verachen/Fun-ASR
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add decode script

Browse Source
This commit is contained in:
pengzhendong
2026-01-04 15:12:17 +08:00
parent c35ce7601a
commit 403d305df9
5 changed files with 85 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

45
decode.py Normal file
View File

@ -0,0 +1,45 @@
import argparse
import os
import torch
from model import FunASRNano
def main():
parser = argparse.ArgumentParser(description="FunASR-Nano")
parser.add_argument("--scp-file", type=str, required=True)
parser.add_argument("--output-file", type=str, required=True)
parser.add_argument(
"--model-dir", type=str, default="FunAudioLLM/Fun-ASR-Nano-2512"
)
args = parser.parse_args()
device = (
"cuda:0"
if torch.cuda.is_available()
else "mps"
if torch.backends.mps.is_available()
else "cpu"
)
m, kwargs = FunASRNano.from_pretrained(model=args.model_dir, device=device)
m.eval()
output_dir = os.path.dirname(args.output_file)
if output_dir and not os.path.exists(output_dir):
os.makedirs(output_dir, exist_ok=True)
with open(args.scp_file, "r", encoding="utf-8") as f1:
with open(args.output_file, "w", encoding="utf-8") as f2:
for line in f1:
line = line.strip()
if not line:
continue
parts = line.split(maxsplit=1)
if len(parts) == 2:
text = m.inference(data_in=[parts[1]], **kwargs)[0][0]["text"]
f2.write(f"{parts[0]}\t{text}\n")
if __name__ == "__main__":
main()

6
demo1.py
View File

@ -7,7 +7,9 @@ def main():
device = (
"cuda:0"
if torch.cuda.is_available()
else "mps" if torch.backends.mps.is_available() else "cpu"
else "mps"
if torch.backends.mps.is_available()
else "cpu"
)
model = AutoModel(
model=model_dir,
@ -28,7 +30,7 @@ def main():
# 匈牙利语、爱尔兰语、拉脱维亚语、立陶宛语、马耳他语、波兰语、葡萄牙语、罗马尼亚语、
# 斯洛伐克语、斯洛文尼亚语、瑞典语 for Fun-ASR-MLT-Nano-2512
language="中文",
itn=True, # or False
itn=True, # or False
)
text = res[0]["text"]
print(text)

4
demo2.py
View File

@ -8,7 +8,9 @@ def main():
device = (
"cuda:0"
if torch.cuda.is_available()
else "mps" if torch.backends.mps.is_available() else "cpu"
else "mps"
if torch.backends.mps.is_available()
else "cpu"
)
m, kwargs = FunASRNano.from_pretrained(model=model_dir, device=device)
m.eval()

15
model.py
View File

@ -117,7 +117,9 @@ class FunASRNano(nn.Module):
if init_param_path is not None:
src_state = torch.load(init_param_path, map_location="cpu")
flag = self.ctc_decoder.load_state_dict(src_state, strict=False)
logging.info(f"Loading ctc_decoder ckpt: {init_param_path}, status: {flag}")
logging.info(
f"Loading ctc_decoder ckpt: {init_param_path}, status: {flag}"
)
freeze = ctc_decoder_conf.get("freeze", False)
if freeze:
for _, param in self.ctc_decoder.named_parameters():
@ -127,7 +129,12 @@ class FunASRNano(nn.Module):
ctc_conf = kwargs.get("ctc_conf", {})
self.blank_id = ctc_conf.get("blank_id", ctc_vocab_size - 1)
self.ctc_weight = kwargs.get("ctc_weight", 0.3)
self.ctc = CTC(odim=ctc_vocab_size, encoder_output_size=audio_encoder_output_size, blank_id=self.blank_id, **ctc_conf)
self.ctc = CTC(
odim=ctc_vocab_size,
encoder_output_size=audio_encoder_output_size,
blank_id=self.blank_id,
**ctc_conf,
)
self.detach_ctc_decoder = kwargs.get("detach_ctc_decoder", True)
self.error_calculator = None
@ -629,7 +636,7 @@ class FunASRNano(nn.Module):
with torch.autocast(
device_type=device_type if device_type in ["cuda", "xpu", "mps"] else "cpu",
enabled=True if llm_dtype != "fp32" else False,
dtype=dtype_map[llm_dtype]
dtype=dtype_map[llm_dtype],
):
label = contents["assistant"][-1]
self.llm = self.llm.to(dtype_map[llm_dtype])
@ -677,7 +684,7 @@ class FunASRNano(nn.Module):
response_clean = re.sub(r"[^\w\s\u3000\u4e00-\u9fff]+", "", response)
result_i = {
"key": key[0],
"text": re.sub(r'\s+', ' ', response.replace("/sil", " ")),
"text": re.sub(r"\s+", " ", response.replace("/sil", " ")),
"text_tn": response_clean,
"label": label,
}

34
tools/scp2jsonl.py
View File

@ -17,8 +17,12 @@ def parse_args():
parser.add_argument("--scp-file", type=str, required=True)
parser.add_argument("--transcript-file", type=str, required=True)
parser.add_argument("--jsonl-file", type=str, required=True)
parser.add_argument("--max-workers", type=int, default=8,
help="Number of concurrent workers (default: 8)")
parser.add_argument(
"--max-workers",
type=int,
default=8,
help="Number of concurrent workers (default: 8)",
)
return parser.parse_args()
@ -59,11 +63,14 @@ class LineProcessor:
data = {
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": f"语音转写：<|startofspeech|>!{wav_path}<|endofspeech|>"},
{"role": "assistant", "content": text}
{
"role": "user",
"content": f"语音转写：<|startofspeech|>!{wav_path}<|endofspeech|>",
},
{"role": "assistant", "content": text},
],
"speech_length": int((duration * 1000 - 25) // 10 + 1),
"text_length": len(self.tokenizer.tokenize(text))
"text_length": len(self.tokenizer.tokenize(text)),
}
return {"success": data, "utt": utt1}
@ -79,7 +86,9 @@ def main():
transcript_lines = f2.readlines()
if len(scp_lines) != len(transcript_lines):
print(f"Warning: Line count mismatch - scp: {len(scp_lines)}, transcript: {len(transcript_lines)}")
print(
f"Warning: Line count mismatch - scp: {len(scp_lines)}, transcript: {len(transcript_lines)}"
)
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-0.6B")
processor = LineProcessor(tokenizer)
@ -93,8 +102,10 @@ def main():
with tqdm(total=len(data_pairs), desc="Processing") as pbar:
with ThreadPoolExecutor(max_workers=args.max_workers) as executor:
with open(args.jsonl_file, "w") as f_out:
futures = {executor.submit(processor.process_line, pair): i
for i, pair in enumerate(data_pairs)}
futures = {
executor.submit(processor.process_line, pair): i
for i, pair in enumerate(data_pairs)
}
for future in as_completed(futures):
result = future.result()
@ -109,10 +120,9 @@ def main():
error_messages.append(result["error"])
pbar.update(1)
pbar.set_postfix({
"processed": processed_count,
"failed": failed_count
})
pbar.set_postfix(
{"processed": processed_count, "failed": failed_count}
)
print(f"\nProcessing completed:")
print(f" Total lines: {len(data_pairs)}")