DK
DK
> Tensorflow is 1.12 and Keras is 2.2.4. I'm using the dataset API so maybe that's the problem. I use the same version as you do. Have you solved this...
> @FB-wh No I haven't. Waiting for the next version of tensorflow. Golbstein had solved the problem #11605. I've tried this method, and it works. https://github.com/Golbstein/KerasExtras/blob/master/BN16.py
 I'm using the TinyLlama-1.1B-Chat-v1.0 model, but soon the gradient will become nan.
I also meet this problem.How to solve it please？
any update?
CUDA_VISIBLE_DEVICES=0,1 torchrun --nproc_per_node 2 scripts/inference.py configs/opensora-v1-2/inference/sample.py --num-frames 2s --resolution 360p --aspect-ratio 9:16 --prompt "a beautiful waterfall" OOM
Have you solved it?