Update q8 weights (reduce_range=True, per_channel=True) (#1)

- Update q8 weights (reduce_range=True, per_channel=True) (ee2abfb8710b18e47303b781603d424f9d5d5edb)
- Delete onnx/decoder_model_merged_quantized.onnx (01f6aae906fdfca2dc87c3fc3c5536da4b8e96bf)
- Rename onnx/model_quantized.onnx to onnx/decoder_model_merged_quantized.onnx (1e9db7c2c298df45987ecda3e922e057448cf0a5)

onnx/decoder_model_merged_quantized.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7cdc95f39bf11c41805ea4e274236dc9807a3ad18879b34fcef73adeaf7a5715
-size 482363069
+oid sha256:5eddb33de046ceabef9072b55b203cf3065a654e62a6d07efc5dbc96d4561dfc
+size 484413840