v0.45.0

See https://github.com/quic/ai-hub-models/releases/v0.45.0 for changelog.

LICENSE

@@ -0,0 +1 @@
+The license of the original trained model can be found at https://github.com/huggingface/transformers/blob/v4.42.3/LICENSE.

README.md

@@ -0,0 +1,268 @@
+---
+library_name: pytorch
+license: other
+tags:
+- foundation
+- android
+pipeline_tag: automatic-speech-recognition
+
+---
+
+![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/zipformer/web-assets/model_demo.png)
+
+# Zipformer: Optimized for Mobile Deployment
+## Transformer-based automatic speech recognition (ASR) model for English and Chinese language
+
+
+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.
+
+This model is an implementation of Zipformer found [here](https://github.com/k2-fsa/icefall).
+
+
+This repository provides scripts to run Zipformer on Qualcomm® devices.
+More details on model performance across various devices, can be found
+[here](https://aihub.qualcomm.com/models/zipformer).
+
+
+
+### Model Details
+
+- **Model Type:** Model_use_case.speech_recognition
+- **Model Stats:**
+  - Model checkpoint: pfluo/k2fsa-zipformer-chinese-english-mixed
+  - Input resolution: 80x71 (0.71 seconds audio)
+  - Max decoded sequence length: 200 tokens
+  - Number of parameters (ZipformerEncoder): 63.2M
+  - Model size (ZipformerEncoder) (float): 242 MB
+  - Number of parameters (ZipformerDecoder): 3.47M
+  - Model size (ZipformerDecoder) (float): 13.2 MB
+  - Number of parameters (ZipformerJoiner): 3.21M
+  - Model size (ZipformerJoiner) (float): 12.2 MB
+
+| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
+|---|---|---|---|---|---|---|---|---|
+| ZipformerEncoder | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_CONTEXT_BINARY | 23.393 ms | 3 - 10 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_CONTEXT_BINARY | 11.889 ms | 2 - 12 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_CONTEXT_BINARY | 8.84 ms | 3 - 5 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 9.665 ms | 10 - 11 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_CONTEXT_BINARY | 41.908 ms | 3 - 11 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | SA7255P ADP | Qualcomm® SA7255P | QNN_CONTEXT_BINARY | 23.393 ms | 3 - 10 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | SA8295P ADP | Qualcomm® SA8295P | QNN_CONTEXT_BINARY | 13.206 ms | 0 - 5 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | SA8775P ADP | Qualcomm® SA8775P | QNN_CONTEXT_BINARY | 41.908 ms | 3 - 11 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_CONTEXT_BINARY | 6.294 ms | 2 - 11 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 7.014 ms | 9 - 17 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_CONTEXT_BINARY | 5.417 ms | 3 - 15 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 5.879 ms | 9 - 21 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_CONTEXT_BINARY | 4.835 ms | 3 - 13 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | PRECOMPILED_QNN_ONNX | 5.367 ms | 10 - 20 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_CONTEXT_BINARY | 9.385 ms | 3 - 3 MB | NPU | Use Export Script |
+| ZipformerEncoder | float | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 9.807 ms | 142 - 142 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_CONTEXT_BINARY | 0.302 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_CONTEXT_BINARY | 0.098 ms | 0 - 9 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_CONTEXT_BINARY | 0.073 ms | 0 - 2 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 0.211 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_CONTEXT_BINARY | 0.583 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | SA7255P ADP | Qualcomm® SA7255P | QNN_CONTEXT_BINARY | 0.302 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | SA8295P ADP | Qualcomm® SA8295P | QNN_CONTEXT_BINARY | 0.319 ms | 0 - 5 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | SA8775P ADP | Qualcomm® SA8775P | QNN_CONTEXT_BINARY | 0.583 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_CONTEXT_BINARY | 0.055 ms | 0 - 7 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 0.2 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_CONTEXT_BINARY | 0.05 ms | 0 - 9 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 0.157 ms | 0 - 12 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_CONTEXT_BINARY | 0.045 ms | 0 - 10 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | PRECOMPILED_QNN_ONNX | 0.155 ms | 0 - 11 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_CONTEXT_BINARY | 0.151 ms | 0 - 0 MB | NPU | Use Export Script |
+| ZipformerDecoder | float | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 0.154 ms | 7 - 7 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_CONTEXT_BINARY | 0.465 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_CONTEXT_BINARY | 0.231 ms | 0 - 9 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_CONTEXT_BINARY | 0.189 ms | 0 - 1 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | PRECOMPILED_QNN_ONNX | 0.352 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_CONTEXT_BINARY | 0.328 ms | 0 - 5 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | SA7255P ADP | Qualcomm® SA7255P | QNN_CONTEXT_BINARY | 0.465 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | SA8295P ADP | Qualcomm® SA8295P | QNN_CONTEXT_BINARY | 0.454 ms | 0 - 5 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | SA8775P ADP | Qualcomm® SA8775P | QNN_CONTEXT_BINARY | 0.328 ms | 0 - 5 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_CONTEXT_BINARY | 0.154 ms | 0 - 8 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | PRECOMPILED_QNN_ONNX | 0.328 ms | 0 - 7 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_CONTEXT_BINARY | 0.132 ms | 0 - 9 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | PRECOMPILED_QNN_ONNX | 0.23 ms | 0 - 7 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_CONTEXT_BINARY | 0.13 ms | 0 - 9 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | PRECOMPILED_QNN_ONNX | 0.226 ms | 0 - 11 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_CONTEXT_BINARY | 0.263 ms | 0 - 0 MB | NPU | Use Export Script |
+| ZipformerJoiner | float | Snapdragon X Elite CRD | Snapdragon® X Elite | PRECOMPILED_QNN_ONNX | 0.247 ms | 6 - 6 MB | NPU | Use Export Script |
+
+
+
+
+## Installation
+
+
+Install the package via pip:
+```bash
+# NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported.
+pip install torch==2.9.0+cpu -f https://download.pytorch.org/whl/torch/
+pip install "qai-hub-models[zipformer]" k2==1.24.4.dev20251029+cpu.torch2.9.0 -f https://k2-fsa.github.io/k2/cpu.html
+```
+
+
+## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
+
+Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
+Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
+
+With this API token, you can configure your client to run models on the cloud
+hosted devices.
+```bash
+qai-hub configure --api_token API_TOKEN
+```
+Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
+
+
+
+## Demo off target
+
+The package contains a simple end-to-end demo that downloads pre-trained
+weights and runs this model on a sample input.
+
+```bash
+python -m qai_hub_models.models.zipformer.demo
+```
+
+The above demo runs a reference implementation of pre-processing, model
+inference, and post processing.
+
+**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
+environment, please add the following to your cell (instead of the above).
+```
+%run -m qai_hub_models.models.zipformer.demo
+```
+
+
+### Run model on a cloud-hosted device
+
+In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
+device. This script does the following:
+* Performance check on-device on a cloud-hosted device
+* Downloads compiled assets that can be deployed on-device for Android.
+* Accuracy check between PyTorch and on-device outputs.
+
+```bash
+python -m qai_hub_models.models.zipformer.export
+```
+
+
+
+## How does this work?
+
+This [export script](https://aihub.qualcomm.com/models/zipformer/qai_hub_models/models/Zipformer/export.py)
+leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
+on-device. Lets go through each step below in detail:
+
+Step 1: **Compile model for on-device deployment**
+
+To compile a PyTorch model for on-device deployment, we first trace the model
+in memory using the `jit.trace` and then call the `submit_compile_job` API.
+
+```python
+import torch
+
+import qai_hub as hub
+from qai_hub_models.models.zipformer import Model
+
+# Load the model
+torch_model = Model.from_pretrained()
+
+# Device
+device = hub.Device("Samsung Galaxy S25")
+
+# Trace model
+input_shape = torch_model.get_input_spec()
+sample_inputs = torch_model.sample_inputs()
+
+pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
+
+# Compile model on a specific device
+compile_job = hub.submit_compile_job(
+    model=pt_model,
+    device=device,
+    input_specs=torch_model.get_input_spec(),
+)
+
+# Get target model to run on-device
+target_model = compile_job.get_target_model()
+
+```
+
+
+Step 2: **Performance profiling on cloud-hosted device**
+
+After compiling models from step 1. Models can be profiled model on-device using the
+`target_model`. Note that this scripts runs the model on a device automatically
+provisioned in the cloud.  Once the job is submitted, you can navigate to a
+provided job URL to view a variety of on-device performance metrics.
+```python
+profile_job = hub.submit_profile_job(
+    model=target_model,
+    device=device,
+)
+        
+```
+
+Step 3: **Verify on-device accuracy**
+
+To verify the accuracy of the model on-device, you can run on-device inference
+on sample input data on the same cloud hosted device.
+```python
+input_data = torch_model.sample_inputs()
+inference_job = hub.submit_inference_job(
+    model=target_model,
+    device=device,
+    inputs=input_data,
+)
+    on_device_output = inference_job.download_output_data()
+
+```
+With the output of the model, you can compute like PSNR, relative errors or
+spot check the output with expected output.
+
+**Note**: This on-device profiling and inference requires access to Qualcomm®
+AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
+
+
+
+
+## Deploying compiled model to Android
+
+
+The models can be deployed using multiple runtimes:
+- TensorFlow Lite (`.tflite` export): [This
+  tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
+  guide to deploy the .tflite model in an Android application.
+
+
+- QNN (`.so` export ): This [sample
+  app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
+provides instructions on how to use the `.so` shared library  in an Android application.
+
+
+## View on Qualcomm® AI Hub
+Get more details on Zipformer's performance across various devices [here](https://aihub.qualcomm.com/models/zipformer).
+Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
+
+
+## License
+* The license for the original implementation of Zipformer can be found
+  [here](https://github.com/huggingface/transformers/blob/v4.42.3/LICENSE).
+
+
+
+## References
+* [Zipformer A faster and better encoder for automatic speech recognition](https://openreview.net/forum?id=9WD9KwssyT)
+* [Source Model Implementation](https://github.com/k2-fsa/icefall)
+
+
+
+## Community
+* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
+* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).
+
+

precompiled/qualcomm-qcs8275-proxy/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5363413e257201699a5a517e5c05cfd9a5e60ce7af085ae72c52cb8a8f5a768d
+size 7045120

precompiled/qualcomm-qcs8275-proxy/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e7ce06d26dcbafd44e26f3521aa1ccbc0ea69ae44c5802a3ec348813d6c3a9a
+size 151408640

precompiled/qualcomm-qcs8275-proxy/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a0c7f430dfad84afe210c6ed4546bc06b5f9c3e96ee27079e33d34c19f4db42
+size 6529024

precompiled/qualcomm-qcs8275-proxy/tool-versions.yaml

@@ -0,0 +1,3 @@
+tool_versions:
+  qnn_context_binary:
+    qairt: 2.41.0.251128145156_191518-auto

precompiled/qualcomm-qcs8450-proxy/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e715304ca5d1ce417fe3ebf09bdc87536f7d588191e338a55bcc58b427706680
+size 7045120

precompiled/qualcomm-qcs8450-proxy/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f3e02f37dca952c7fe1d2c19c6952aa1b05443e761c559ecd58ed9fd856cb21
+size 150818816

precompiled/qualcomm-qcs8450-proxy/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9b7d9f199a0388cb3f9c06e842a593f54d40b78bcb96eec683faf997101e55e
+size 6500352

precompiled/qualcomm-qcs8450-proxy/tool-versions.yaml

@@ -0,0 +1,3 @@
+tool_versions:
+  qnn_context_binary:
+    qairt: 2.41.0.251128145156_191518

precompiled/qualcomm-qcs8550-proxy/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:375e2e17df6468f302d102605bf96e4d2dafb88dda8c0326b1b658566937ae3d
+size 7045120

precompiled/qualcomm-qcs8550-proxy/Zipformer_ZipformerDecoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77ac8aa72661e27eeec75623e9d79e0050f16a38b6b8bc55e0264a7fb7adcbe4
+size 6501936

precompiled/qualcomm-qcs8550-proxy/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13ed7d514daec2f6d6ba279ca05c665ab78c8174e9297a8c874009085c70e709
+size 151408640

precompiled/qualcomm-qcs8550-proxy/Zipformer_ZipformerEncoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e2f45c2daa00177397145225c002b4683bdcb2f9c78c7f13dfbbded51bd33d3
+size 118423850

precompiled/qualcomm-qcs8550-proxy/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:123494094260a4e75518df48c335cdd5c405e0e9499090ed6584075d6442a040
+size 6529024

precompiled/qualcomm-qcs8550-proxy/Zipformer_ZipformerJoiner_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e16e47f35a6969b8c6202deadea239c261370822e59b3c56f4201d2c934d6e46
+size 5776036

precompiled/qualcomm-qcs8550-proxy/tool-versions.yaml

@@ -0,0 +1,4 @@
+tool_versions:
+  precompiled_qnn_onnx:
+    qairt: 2.37.1.250807093845_124904
+    onnx_runtime: 1.23.0

precompiled/qualcomm-qcs9075-proxy/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f8b0bc321754d064fd41e62589532b29288f6cbd095ba64787e66b8cfc2e9e1
+size 7045120

precompiled/qualcomm-qcs9075-proxy/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7d689cea00cb723269f4513baf6f3202fb5f58154d32026d227d5fe56447304
+size 151408640

precompiled/qualcomm-qcs9075-proxy/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bcdbd48d58b024ffd9bc27b8c04314d9e10de4dbb48ffbd0bd0b7fa430ca19d
+size 6529024

precompiled/qualcomm-qcs9075-proxy/tool-versions.yaml

@@ -0,0 +1,3 @@
+tool_versions:
+  qnn_context_binary:
+    qairt: 2.41.0.251128145156_191518-auto

precompiled/qualcomm-sa7255p/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5363413e257201699a5a517e5c05cfd9a5e60ce7af085ae72c52cb8a8f5a768d
+size 7045120

precompiled/qualcomm-sa7255p/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e7ce06d26dcbafd44e26f3521aa1ccbc0ea69ae44c5802a3ec348813d6c3a9a
+size 151408640

precompiled/qualcomm-sa7255p/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a0c7f430dfad84afe210c6ed4546bc06b5f9c3e96ee27079e33d34c19f4db42
+size 6529024

precompiled/qualcomm-sa7255p/tool-versions.yaml

@@ -0,0 +1,3 @@
+tool_versions:
+  qnn_context_binary:
+    qairt: 2.41.0.251128145156_191518-auto

precompiled/qualcomm-sa8295p/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cf069b6d9b4a5e233d6eda478d32ff5aba8f552a15fe92caca379b0721c45e2
+size 7045120

precompiled/qualcomm-sa8295p/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:011384fd6798c84ac99c6894fb4d8cb94337a85bd4523833925a11676b982892
+size 150794240

precompiled/qualcomm-sa8295p/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35210c56260ee29b365a13b0e1b68cf0d3bb04e98d888cc7c4778730aada4d15
+size 6500352

precompiled/qualcomm-sa8295p/tool-versions.yaml

@@ -0,0 +1,3 @@
+tool_versions:
+  qnn_context_binary:
+    qairt: 2.41.0.251128145156_191518-auto

precompiled/qualcomm-sa8775p/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f8b0bc321754d064fd41e62589532b29288f6cbd095ba64787e66b8cfc2e9e1
+size 7045120

precompiled/qualcomm-sa8775p/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7d689cea00cb723269f4513baf6f3202fb5f58154d32026d227d5fe56447304
+size 151408640

precompiled/qualcomm-sa8775p/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bcdbd48d58b024ffd9bc27b8c04314d9e10de4dbb48ffbd0bd0b7fa430ca19d
+size 6529024

precompiled/qualcomm-sa8775p/tool-versions.yaml

@@ -0,0 +1,3 @@
+tool_versions:
+  qnn_context_binary:
+    qairt: 2.41.0.251128145156_191518-auto

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79024f5d6ea102d1b5a9dc3ef5152571e32ac0d9bc6d06dbf5bc50ac1a99f5aa
+size 7045120

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Zipformer_ZipformerDecoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fb2b946ef3304467d6bf4afd95f7a5f348975d83d0080e7001c6ff9a42d7e5a
+size 6501931

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14c778faddd4584bcb432c01f919e8e1f87571396fca730b15027f049888853a
+size 151379968

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Zipformer_ZipformerEncoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf1f35d67a152bddc6c0d6cb898f87088cc17c003ce4e93e622f53abc6f8265f
+size 118412396

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3733e7e5fb4623c12dfb0de06baa0a7386af67faadaa1aba251e1dc4b94e63c2
+size 6529024

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/Zipformer_ZipformerJoiner_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21519e0df17a95012d77a3eae04622bd2740bb27d8306026328976bb66ffecaa
+size 5776047

precompiled/qualcomm-snapdragon-8-elite-for-galaxy/tool-versions.yaml

@@ -0,0 +1,4 @@
+tool_versions:
+  precompiled_qnn_onnx:
+    qairt: 2.37.1.250807093845_124904
+    onnx_runtime: 1.23.0

precompiled/qualcomm-snapdragon-8-elite-gen5/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8583097b1d1367f0314d598a06bce244ee9243e3da845af3d842f8fa58817c59
+size 7045120

precompiled/qualcomm-snapdragon-8-elite-gen5/Zipformer_ZipformerDecoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29e5a6fef628d2e46b6f634f919abfc3650fe058d9c229f06d401fe3fecf4ca0
+size 6502284

precompiled/qualcomm-snapdragon-8-elite-gen5/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db20f0464d17f88b7b9d05b5b519253a85e5090b725a737ac39cdb58b3a67537
+size 151760896

precompiled/qualcomm-snapdragon-8-elite-gen5/Zipformer_ZipformerEncoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8296211b87b5e7e37160437c3a656b3e2788c8119a09b8014c7af37f68e9f69f
+size 118521474

precompiled/qualcomm-snapdragon-8-elite-gen5/Zipformer_ZipformerJoiner_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8300e75134f7634694a7dd4e89e05aad4110470acfcfdf3997f1c5cad8fa8201
+size 6533120

precompiled/qualcomm-snapdragon-8-elite-gen5/Zipformer_ZipformerJoiner_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:928d702b609926dfc586da647f9d52b97cedde9529b7f58a6ab19d5642ba3cf4
+size 5777070

precompiled/qualcomm-snapdragon-8-elite-gen5/tool-versions.yaml

@@ -0,0 +1,4 @@
+tool_versions:
+  precompiled_qnn_onnx:
+    qairt: 2.37.1.250807093845_124904
+    onnx_runtime: 1.23.0

precompiled/qualcomm-snapdragon-8gen3/Zipformer_ZipformerDecoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df7cfc2c3a450cce4c75dc85f9e30bf09362223a64b256bd214608b1141e2ce1
+size 7045120

precompiled/qualcomm-snapdragon-8gen3/Zipformer_ZipformerDecoder_float.onnx.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d58a4e0c407be5352f522e23112dfa9b392944f6ad4bf4cd22e7d4ea548e0c8
+size 6501930

precompiled/qualcomm-snapdragon-8gen3/Zipformer_ZipformerEncoder_float.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:929972502ee35e5cff8b166a239203ea4dbf963ad248f97c94b602086010bb65
+size 151408640