qaihm-bot commited on
Commit
0a81a69
·
verified ·
1 Parent(s): 8309cf4

See https://github.com/quic/ai-hub-models/releases/v0.29.1 for changelog.

README.md CHANGED
@@ -38,64 +38,64 @@ More details on model performance across various devices, can be found
38
 
39
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
40
  |---|---|---|---|---|---|---|---|---|
41
- | WhisperEncoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 419.271 ms | 20 - 42 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
42
- | WhisperEncoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 142.28 ms | 0 - 10 MB | NPU | Use Export Script |
43
- | WhisperEncoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 134.033 ms | 21 - 64 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
44
- | WhisperEncoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 114.257 ms | 0 - 236 MB | NPU | Use Export Script |
45
- | WhisperEncoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 130.771 ms | 20 - 69 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
46
- | WhisperEncoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 68.697 ms | 1 - 3 MB | NPU | Use Export Script |
47
- | WhisperEncoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 168.846 ms | 20 - 45 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
48
- | WhisperEncoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 72.128 ms | 1 - 15 MB | NPU | Use Export Script |
49
- | WhisperEncoderInf | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 419.271 ms | 20 - 42 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
50
- | WhisperEncoderInf | float | SA7255P ADP | Qualcomm® SA7255P | QNN | 142.28 ms | 0 - 10 MB | NPU | Use Export Script |
51
- | WhisperEncoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 101.236 ms | 3 - 37 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
52
- | WhisperEncoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 69.056 ms | 1 - 10 MB | NPU | Use Export Script |
53
- | WhisperEncoderInf | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 113.839 ms | 20 - 49 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
54
- | WhisperEncoderInf | float | SA8295P ADP | Qualcomm® SA8295P | QNN | 93.516 ms | 1 - 19 MB | NPU | Use Export Script |
55
- | WhisperEncoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 98.63 ms | 20 - 69 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
56
- | WhisperEncoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 72.489 ms | 1 - 3 MB | NPU | Use Export Script |
57
- | WhisperEncoderInf | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 168.846 ms | 20 - 45 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
58
- | WhisperEncoderInf | float | SA8775P ADP | Qualcomm® SA8775P | QNN | 72.128 ms | 1 - 15 MB | NPU | Use Export Script |
59
- | WhisperEncoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 114.249 ms | 20 - 68 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
60
- | WhisperEncoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 69.251 ms | 0 - 83 MB | NPU | Use Export Script |
61
- | WhisperEncoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 65.119 ms | 26 - 168 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
62
- | WhisperEncoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 78.166 ms | 20 - 54 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
63
- | WhisperEncoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 57.663 ms | 1 - 239 MB | NPU | Use Export Script |
64
- | WhisperEncoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 52.51 ms | 37 - 366 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
65
- | WhisperEncoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 76.181 ms | 19 - 49 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
66
- | WhisperEncoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 45.212 ms | 0 - 228 MB | NPU | Use Export Script |
67
- | WhisperEncoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 41.854 ms | 41 - 349 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
68
- | WhisperEncoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 66.685 ms | 0 - 0 MB | NPU | Use Export Script |
69
- | WhisperEncoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 66.088 ms | 67 - 67 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
70
- | WhisperDecoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 6.898 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
71
- | WhisperDecoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 4.013 ms | 9 - 19 MB | NPU | Use Export Script |
72
- | WhisperDecoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 4.863 ms | 3 - 76 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
73
- | WhisperDecoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 3.066 ms | 4 - 53 MB | NPU | Use Export Script |
74
- | WhisperDecoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 3.846 ms | 3 - 27 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
75
- | WhisperDecoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 2.214 ms | 10 - 14 MB | NPU | Use Export Script |
76
- | WhisperDecoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 4.723 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
77
- | WhisperDecoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 2.867 ms | 9 - 20 MB | NPU | Use Export Script |
78
- | WhisperDecoderInf | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 6.898 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
79
- | WhisperDecoderInf | float | SA7255P ADP | Qualcomm® SA7255P | QNN | 4.013 ms | 9 - 19 MB | NPU | Use Export Script |
80
- | WhisperDecoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 3.782 ms | 3 - 38 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
81
- | WhisperDecoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 2.313 ms | 10 - 13 MB | NPU | Use Export Script |
82
- | WhisperDecoderInf | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 4.94 ms | 3 - 65 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
83
- | WhisperDecoderInf | float | SA8295P ADP | Qualcomm® SA8295P | QNN | 3.062 ms | 1 - 18 MB | NPU | Use Export Script |
84
- | WhisperDecoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 3.813 ms | 3 - 29 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
85
- | WhisperDecoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 2.138 ms | 10 - 13 MB | NPU | Use Export Script |
86
- | WhisperDecoderInf | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 4.723 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
87
- | WhisperDecoderInf | float | SA8775P ADP | Qualcomm® SA8775P | QNN | 2.867 ms | 9 - 20 MB | NPU | Use Export Script |
88
- | WhisperDecoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 3.782 ms | 3 - 28 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
89
- | WhisperDecoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 2.149 ms | 4 - 24 MB | NPU | Use Export Script |
90
- | WhisperDecoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 4.083 ms | 6 - 241 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
91
- | WhisperDecoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 2.942 ms | 3 - 79 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
92
- | WhisperDecoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 1.658 ms | 4 - 57 MB | NPU | Use Export Script |
93
- | WhisperDecoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.029 ms | 26 - 102 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
94
- | WhisperDecoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 2.542 ms | 1 - 72 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
95
- | WhisperDecoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 1.329 ms | 4 - 53 MB | NPU | Use Export Script |
96
- | WhisperDecoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 2.991 ms | 0 - 69 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
97
- | WhisperDecoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.087 ms | 10 - 10 MB | NPU | Use Export Script |
98
- | WhisperDecoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 3.642 ms | 74 - 74 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
99
 
100
 
101
 
@@ -159,8 +159,8 @@ Profiling Results
159
  WhisperEncoderInf
160
  Device : cs_8275 (ANDROID 14)
161
  Runtime : TFLITE
162
- Estimated inference time (ms) : 419.3
163
- Estimated peak memory usage (MB): [20, 42]
164
  Total # Ops : 271
165
  Compute Unit(s) : npu (0 ops) gpu (260 ops) cpu (11 ops)
166
 
 
38
 
39
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
40
  |---|---|---|---|---|---|---|---|---|
41
+ | WhisperEncoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 402.002 ms | 19 - 41 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
42
+ | WhisperEncoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 143.211 ms | 1 - 10 MB | NPU | Use Export Script |
43
+ | WhisperEncoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 144.102 ms | 20 - 61 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
44
+ | WhisperEncoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 122.102 ms | 0 - 230 MB | NPU | Use Export Script |
45
+ | WhisperEncoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 104.553 ms | 3 - 48 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
46
+ | WhisperEncoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 69.346 ms | 1 - 3 MB | NPU | Use Export Script |
47
+ | WhisperEncoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 168.883 ms | 20 - 45 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
48
+ | WhisperEncoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 72.23 ms | 1 - 15 MB | NPU | Use Export Script |
49
+ | WhisperEncoderInf | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 402.002 ms | 19 - 41 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
50
+ | WhisperEncoderInf | float | SA7255P ADP | Qualcomm® SA7255P | QNN | 143.211 ms | 1 - 10 MB | NPU | Use Export Script |
51
+ | WhisperEncoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 99.525 ms | 20 - 68 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
52
+ | WhisperEncoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 69.397 ms | 1 - 3 MB | NPU | Use Export Script |
53
+ | WhisperEncoderInf | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 112.96 ms | 20 - 49 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
54
+ | WhisperEncoderInf | float | SA8295P ADP | Qualcomm® SA8295P | QNN | 93.669 ms | 1 - 19 MB | NPU | Use Export Script |
55
+ | WhisperEncoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 104.169 ms | 20 - 69 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
56
+ | WhisperEncoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 70.891 ms | 0 - 3 MB | NPU | Use Export Script |
57
+ | WhisperEncoderInf | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 168.883 ms | 20 - 45 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
58
+ | WhisperEncoderInf | float | SA8775P ADP | Qualcomm® SA8775P | QNN | 72.23 ms | 1 - 15 MB | NPU | Use Export Script |
59
+ | WhisperEncoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 99.369 ms | 20 - 69 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
60
+ | WhisperEncoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 68.549 ms | 0 - 85 MB | NPU | Use Export Script |
61
+ | WhisperEncoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 62.856 ms | 5 - 144 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
62
+ | WhisperEncoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 85.694 ms | 18 - 55 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
63
+ | WhisperEncoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 58.571 ms | 1 - 240 MB | NPU | Use Export Script |
64
+ | WhisperEncoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 52.535 ms | 30 - 360 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
65
+ | WhisperEncoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 75.915 ms | 20 - 46 MB | GPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
66
+ | WhisperEncoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 44.055 ms | 0 - 228 MB | NPU | Use Export Script |
67
+ | WhisperEncoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 44.895 ms | 37 - 345 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
68
+ | WhisperEncoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 67.455 ms | 0 - 0 MB | NPU | Use Export Script |
69
+ | WhisperEncoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 67.634 ms | 67 - 67 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
70
+ | WhisperDecoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 6.883 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
71
+ | WhisperDecoderInf | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 4.019 ms | 4 - 14 MB | NPU | Use Export Script |
72
+ | WhisperDecoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 4.817 ms | 3 - 74 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
73
+ | WhisperDecoderInf | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 3.193 ms | 4 - 54 MB | NPU | Use Export Script |
74
+ | WhisperDecoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 3.807 ms | 3 - 37 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
75
+ | WhisperDecoderInf | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 2.202 ms | 9 - 11 MB | NPU | Use Export Script |
76
+ | WhisperDecoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 4.758 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
77
+ | WhisperDecoderInf | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 2.856 ms | 8 - 20 MB | NPU | Use Export Script |
78
+ | WhisperDecoderInf | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 6.883 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
79
+ | WhisperDecoderInf | float | SA7255P ADP | Qualcomm® SA7255P | QNN | 4.019 ms | 4 - 14 MB | NPU | Use Export Script |
80
+ | WhisperDecoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 3.898 ms | 3 - 40 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
81
+ | WhisperDecoderInf | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 2.125 ms | 16 - 18 MB | NPU | Use Export Script |
82
+ | WhisperDecoderInf | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 4.83 ms | 3 - 65 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
83
+ | WhisperDecoderInf | float | SA8295P ADP | Qualcomm® SA8295P | QNN | 3.082 ms | 2 - 20 MB | NPU | Use Export Script |
84
+ | WhisperDecoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 3.791 ms | 3 - 26 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
85
+ | WhisperDecoderInf | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 2.249 ms | 5 - 7 MB | NPU | Use Export Script |
86
+ | WhisperDecoderInf | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 4.758 ms | 3 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
87
+ | WhisperDecoderInf | float | SA8775P ADP | Qualcomm® SA8775P | QNN | 2.856 ms | 8 - 20 MB | NPU | Use Export Script |
88
+ | WhisperDecoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 3.792 ms | 3 - 38 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
89
+ | WhisperDecoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 2.204 ms | 4 - 23 MB | NPU | Use Export Script |
90
+ | WhisperDecoderInf | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 3.999 ms | 6 - 184 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
91
+ | WhisperDecoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 2.97 ms | 2 - 75 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
92
+ | WhisperDecoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 1.707 ms | 4 - 57 MB | NPU | Use Export Script |
93
+ | WhisperDecoderInf | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.103 ms | 26 - 102 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
94
+ | WhisperDecoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 2.68 ms | 0 - 70 MB | NPU | [Whisper-Tiny-En.tflite](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.tflite) |
95
+ | WhisperDecoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 1.359 ms | 0 - 48 MB | NPU | Use Export Script |
96
+ | WhisperDecoderInf | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 2.898 ms | 3 - 72 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
97
+ | WhisperDecoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.121 ms | 10 - 10 MB | NPU | Use Export Script |
98
+ | WhisperDecoderInf | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 3.53 ms | 74 - 74 MB | NPU | [Whisper-Tiny-En.onnx](https://huggingface.co/qualcomm/Whisper-Tiny-En/blob/main/Whisper-Tiny-En.onnx) |
99
 
100
 
101
 
 
159
  WhisperEncoderInf
160
  Device : cs_8275 (ANDROID 14)
161
  Runtime : TFLITE
162
+ Estimated inference time (ms) : 402.0
163
+ Estimated peak memory usage (MB): [19, 41]
164
  Total # Ops : 271
165
  Compute Unit(s) : npu (0 ops) gpu (260 ops) cpu (11 ops)
166
 
WhisperDecoder.bin → Whisper-Tiny-En_WhisperDecoderInf.onnx RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ccd9b08481eb4629b707c7629f80042bfd2f293625adaa9a5ee5f9c9e3dbbdfa
3
- size 97452888
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:755ea46c4db83b684f9220ce9ae1c7fbc4408fd61f9dce6e260d846444b9aa18
3
+ size 193091330
WhisperDecoder.onnx → Whisper-Tiny-En_WhisperDecoderInf.tflite RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ad25b689053b731e79da01fa6a64addeaf7cad1f89220cf604b13fcf6402758c
3
- size 193103192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f2e94746b483b48cb02424ad34d83e949c1f5e1cd900fcd24052eca240514e7
3
+ size 113437748
WhisperEncoderInf.onnx → Whisper-Tiny-En_WhisperEncoderInf.onnx RENAMED
File without changes
WhisperEncoder.tflite → Whisper-Tiny-En_WhisperEncoderInf.tflite RENAMED
File without changes
WhisperDecoder.so DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:4cdb56f747c36c24e2232c3881e4763e87b655972f133d641a6122d43eccaf6d
3
- size 97598480
 
 
 
 
WhisperDecoder.tflite DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:4113b4bfd4261f080c903fdeb33192334fe6d0fa8c3dd3af4c425b548fa6b4a0
3
- size 113438084
 
 
 
 
WhisperDecoderInf.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:5ca932bc13933d3992bea873b31928c670e855a700070aaaee0b907135764355
3
- size 97445632
 
 
 
 
WhisperDecoderInf.onnx DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:67348f533353143db2c222b8280991585b7052da755142cfd58e0458b6bf9a10
3
- size 193090488
 
 
 
 
WhisperDecoderInf.so DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1a35092f9a5b2284ae56e80d6c86437cb197d125519f45b73f718874022465ac
3
- size 97430472
 
 
 
 
WhisperDecoderInf.tflite DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:8f07d10ee791d55f719355d5a2a99870ef13928634b4abc9f97eee2735de874d
3
- size 113437964
 
 
 
 
WhisperEncoder.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:58683d7e76552e280d4b6672063d0096e7fe57c12f365cbcc2bbb81ca8792d27
3
- size 32284240
 
 
 
 
WhisperEncoder.onnx DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:75084c29db415f9c76652144a877ca3de9d99dcbbaefa4c081608a02832fe855
3
- size 37604846
 
 
 
 
WhisperEncoder.so DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1871ce785ff765ca38e7a74eca0830fe8c2a3e5bafb79ce537e377b7364b528b
3
- size 19692592
 
 
 
 
WhisperEncoderInf.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ed6a21dd54e34a31e5dc586675c8f8d2e72f2f6270316c0bc28c8d31076d182c
3
- size 32285360
 
 
 
 
WhisperEncoderInf.so DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:a34c83fa3baa639bbd7c841131a79c4795c13dd52bbf7d096dd72f2270094867
3
- size 19623584
 
 
 
 
WhisperEncoderInf.tflite DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:802d5e08a6ab7e8794f9826c939239d5063012e6144a931be5a65bdf392b21d7
3
- size 37612748