apple m1?
iamcos opened this issue · 2 comments
iamcos commented
Error message ive got:tango/venv/lib/python3.10/site-packages/torch/cuda/init.py", line 221, in _lazy_init
raise AssertionError("Torch not compiled with CUDA enabled")
AssertionError: Torch not compiled with CUDA enabled
iamcos commented
Did the following:
tango = Tango("declare-lab/tango", "mps")
Current error:
/venv/lib/python3.10/site-packages/transformers/models/t5/modeling_t5.py", line 414, in _relative_position_bucket
relative_position = torch.abs(relative_position)
TypeError: Operation 'abs_out_mps()' does not support input type 'int64' in MPS backend.
iamcos commented
After installing latest nightly pytorch build the problem went away, now its generating.
It takes aproximately 84 seconds for one generation