tuna2134/sbv2-v2
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

Fix: Inference failure with some ExecutionProviders

Browse Source 
length_scale, etc. were passed as one-dimensional arrays when they should have been passed as scalar values.
This commit is contained in:
tsukumi
2024-09-28 17:36:26 +09:00
parent 7793e9da43
commit c4b6c169fc
1 changed files with 8 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
style_bert_vits2/models/infer_onnx.py
View File

@@ -165,10 +165,10 @@ def infer_onnx(
lang_ids,
ja_bert,
style_vec_tensor,
np.array([length_scale], dtype=np.float32),
np.array([sdp_ratio], dtype=np.float32),
np.array([noise_scale], dtype=np.float32),
np.array([noise_scale_w], dtype=np.float32),
np.array(length_scale, dtype=np.float32),
np.array(sdp_ratio, dtype=np.float32),
np.array(noise_scale, dtype=np.float32),
np.array(noise_scale_w, dtype=np.float32),
]
else:
input_tensor = [
@@ -181,10 +181,10 @@ def infer_onnx(
ja_bert,
en_bert,
style_vec_tensor,
np.array([length_scale], dtype=np.float32),
np.array([sdp_ratio], dtype=np.float32),
np.array([noise_scale], dtype=np.float32),
np.array([noise_scale_w], dtype=np.float32),
np.array(length_scale, dtype=np.float32),
np.array(sdp_ratio, dtype=np.float32),
np.array(noise_scale, dtype=np.float32),
np.array(noise_scale_w, dtype=np.float32),
]
# 入力テンソルを転送する GPU デバイスを取得