From 61e4203fbf80c6951de777b66e63eea480f283b8 Mon Sep 17 00:00:00 2001
From: zhifu gao <zhifu.gzf@alibaba-inc.com>
Date: 星期四, 20 六月 2024 17:14:29 +0800
Subject: [PATCH] sensevoice (#1834)
---
funasr/auto/auto_model.py | 19 +++++++------------
1 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/funasr/auto/auto_model.py b/funasr/auto/auto_model.py
index 603c0a0..57a15db 100644
--- a/funasr/auto/auto_model.py
+++ b/funasr/auto/auto_model.py
@@ -92,7 +92,8 @@
if isinstance(data_i, str) and os.path.exists(data_i):
key = misc.extract_filename_without_extension(data_i)
else:
- key = "rand_key_" + "".join(random.choice(chars) for _ in range(13))
+ if key is None:
+ key = "rand_key_" + "".join(random.choice(chars) for _ in range(13))
key_list.append(key)
else: # raw text; audio sample point, fbank; bytes
@@ -213,7 +214,6 @@
deep_update(model_conf, kwargs.get("model_conf", {}))
deep_update(model_conf, kwargs)
model = model_class(**model_conf, vocab_size=vocab_size)
- model.to(device)
# init_param
init_param = kwargs.get("init_param", None)
@@ -236,6 +236,7 @@
model.to(torch.float16)
elif kwargs.get("bf16", False):
model.to(torch.bfloat16)
+ model.to(device)
return model, kwargs
def __call__(self, *args, **cfg):
@@ -324,7 +325,7 @@
input, input_len=input_len, model=self.vad_model, kwargs=self.vad_kwargs, **cfg
)
end_vad = time.time()
-
+
# FIX(gcf): concat the vad clips for sense vocie model for better aed
if kwargs.get("merge_vad", False):
for i in range(len(res)):
@@ -466,7 +467,7 @@
result[k] = restored_data[j][k]
else:
result[k] += restored_data[j][k]
-
+
if not len(result["text"].strip()):
continue
return_raw_text = kwargs.get("return_raw_text", False)
@@ -481,7 +482,7 @@
if return_raw_text:
result["raw_text"] = raw_text
result["text"] = punc_res[0]["text"]
-
+
# speaker embedding cluster after resorted
if self.spk_model is not None and kwargs.get("return_spk_res", True):
if raw_text is None:
@@ -602,12 +603,6 @@
)
with torch.no_grad():
-
- if type == "onnx":
- export_dir = export_utils.export_onnx(model=model, data_in=data_list, **kwargs)
- else:
- export_dir = export_utils.export_torchscripts(
- model=model, data_in=data_list, **kwargs
- )
+ export_dir = export_utils.export(model=model, data_in=data_list, **kwargs)
return export_dir
--
Gitblit v1.9.1