Skip to content

fix: use multi-GPU RWKV strategy from visible CUDA devices #3882

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Chessing234 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: use multi-GPU RWKV strategy from visible CUDA devices #3882

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 7 additions & 3 deletions fastchat/model/rwkv_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,9 +17,13 @@ def __init__(self, model_path):
"Experimental support. Please use ChatRWKV if you want to chat with RWKV"
)
self.config = SimpleNamespace(is_encoder_decoder=False)
self.model = RWKV(model=model_path, strategy="cuda fp16")
# two GPUs
# self.model = RWKV(model=model_path, strategy="cuda:0 fp16 *20 -> cuda:1 fp16")
n = torch.cuda.device_count()
strategy = (
" -> ".join(f"cuda:{i} fp16" for i in range(n))
if n > 1
else "cuda fp16"
)
self.model = RWKV(model=model_path, strategy=strategy)

self.tokenizer = None
self.model_path = model_path
Expand Down
Loading