Hugging Face's logo

qualcomm
/
Stable-Diffusion-v1.5

Unconditional Image Generation
PyTorch
generative_ai
android
Model card Files
xet
Community
qaihm-bot commited on
Commit
037d5ab
·
verified ·
1 Parent(s): 07af287

v0.43.0

Browse files

See https://github.com/quic/ai-hub-models/releases/v0.43.0 for changelog.

Files changed (20) hide show
  1. README.md +21 -18
  2. precompiled/qualcomm-qcs8550-proxy/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip +2 -2
  3. precompiled/qualcomm-qcs8550-proxy/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip +2 -2
  4. precompiled/qualcomm-qcs8550-proxy/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip +2 -2
  5. precompiled/qualcomm-snapdragon-7gen4/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip +3 -0
  6. precompiled/qualcomm-snapdragon-7gen4/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip +3 -0
  7. precompiled/qualcomm-snapdragon-7gen4/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip +3 -0
  8. precompiled/qualcomm-snapdragon-7gen4/tool-versions.yaml +4 -0
  9. precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip +2 -2
  10. precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip +1 -1
  11. precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip +2 -2
  12. precompiled/qualcomm-snapdragon-8-elite-gen5/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip +2 -2
  13. precompiled/qualcomm-snapdragon-8-elite-gen5/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip +2 -2
  14. precompiled/qualcomm-snapdragon-8-elite-gen5/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip +2 -2
  15. precompiled/qualcomm-snapdragon-8gen3/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip +2 -2
  16. precompiled/qualcomm-snapdragon-8gen3/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip +1 -1
  17. precompiled/qualcomm-snapdragon-8gen3/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip +2 -2
  18. precompiled/qualcomm-snapdragon-x-elite/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip +1 -1
  19. precompiled/qualcomm-snapdragon-x-elite/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip +2 -2
  20. precompiled/qualcomm-snapdragon-x-elite/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip +1 -1
README.md CHANGED
@@ -33,21 +33,24 @@ More details on model performance across various devices, can be found
33
 
34
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
35
  |---|---|---|---|---|---|---|---|---|
36
- | text_encoder | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 5.467 ms | 0 - 162 MB | NPU | Use Export Script |
37
- | text_encoder | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 3.961 ms | 0 - 19 MB | NPU | Use Export Script |
38
- | text_encoder | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 3.102 ms | 0 - 15 MB | NPU | Use Export Script |
39
- | text_encoder | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | PRECOMPILED_QNN_ONNX | 2.623 ms | 0 - 10 MB | NPU | Use Export Script |
40
- | text_encoder | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 5.667 ms | 157 - 157 MB | NPU | Use Export Script |
41
- | unet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 112.467 ms | 0 - 899 MB | NPU | Use Export Script |
42
- | unet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 80.436 ms | 0 - 17 MB | NPU | Use Export Script |
43
- | unet | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 63.49 ms | 0 - 15 MB | NPU | Use Export Script |
44
- | unet | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | PRECOMPILED_QNN_ONNX | 47.113 ms | 0 - 8 MB | NPU | Use Export Script |
45
- | unet | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 113.47 ms | 842 - 842 MB | NPU | Use Export Script |
46
- | vae | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 220.115 ms | 3 - 6 MB | NPU | Use Export Script |
47
- | vae | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 162.862 ms | 3 - 22 MB | NPU | Use Export Script |
48
- | vae | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 146.475 ms | 3 - 18 MB | NPU | Use Export Script |
49
- | vae | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | PRECOMPILED_QNN_ONNX | 94.369 ms | 3 - 13 MB | NPU | Use Export Script |
50
- | vae | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 218.188 ms | 59 - 59 MB | NPU | Use Export Script |
 
 
 
51
 
52
  ## Deploy to Snapdragon X Elite NPU
53
  Please follow the [Stable Diffusion Windows App](https://github.com/quic/ai-hub-apps/tree/main/apps/windows/python/StableDiffusion) tutorial to quantize model with custom weights.
@@ -68,9 +71,9 @@ pip install "qai-hub-models[stable-diffusion-v1-5]"
68
  ```
69
 
70
 
71
- ## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
72
 
73
- Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
74
  Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
75
 
76
  With this API token, you can configure your client to run models on the cloud
@@ -78,7 +81,7 @@ hosted devices.
78
  ```bash
79
  qai-hub configure --api_token API_TOKEN
80
  ```
81
- Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information.
82
 
83
 
84
 
 
33
 
34
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
35
  |---|---|---|---|---|---|---|---|---|
36
+ | text_encoder | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 5.484 ms | 0 - 162 MB | NPU | Use Export Script |
37
+ | text_encoder | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 3.945 ms | 0 - 22 MB | NPU | Use Export Script |
38
+ | text_encoder | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 3.106 ms | 0 - 11 MB | NPU | Use Export Script |
39
+ | text_encoder | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | PRECOMPILED_QNN_ONNX | 5.757 ms | 0 - 14 MB | NPU | Use Export Script |
40
+ | text_encoder | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | PRECOMPILED_QNN_ONNX | 2.619 ms | 0 - 10 MB | NPU | Use Export Script |
41
+ | text_encoder | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 5.646 ms | 157 - 157 MB | NPU | Use Export Script |
42
+ | unet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 112.731 ms | 0 - 899 MB | NPU | Use Export Script |
43
+ | unet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 79.969 ms | 0 - 16 MB | NPU | Use Export Script |
44
+ | unet | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 63.819 ms | 0 - 21 MB | NPU | Use Export Script |
45
+ | unet | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | PRECOMPILED_QNN_ONNX | 172.669 ms | 0 - 10 MB | NPU | Use Export Script |
46
+ | unet | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | PRECOMPILED_QNN_ONNX | 46.846 ms | 0 - 7 MB | NPU | Use Export Script |
47
+ | unet | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 113.219 ms | 842 - 842 MB | NPU | Use Export Script |
48
+ | vae | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 219.968 ms | 3 - 6 MB | NPU | Use Export Script |
49
+ | vae | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 162.551 ms | 3 - 22 MB | NPU | Use Export Script |
50
+ | vae | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 147.035 ms | 3 - 14 MB | NPU | Use Export Script |
51
+ | vae | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | PRECOMPILED_QNN_ONNX | 445.273 ms | 3 - 17 MB | NPU | Use Export Script |
52
+ | vae | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | PRECOMPILED_QNN_ONNX | 89.9 ms | 3 - 13 MB | NPU | Use Export Script |
53
+ | vae | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 218.025 ms | 59 - 59 MB | NPU | Use Export Script |
54
 
55
  ## Deploy to Snapdragon X Elite NPU
56
  Please follow the [Stable Diffusion Windows App](https://github.com/quic/ai-hub-apps/tree/main/apps/windows/python/StableDiffusion) tutorial to quantize model with custom weights.
 
71
  ```
72
 
73
 
74
+ ## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
75
 
76
+ Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
77
  Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
78
 
79
  With this API token, you can configure your client to run models on the cloud
 
81
  ```bash
82
  qai-hub configure --api_token API_TOKEN
83
  ```
84
+ Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
85
 
86
 
87
 
precompiled/qualcomm-qcs8550-proxy/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1d45ef3f13f0119cd375e7d132f68274e87642f4c329d8209e6db3989a80daf4
3
- size 127298819
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ecbae405ddce459427fb1c39a61baf9f8816a7a57edbe9ea3309579b3504d998
3
+ size 127298844
precompiled/qualcomm-qcs8550-proxy/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:92d64b0d8ce80b6909febb03fb47db482f4c1063924b9b2c1124d3511b814e07
3
- size 567379636
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e734a6b066276fc63462d8687e388179cc7061357ce08a861541df5fa688e12
3
+ size 567379568
precompiled/qualcomm-qcs8550-proxy/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:61d43c7fcdabf06acd08421bbaf4d61d723264165b5eeaccb5925d625c4f5a3e
3
- size 40400245
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b070a836d582d03eb58935dce0d20f0e225e0ed6f093d42be18068833718da7
3
+ size 40400248
precompiled/qualcomm-snapdragon-7gen4/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1459199143e572dca58f4f2f68e00e4e4a9b9f4e6a465e4d63e9217865cf9b33
3
+ size 127216124
precompiled/qualcomm-snapdragon-7gen4/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fde214648e7b6487560eba143ce02b796c44e7087c1a53d6a4441b5fbb65849
3
+ size 567388832
precompiled/qualcomm-snapdragon-7gen4/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf5d3b8a4973dc12b9494dc3c08f71a16a26a9ed6e5a7034b4ebee5110c22c2c
3
+ size 42032497
precompiled/qualcomm-snapdragon-7gen4/tool-versions.yaml ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ tool_versions:
2
+ precompiled_qnn_onnx:
3
+ qairt: 2.37.1.250807093845_124904
4
+ onnx_runtime: 1.23.0
precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:54b2c9808dfbe94f74db0ab30114fe63b5cee5ea93f5309aa584a3e3858d8719
3
- size 127338643
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c84e6775f9056819ddd812fcae270ced0d49318b56385c715c2bc51fa57354dc
3
+ size 127338603
precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0fa9539678ba97fc99a109355ac0129cd002623c1d0329ddc7c9c02431618dd2
3
  size 567324228
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e4ef2fcb0925e98978c17dbc7cfef522ee5492dc8dbf56601a2c06de8faaa76
3
  size 567324228
precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:87f252d83eb9b023e6bb825883c822f810114017a05a9e48968f7792028056db
3
- size 40185429
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f75476a41359dd6fd9c8b4a0e3ecfb92e9fe0ce5b11f78cc61b25a2a41afea77
3
+ size 40185425
precompiled/qualcomm-snapdragon-8-elite-gen5/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8ccb5a4a319de9b7c0ecda1195c55bf71b6dcf3b405acff1102d945b289bd6f9
3
- size 127368221
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d437bd113a5b56bc09c9e5d19996ac8a46fbaeca1463e43e4efc714050c4240
3
+ size 127368190
precompiled/qualcomm-snapdragon-8-elite-gen5/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7df9e76353d49c43d22ba6ebe8a86b7d6a6d5e21bc9ab24ad13bb8dd2c4bc7f0
3
- size 567255038
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b7651ac224efd34f718cb3f42b1881e693b3d0ca5f5f3c0cf7a50b733df9c3c
3
+ size 567255009
precompiled/qualcomm-snapdragon-8-elite-gen5/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b56daed515ca0923e74e53347803140f369e986e208ce6009c8a8f511190dacc
3
- size 40465857
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9a731beb45b05478e6b72750cdf5cecac0526773c8810eb0203ec8bcd3cfbeb
3
+ size 40465862
precompiled/qualcomm-snapdragon-8gen3/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2b03bf7a07b469fe55a9595921faa5bb5a4570f896e02c787e622e057fc9eba2
3
- size 127293437
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2c73ecc3bdc9b11c69420ebf46731794ba9fe9391f11b4681e8b2243e475813
3
+ size 127293451
precompiled/qualcomm-snapdragon-8gen3/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3edb35895f52478c1df6cfb0732f480ef85a03335c499ebd116771a9a63bc048
3
  size 567424733
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e8f0bbacb22f5ce0a6f37fdfb46efbf778ae1f8371abb48d5a62077ecc970a04
3
  size 567424733
precompiled/qualcomm-snapdragon-8gen3/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dc8ff9835e867c6d2e6c6de3510498acbe80800024c7c39215581a04190ffe5c
3
- size 40296296
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3380edcd0ef303631d019df519f75653a6e3e1601915c3bad2a85aa5faf05731
3
+ size 40296297
precompiled/qualcomm-snapdragon-x-elite/Stable-Diffusion-v1.5_text_encoder_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2e596035c4e98a5dcd1cb8ba118b5a7b0dfb965b4f5e059a7503a9d5ee026b02
3
  size 127300657
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:598ed7c6d3217bebb81a278fe29e15b1242331fd57de9e91f6a85cdc686fbfb3
3
  size 127300657
precompiled/qualcomm-snapdragon-x-elite/Stable-Diffusion-v1.5_unet_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ce130a4170a63e5b2d21c07973d91c177fdfbb57aef2337205e9ced1cf5a0c03
3
- size 566661987
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34cc8a3819538a11a0e2d6740d35df658604a576e9d3f4a56398539675c062b2
3
+ size 566661986
precompiled/qualcomm-snapdragon-x-elite/Stable-Diffusion-v1.5_vae_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ed3d41490487a95909bec091424369c18dc0108e2d920f4284253e9210e8c19b
3
  size 40415225
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7abb6582fbe5998bcd9b01650013889b060014b17d1790901dea4c3cc98f0ef7
3
  size 40415225