Add ONNX weights (#1)

- Add ONNX weights (1a5e9052cb16a2a69e1b51aec90d2f335ef486b8)


Co-authored-by: Joshua <Xenova@users.noreply.huggingface.co>

onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4ab4c480f402e70429d67dcc80a2bd88476c166824dcb1d10602566c485fd0f
+size 1741981193

onnx/model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d03bd791409057b78184fa7726effdf5429114a68a7757939aa0d350761f7697
+size 642585535

onnx/quantize_config.json

@@ -0,0 +1,50 @@
+{
+    "per_channel": true,
+    "reduce_range": true,
+    "per_model_config": {
+        "model": {
+            "op_types": [
+                "Ceil",
+                "Transpose",
+                "Abs",
+                "GatherElements",
+                "Sign",
+                "Unsqueeze",
+                "Pow",
+                "Log",
+                "Erf",
+                "Shape",
+                "Softmax",
+                "Div",
+                "Squeeze",
+                "Expand",
+                "Constant",
+                "Sqrt",
+                "Gemm",
+                "Less",
+                "ConstantOfShape",
+                "Range",
+                "Clip",
+                "LessOrEqual",
+                "MatMul",
+                "ReduceMean",
+                "Identity",
+                "Where",
+                "Gather",
+                "Greater",
+                "Neg",
+                "Equal",
+                "Concat",
+                "Add",
+                "Slice",
+                "Mul",
+                "Tile",
+                "Cast",
+                "Reshape",
+                "And",
+                "Sub"
+            ],
+            "weight_type": "QInt8"
+        }
+    }
+}