Closed TalentBoy2333 closed 2 weeks ago
Thanks for sharing!
Another workaround is:
python scripts/inference.py configs/opensora-v1-2/inference/sample.py \
--num-frames 4s --resolution 720p \
--layernorm-kernel False --flash-attn False \
--prompt "a beautiful waterfall"
试过关掉 flash-attn
, 在 2 张 A100 上 cuda oom 了
Yes, without flash-attn 720p it costs too much memory.
Maybe I should make more money and buy some GPUs that don't exist in China. :doge
运行
报错
参考 https://github.com/Dao-AILab/flash-attention/issues/966 解决