Add files using upload-large-folder tool

README.md

@@ -0,0 +1,27 @@
+---
+license: gpl-3.0
+tags:
+  - voice-conversion
+  - rvc
+  - onnx
+  - browser
+---
+
+# RVC browser runtime
+
+Browser-compatible FP16 ContentVec and RMVPE graphs used by the fully
+client-side Nepali RVC demo.
+
+The source graphs are from
+[`wok000/weights_gpl`](https://huggingface.co/wok000/weights_gpl/tree/d1b578cfcdf944dd3910786f7604754c79234b36),
+which licenses its model files under GPL-3.0. The graphs here retain FP32 input
+and output tensors, convert internal floating-point weights and computation to
+FP16, and are topologically sorted for ONNX Runtime Web 1.22.
+
+| File | Bytes | SHA-256 |
+|---|---:|---|
+| `contentvec.fp16.onnx` | 189,437,475 | `205d92e4dea32ba9f309608e05209c01f0f8e35ed927c6d80ac98464e91cca83` |
+| `rmvpe.fp16.onnx` | 181,070,368 | `22cd95fc3f6dd91619c0ba62cf3aff51339d617cdffbf5eabd915d4d58e8bfb2` |
+
+The files are static model assets. Audio is processed by ONNX Runtime inside
+the browser and is not sent to this repository or an inference API.

contentvec.fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:205d92e4dea32ba9f309608e05209c01f0f8e35ed927c6d80ac98464e91cca83
+size 189437475

manifest.json

@@ -0,0 +1,18 @@
+{
+  "version": "1.0.0",
+  "source": {
+    "repo": "wok000/weights_gpl",
+    "revision": "d1b578cfcdf944dd3910786f7604754c79234b36",
+    "license": "GPL-3.0"
+  },
+  "files": {
+    "contentvec.fp16.onnx": {
+      "bytes": 189437475,
+      "sha256": "205d92e4dea32ba9f309608e05209c01f0f8e35ed927c6d80ac98464e91cca83"
+    },
+    "rmvpe.fp16.onnx": {
+      "bytes": 181070368,
+      "sha256": "22cd95fc3f6dd91619c0ba62cf3aff51339d617cdffbf5eabd915d4d58e8bfb2"
+    }
+  }
+}

rmvpe.fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22cd95fc3f6dd91619c0ba62cf3aff51339d617cdffbf5eabd915d4d58e8bfb2
+size 181070368