ms | 2 - 4 MB | NPU
119
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | 0.102 ms | 0 - 9 MB | NPU
120
  | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 0.319 ms | 0 - 8 MB | NPU
121
  | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 0.33 ms | 0 - 5 MB | NPU
122
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.047 ms | 0 - 13 MB | NPU
123
- | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.041 ms | 0 - 10 MB | NPU
124
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X Elite | 0.171 ms | 0 - 0 MB | NPU
125
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Gen 3 Mobile | 0.056 ms | 0 - 7 MB | NPU
126
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8275 (Proxy) | 0.322 ms | 0 - 8 MB | NPU
127
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8550 (Proxy) | 0.063 ms | 0 - 1 MB | NPU
128
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA8775P | 0.215 ms | 0 - 5 MB | NPU
129
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS9075 | 0.134 ms | 0 - 2 MB | NPU
130
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA7255P | 0.322 ms | 0 - 8 MB | NPU
131
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.046 ms | 0 - 9 MB | NPU
132
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.053 ms | 0 - 10 MB | NPU
133
- | ZipformerDecoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X2 Elite | 0.144 ms | 0 - 0 MB | NPU
134
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 9.771 ms | 142 - 142 MB | NPU
135
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 6.505 ms | 9 - 17 MB | NPU
136
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 9.062 ms | 9 - 10 MB | NPU
137
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 10.363 ms | 8 - 12 MB | NPU
138
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 5.384 ms | 9 - 17 MB | NPU
139
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 4.878 ms | 10 - 20 MB | NPU
140
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 5.488 ms | 142 - 142 MB | NPU
141
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X Elite | 14.96 ms | 146 - 146 MB | NPU
142
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Gen 3 Mobile | 12.414 ms | 4 - 15 MB | NPU
143
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS8550 (Proxy) | 16.212 ms | 4 - 6 MB | NPU
144
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS9075 | 18.591 ms | 4 - 7 MB | NPU
145
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | 8.621 ms | 4 - 11 MB | NPU
146
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | 8.046 ms | 4 - 14 MB | NPU
147
- | ZipformerEncoder | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X2 Elite | 9.253 ms | 147 - 147 MB | NPU
148
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 9.729 ms | 3 - 3 MB | NPU
149
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 6.6 ms | 2 - 10 MB | NPU
150
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 23.718 ms | 3 - 11 MB | NPU
151
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 9.066 ms | 3 - 5 MB | NPU
152
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 10.992 ms | 3 - 11 MB | NPU
153
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 10.214 ms | 3 - 7 MB | NPU
154
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | 11.845 ms | 2 - 12 MB | NPU
155
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 23.718 ms | 3 - 11 MB | NPU
156
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 13.295 ms | 3 - 8 MB | NPU
157
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 5.433 ms | 3 - 15 MB | NPU
158
- | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 4.738 ms | 3 - 12 MB | NPU
159
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X Elite | 14.608 ms | 1 - 1 MB | NPU
160
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Gen 3 Mobile | 11.568 ms | 1 - 8 MB | NPU
161
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8275 (Proxy) | 27.668 ms | 1 - 8 MB | NPU
162
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8550 (Proxy) | 15.779 ms | 1 - 5 MB | NPU
163
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA8775P | 17.364 ms | 1 - 10 MB | NPU
164
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS9075 | 17.302 ms | 1 - 3 MB | NPU
165
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA7255P | 27.668 ms | 1 - 8 MB | NPU
166
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | 8.509 ms | 1 - 9 MB | NPU
167
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | 7.924 ms | 1 - 11 MB | NPU
168
- | ZipformerEncoder | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X2 Elite | 9.366 ms | 1 - 1 MB | NPU
169
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 0.262 ms | 6 - 6 MB | NPU
170
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.149 ms | 0 - 6 MB | NPU
171
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.19 ms | 0 - 8 MB | NPU
172
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 0.299 ms | 0 - 3 MB | NPU
173
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.134 ms | 0 - 8 MB | NPU
 
174
  | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.132 ms | 0 - 10 MB | NPU
175
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 0.079 ms | 6 - 6 MB | NPU
176
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X Elite | 0.25 ms | 6 - 6 MB | NPU
177
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Gen 3 Mobile | 0.157 ms | 0 - 7 MB | NPU
178
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS8550 (Proxy) | 0.187 ms | 0 - 8 MB | NPU
179
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Qualcomm® QCS9075 | 0.253 ms | 0 - 3 MB | NPU
180
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.132 ms | 0 - 7 MB | NPU
181
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.135 ms | 0 - 10 MB | NPU
182
- | ZipformerJoiner | PRECOMPILED_QNN_ONNX | w16a16 | Snapdragon® X2 Elite | 0.083 ms | 6 - 6 MB | NPU
183
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 0.246 ms | 0 - 0 MB | NPU
184
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 0.159 ms | 0 - 7 MB | NPU
185
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 0.472 ms | 0 - 7 MB | NPU
186
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 0.187 ms | 0 - 1 MB | NPU
187
  | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 0.324 ms | 0 - 5 MB | NPU
188
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 0.235 ms | 0 - 2 MB | NPU
189
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | 0.231 ms | 0 - 9 MB | NPU
190
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 0.472 ms | 0 - 7 MB | NPU
191
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 0.469 ms | 0 - 5 MB | NPU
192
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.133 ms | 0 - 13 MB | NPU
193
- | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.133 ms | 0 - 10 MB | NPU
194
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X Elite | 0.253 ms | 0 - 0 MB | NPU
195
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Gen 3 Mobile | 0.151 ms | 0 - 7 MB | NPU
196
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8275 (Proxy) | 0.438 ms | 0 - 9 MB | NPU
197
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS8550 (Proxy) | 0.183 ms | 0 - 1 MB | NPU
198
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA8775P | 0.883 ms | 0 - 8 MB | NPU
199
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® QCS9075 | 0.233 ms | 2 - 4 MB | NPU
200
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Qualcomm® SA7255P | 0.438 ms | 0 - 9 MB | NPU
201
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.135 ms | 0 - 9 MB | NPU
202
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.135 ms | 0 - 10 MB | NPU
203
- | ZipformerJoiner | QNN_CONTEXT_BINARY | w16a16 | Snapdragon® X2 Elite | 0.176 ms | 0 - 0 MB | NPU
204
 
205
  ## License
206
  * The license for the original implementation of Zipformer can be found
 
15
  Zipformer streaming ASR (Automatic Speech Recognition) model is a state-of-the-art system designed for transcribing spoken language into written text streamingly. This model is based on the transformer architecture and has been optimized for edge inference by replacing linear layers with convolutional (conv) layers. It exhibits robust performance in realistic, noisy environments, making it highly reliable for real-world applications. Specifically, it excels in long-form transcription, capable of accurately transcribing audios. Time to the first token is the encoder's latency, while time to each additional token is joiner's latency, where we assume a max decoded length specified below.
16
 
17
  This is based on the implementation of Zipformer found [here](https://github.com/k2-fsa/icefall).
18
+ This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/zipformer) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
 
20
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
 
 
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
+ | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_x2_elite.zip)
32
+ | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_x_elite.zip)
33
+ | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_8gen3.zip)
34
+ | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_qcs8550_proxy.zip)
35
+ | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
36
+ | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_gen5.zip)
37
+ | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-float-qualcomm_qcs9075.zip)
38
+ | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X2 Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_snapdragon_x2_elite.zip)
39
+ | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X Elite | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_snapdragon_x_elite.zip)
40
+ | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_snapdragon_8gen3.zip)
41
+ | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS8550 (Proxy) | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_qcs8550_proxy.zip)
42
+ | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_snapdragon_8_elite_for_galaxy.zip)
43
+ | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_snapdragon_8_elite_gen5.zip)
44
+ | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS9075 | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-precompiled_qnn_onnx-mixed-qualcomm_qcs9075.zip)
45
+ | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_x_elite.zip)
46
+ | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_8gen3.zip)
47
+ | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_qcs8550_proxy.zip)
48
+ | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_sa8775p.zip)
49
+ | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
50
+ | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_snapdragon_8_elite_gen5.zip)
51
+ | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_sa7255p.zip)
52
+ | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_sa8295p.zip)
53
+ | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_qcs9075.zip)
54
+ | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-float-qualcomm_qcs8450_proxy.zip)
55
+ | QNN_CONTEXT_BINARY | mixed | Snapdragon® X2 Elite | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_snapdragon_x2_elite.zip)
56
+ | QNN_CONTEXT_BINARY | mixed | Snapdragon® X Elite | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_snapdragon_x_elite.zip)
57
+ | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_snapdragon_8gen3.zip)
58
+ | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8550 (Proxy) | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_qcs8550_proxy.zip)
59
+ | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA8775P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_sa8775p.zip)
60
+ | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_snapdragon_8_elite_for_galaxy.zip)
61
+ | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_snapdragon_8_elite_gen5.zip)
62
+ | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA7255P | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_sa7255p.zip)
63
+ | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS9075 | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/releases/v0.48.0/zipformer-qnn_context_binary-mixed-qualcomm_qcs9075.zip)
 
 
64
 
65
  For more device-specific assets and performance metrics, visit **[Zipformer on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/zipformer)**.
66
 
67
 
68
  ### Option 2: Export with Custom Configurations
69
 
70
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/zipformer) Python library to compile and export the model with your own:
71
  - Custom weights (e.g., fine-tuned checkpoints)
72
  - Custom input shapes
73
  - Target device and runtime configurations
74
 
75
  This option is ideal if you need to customize the model beyond the default configuration provided here.
76
 
77
+ See our repository for [Zipformer on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/zipformer) for usage instructions.
78
 
79
  ## Model Details
80
 
 
94
  ## Performance Summary
95
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
96
  |---|---|---|---|---|---|---
97
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 0.036 ms | 7 - 7 MB | NPU
98
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 0.165 ms | 7 - 7 MB | NPU
99
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.06 ms | 0 - 7 MB | NPU
100
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.071 ms | 0 - 8 MB | NPU
101
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 0.149 ms | 0 - 3 MB | NPU
102
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.051 ms | 0 - 8 MB | NPU
103
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.051 ms | 0 - 10 MB | NPU
104
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X2 Elite | 0.053 ms | 7 - 7 MB | NPU
105
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X Elite | 0.155 ms | 7 - 7 MB | NPU
106
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Gen 3 Mobile | 0.066 ms | 0 - 7 MB | NPU
107
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS8550 (Proxy) | 0.076 ms | 0 - 8 MB | NPU
108
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS9075 | 0.155 ms | 0 - 3 MB | NPU
109
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite For Galaxy Mobile | 0.054 ms | 0 - 11 MB | NPU
110
+ | ZipformerDecoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite Gen 5 Mobile | 0.048 ms | 0 - 10 MB | NPU
111
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 0.155 ms | 0 - 0 MB | NPU
112
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 0.055 ms | 0 - 7 MB | NPU
113
  | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 0.319 ms | 0 - 8 MB | NPU
114
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 0.071 ms | 0 - 1 MB | NPU
115
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 0.588 ms | 0 - 8 MB | NPU
116
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 0.125 ms | 2 - 4 MB | NPU
117
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | 0.099 ms | 0 - 9 MB | NPU
118
  | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 0.319 ms | 0 - 8 MB | NPU
119
  | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 0.33 ms | 0 - 5 MB | NPU
120
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.043 ms | 0 - 9 MB | NPU
121
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.047 ms | 0 - 10 MB | NPU
122
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® X2 Elite | 0.13 ms | 0 - 0 MB | NPU
123
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® X Elite | 0.173 ms | 0 - 0 MB | NPU
124
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Gen 3 Mobile | 0.053 ms | 0 - 7 MB | NPU
125
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8275 (Proxy) | 0.314 ms | 0 - 8 MB | NPU
126
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8550 (Proxy) | 0.075 ms | 0 - 1 MB | NPU
127
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA8775P | 0.222 ms | 0 - 9 MB | NPU
128
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS9075 | 0.132 ms | 2 - 4 MB | NPU
129
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA7255P | 0.314 ms | 0 - 8 MB | NPU
130
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite For Galaxy Mobile | 0.047 ms | 0 - 13 MB | NPU
131
+ | ZipformerDecoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite Gen 5 Mobile | 0.042 ms | 0 - 10 MB | NPU
132
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 5.469 ms | 142 - 142 MB | NPU
133
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 9.747 ms | 142 - 142 MB | NPU
134
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 6.746 ms | 9 - 17 MB | NPU
135
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 9.096 ms | 0 - 159 MB | NPU
136
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 10.377 ms | 8 - 12 MB | NPU
137
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 5.463 ms | 9 - 17 MB | NPU
138
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 4.858 ms | 10 - 20 MB | NPU
139
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X2 Elite | 8.256 ms | 75 - 75 MB | NPU
140
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X Elite | 13.714 ms | 75 - 75 MB | NPU
141
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Gen 3 Mobile | 10.805 ms | 9 - 17 MB | NPU
142
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS8550 (Proxy) | 14.699 ms | 9 - 10 MB | NPU
143
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS9075 | 16.848 ms | 8 - 12 MB | NPU
144
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite For Galaxy Mobile | 7.568 ms | 9 - 21 MB | NPU
145
+ | ZipformerEncoder | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite Gen 5 Mobile | 7.196 ms | 10 - 20 MB | NPU
146
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 9.632 ms | 3 - 3 MB | NPU
147
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 6.616 ms | 2 - 10 MB | NPU
148
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 23.756 ms | 3 - 10 MB | NPU
149
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 8.982 ms | 3 - 4 MB | NPU
150
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 10.941 ms | 3 - 11 MB | NPU
151
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 10.192 ms | 3 - 7 MB | NPU
152
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | 11.941 ms | 2 - 11 MB | NPU
153
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 23.756 ms | 3 - 10 MB | NPU
154
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 13.226 ms | 3 - 8 MB | NPU
155
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 5.498 ms | 3 - 11 MB | NPU
156
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 4.809 ms | 3 - 13 MB | NPU
157
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® X2 Elite | 8.739 ms | 3 - 3 MB | NPU
158
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® X Elite | 13.254 ms | 3 - 3 MB | NPU
159
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Gen 3 Mobile | 10.769 ms | 2 - 11 MB | NPU
160
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8275 (Proxy) | 25.125 ms | 3 - 11 MB | NPU
161
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8550 (Proxy) | 14.648 ms | 3 - 4 MB | NPU
162
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA8775P | 16.191 ms | 3 - 12 MB | NPU
163
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS9075 | 16.109 ms | 3 - 7 MB | NPU
164
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA7255P | 25.125 ms | 3 - 11 MB | NPU
165
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite For Galaxy Mobile | 7.286 ms | 3 - 12 MB | NPU
166
+ | ZipformerEncoder | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite Gen 5 Mobile | 7.011 ms | 3 - 14 MB | NPU
167
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 0.117 ms | 6 - 6 MB | NPU
168
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 0.252 ms | 6 - 6 MB | NPU
169
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.15 ms | 0 - 7 MB | NPU
170
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.194 ms | 0 - 8 MB | NPU
171
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 0.258 ms | 0 - 3 MB | NPU
172
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.137 ms | 0 - 7 MB | NPU
173
  | ZipformerJoiner | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.132 ms | 0 - 10 MB | NPU
174
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X2 Elite | 0.085 ms | 6 - 6 MB | NPU
175
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® X Elite | 0.278 ms | 6 - 6 MB | NPU
176
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Gen 3 Mobile | 0.152 ms | 0 - 7 MB | NPU
177
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS8550 (Proxy) | 0.199 ms | 0 - 8 MB | NPU
178
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Qualcomm® QCS9075 | 0.272 ms | 0 - 3 MB | NPU
179
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite For Galaxy Mobile | 0.137 ms | 0 - 8 MB | NPU
180
+ | ZipformerJoiner | PRECOMPILED_QNN_ONNX | mixed | Snapdragon® 8 Elite Gen 5 Mobile | 0.133 ms | 0 - 10 MB | NPU
181
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 0.251 ms | 0 - 0 MB | NPU
182
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 0.152 ms | 0 - 7 MB | NPU
183
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 0.479 ms | 0 - 8 MB | NPU
184
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 0.191 ms | 0 - 2 MB | NPU
 
185
  | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 0.324 ms | 0 - 5 MB | NPU
186
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 0.234 ms | 0 - 2 MB | NPU
187
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 (Proxy) | 0.232 ms | 0 - 9 MB | NPU
188
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 0.479 ms | 0 - 8 MB | NPU
189
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 0.452 ms | 0 - 5 MB | NPU
190
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.131 ms | 0 - 9 MB | NPU
191
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.133 ms | 0 - 9 MB | NPU
192
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Snapdragon® X2 Elite | 0.203 ms | 0 - 0 MB | NPU
193
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Snapdragon® X Elite | 0.269 ms | 0 - 0 MB | NPU
194
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Gen 3 Mobile | 0.163 ms | 0 - 7 MB | NPU
195
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8275 (Proxy) | 0.454 ms | 0 - 9 MB | NPU
196
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS8550 (Proxy) | 0.191 ms | 0 - 1 MB | NPU
197
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA8775P | 0.333 ms | 0 - 5 MB | NPU
198
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Qualcomm® QCS9075 | 0.255 ms | 0 - 2 MB | NPU
199
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Qualcomm® SA7255P | 0.454 ms | 0 - 9 MB | NPU
200
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite For Galaxy Mobile | 0.143 ms | 0 - 9 MB | NPU
201
+ | ZipformerJoiner | QNN_CONTEXT_BINARY | mixed | Snapdragon® 8 Elite Gen 5 Mobile | 0.137 ms | 0 - 10 MB | NPU
202
 
203
  ## License
204
  * The license for the original implementation of Zipformer can be found