github
/
minimind-o
mirror of https://github.com/jingyaogong/minimind-o.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Compare commits

base: github:f7b0a325c13540ebab00b8e753b6f18a77d5c771
Branches Tags
github:master
github:website
..
compare: github:0d1ae504f29397820f097aca965e46570f803c0f
Branches Tags
github:master
github:website

No commits in common. "f7b0a325c13540ebab00b8e753b6f18a77d5c771" and "0d1ae504f29397820f097aca965e46570f803c0f" have entirely different histories.
f7b0a325c1 ... 0d1ae504f2

2 changed files with 2 additions and 3 deletions
Show Stats Expand all files Collapse all files

1
requirements.txt
View File

@ -6,7 +6,6 @@ transformers==4.57.6
speechbrain==1.1.0 speechbrain==1.1.0
onnxruntime==1.23.2 onnxruntime==1.23.2
funasr==1.3.1 funasr==1.3.1
addict==2.4.0
flask==3.0.3 flask==3.0.3
flask-cors==6.0.1 flask-cors==6.0.1
flask-sock==0.7.0 flask-sock==0.7.0

4
webui/web_demo.py
View File

@ -69,7 +69,7 @@ def build_ids(prompt, history):
def _mimi_decode(frames): def _mimi_decode(frames):
codes = [f for f in frames if f and len(f) == 8] codes = [f for f in frames if f and len(f) == 8]
if not codes or not M['mimi']: return None if not codes or not M['mimi']: return None
mc = torch.tensor(codes, dtype=torch.long, device=M['device']).T.unsqueeze(0) mc = torch.tensor(codes, dtype=torch.long).T.unsqueeze(0)
mc = torch.where(mc >= 2049, torch.zeros_like(mc), mc) mc = torch.where(mc >= 2049, torch.zeros_like(mc), mc)
with torch.no_grad(): with torch.no_grad():
au = M['mimi'].decode(mc).audio_values.squeeze().cpu().numpy() au = M['mimi'].decode(mc).audio_values.squeeze().cpu().numpy()
@ -486,7 +486,7 @@ def init_model(args):
au = torch.full((1, 8, 3), 2049, dtype=torch.long, device=args.device) au = torch.full((1, 8, 3), 2049, dtype=torch.long, device=args.device)
M['model'].forward(torch.cat((au, ids.unsqueeze(1)), dim=1)) M['model'].forward(torch.cat((au, ids.unsqueeze(1)), dim=1))
if M['model'].audio_encoder: M['model'].audio_encoder(torch.zeros(1, 100, 560, device=args.device), torch.tensor([100], device=args.device)) if M['model'].audio_encoder: M['model'].audio_encoder(torch.zeros(1, 100, 560, device=args.device), torch.tensor([100], device=args.device))
if M['mimi']: M['mimi'].decode(torch.zeros(1, 8, 1, dtype=torch.long, device=args.device)) if M['mimi']: M['mimi'].decode(torch.zeros(1, 8, 1, dtype=torch.long))
print('Warmup done!') print('Warmup done!')