Hugging Face's logo

qualcomm
/
DeepSpeech2

Automatic Speech Recognition
PyTorch
foundation
real_time
android
Model card Files
xet
 Community
qaihm-bot commited on
Commit
212d7de
·
verified ·
1 Parent(s): d26bb77

v0.61.0

Browse files

See https://github.com/qualcomm/ai-hub-models/releases/v0.61.0 for changelog.

Files changed (2) hide show
  1. README.md +20 -20
  2. release_assets.json +2 -2
README.md CHANGED
@@ -1,6 +1,6 @@
1
  ---
2
  library_name: pytorch
3
- license: other
4
  tags:
5
  - foundation
6
  - real_time
@@ -16,7 +16,7 @@ pipeline_tag: automatic-speech-recognition
16
  DeepSpeech2 is an end-to-end automatic speech recognition (ASR) model. It uses convolutional layers for feature extraction followed by bidirectional recurrent layers and CTC for sequence-to-sequence learning.
17
 
18
  This is based on the implementation of DeepSpeech2 found [here](https://github.com/SeanNaren/deepspeech.pytorch).
19
- 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/v0.60.0/src/qai_hub_models/models/deepspeech2) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
20
 
21
  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.
22
 
@@ -29,21 +29,21 @@ Below are pre-exported model assets ready for deployment.
29
 
30
  | Runtime | Precision | Chipset | SDK Versions | Download |
31
  |---|---|---|---|---|
32
- | TFLITE | float | Universal | | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deepspeech2/releases/v0.60.0/deepspeech2-tflite-float.zip)
33
 
34
  For more device-specific assets and performance metrics, visit **[DeepSpeech2 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/deepspeech2)**.
35
 
36
 
37
  ### Option 2: Export with Custom Configurations
38
 
39
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.60.0/src/qai_hub_models/models/deepspeech2) Python library to compile and export the model with your own:
40
  - Custom weights (e.g., fine-tuned checkpoints)
41
  - Custom input shapes
42
  - Target device and runtime configurations
43
 
44
  This option is ideal if you need to customize the model beyond the default configuration provided here.
45
 
46
- See our repository for [DeepSpeech2 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.60.0/src/qai_hub_models/models/deepspeech2) for usage instructions.
47
 
48
  ## Model Details
49
 
@@ -57,21 +57,21 @@ See our repository for [DeepSpeech2 on GitHub](https://github.com/qualcomm/ai-hu
57
  ## Performance Summary
58
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
59
  |---|---|---|---|---|---|---
60
- | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 4387.282 ms | 147 - 160 MB | CPU
61
- | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 5872.379 ms | 205 - 220 MB | CPU
62
- | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 12161.223 ms | 67 - 551 MB | CPU
63
- | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 18819.817 ms | 109 - 118 MB | CPU
64
- | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 4150.669 ms | 49 - 62 MB | CPU
65
- | DeepSpeech2 | TFLITE | float | Qualcomm® SA8775P | 14169.392 ms | 107 - 117 MB | CPU
66
- | DeepSpeech2 | TFLITE | float | Qualcomm® SA8650P | 14169.392 ms | 107 - 117 MB | CPU
67
- | DeepSpeech2 | TFLITE | float | Qualcomm® SA8255P | 14169.392 ms | 107 - 117 MB | CPU
68
- | DeepSpeech2 | TFLITE | float | Qualcomm® QCS8450 | 5872.379 ms | 205 - 220 MB | CPU
69
- | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 5239.743 ms | 66 - 550 MB | CPU
70
- | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 3156.583 ms | 30 - 43 MB | CPU
71
- | DeepSpeech2 | TFLITE | float | Qualcomm® SA7255P | 18819.817 ms | 109 - 118 MB | CPU
72
- | DeepSpeech2 | TFLITE | float | Qualcomm® SA8295P | 9438.743 ms | 97 - 107 MB | CPU
73
- | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Elite Mobile | 3156.583 ms | 30 - 43 MB | CPU
74
- | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3371.983 ms | 108 - 121 MB | CPU
75
 
76
  ## License
77
  * The license for the original implementation of DeepSpeech2 can be found
 
1
  ---
2
  library_name: pytorch
3
+ license: apache-2.0
4
  tags:
5
  - foundation
6
  - real_time
 
16
  DeepSpeech2 is an end-to-end automatic speech recognition (ASR) model. It uses convolutional layers for feature extraction followed by bidirectional recurrent layers and CTC for sequence-to-sequence learning.
17
 
18
  This is based on the implementation of DeepSpeech2 found [here](https://github.com/SeanNaren/deepspeech.pytorch).
19
+ 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/v0.61.0/src/qai_hub_models/models/deepspeech2) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
20
 
21
  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.
22
 
 
29
 
30
  | Runtime | Precision | Chipset | SDK Versions | Download |
31
  |---|---|---|---|---|
32
+ | TFLITE | float | Universal | | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deepspeech2/releases/v0.61.0/deepspeech2-tflite-float.zip)
33
 
34
  For more device-specific assets and performance metrics, visit **[DeepSpeech2 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/deepspeech2)**.
35
 
36
 
37
  ### Option 2: Export with Custom Configurations
38
 
39
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.61.0/src/qai_hub_models/models/deepspeech2) Python library to compile and export the model with your own:
40
  - Custom weights (e.g., fine-tuned checkpoints)
41
  - Custom input shapes
42
  - Target device and runtime configurations
43
 
44
  This option is ideal if you need to customize the model beyond the default configuration provided here.
45
 
46
+ See our repository for [DeepSpeech2 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.61.0/src/qai_hub_models/models/deepspeech2) for usage instructions.
47
 
48
  ## Model Details
49
 
 
57
  ## Performance Summary
58
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
59
  |---|---|---|---|---|---|---
60
+ | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 4052.36 ms | 188 - 199 MB | CPU
61
+ | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 5914.991 ms | 199 - 213 MB | CPU
62
+ | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 8474.079 ms | 72 - 556 MB | CPU
63
+ | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 18172.649 ms | 113 - 123 MB | CPU
64
+ | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 4083.295 ms | 62 - 69 MB | CPU
65
+ | DeepSpeech2 | TFLITE | float | Qualcomm® SA8775P | 14213.705 ms | 109 - 118 MB | CPU
66
+ | DeepSpeech2 | TFLITE | float | Qualcomm® SA8650P | 14213.705 ms | 109 - 118 MB | CPU
67
+ | DeepSpeech2 | TFLITE | float | Qualcomm® SA8255P | 14213.705 ms | 109 - 118 MB | CPU
68
+ | DeepSpeech2 | TFLITE | float | Qualcomm® QCS8450 | 5914.991 ms | 199 - 213 MB | CPU
69
+ | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 5277.193 ms | 67 - 551 MB | CPU
70
+ | DeepSpeech2 | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 3079.184 ms | 0 - 8 MB | CPU
71
+ | DeepSpeech2 | TFLITE | float | Qualcomm® SA7255P | 18172.649 ms | 113 - 123 MB | CPU
72
+ | DeepSpeech2 | TFLITE | float | Qualcomm® SA8295P | 9452.569 ms | 113 - 123 MB | CPU
73
+ | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Elite Mobile | 3079.184 ms | 0 - 8 MB | CPU
74
+ | DeepSpeech2 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3231.078 ms | 110 - 124 MB | CPU
75
 
76
  ## License
77
  * The license for the original implementation of DeepSpeech2 can be found
release_assets.json CHANGED
@@ -1,11 +1,11 @@
1
  {
2
- "version": "0.60.0",
3
  "precisions": {
4
  "float": {
5
  "universal_assets": {
6
  "tflite": {
7
  "tool_versions": {},
8
- "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deepspeech2/releases/v0.60.0/deepspeech2-tflite-float.zip"
9
  }
10
  }
11
  }
 
1
  {
2
+ "version": "0.61.0",
3
  "precisions": {
4
  "float": {
5
  "universal_assets": {
6
  "tflite": {
7
  "tool_versions": {},
8
+ "download_url": "https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deepspeech2/releases/v0.61.0/deepspeech2-tflite-float.zip"
9
  }
10
  }
11
  }