i'm seeing an error about unknown tensor when running this branch. EDIT: this was resolved with running convert.py from this branch
i'm hitting a Floating point exception immediately following bark_print_statistics for the bark_eval_text_encoder phase. i tried with both f32 and f16 converted weights
is this ready for user testing yet?
i'm seeing an error aboutunknown tensor
when running this branch. EDIT: this was resolved with runningconvert.py
from this branchi'm hitting a
Floating point exception
immediately followingbark_print_statistics
for thebark_eval_text_encoder
phase. i tried with both f32 and f16 converted weightsquantization to q4_0 is also failing.