James Betker ee8ceed6da rework tfd13 further ... - use a gated activation layer for both attention & convs - add a relativistic learned position bias. I believe this is similar to the T5 position encodings but it is simpler and learned - get rid of prepending to the attention matrix - this doesn't really work that well. the model eventually learns to attend one of its heads to these blocks but why not just concat if it is doing that?		2022-07-20 23:28:29 -06:00
..
__init__.py	music!	2022-04-15 21:21:37 -06:00
cheater_gen_ar.py	some stuff	2022-07-13 21:26:25 -06:00
diffwave.py	output variance	2022-05-02 00:10:33 -06:00
encoders.py	some stuff	2022-07-13 21:26:25 -06:00
flat_diffusion.py	fd fix	2022-05-26 20:19:09 -06:00
gpt_music.py	gptmusic work	2022-06-16 15:09:47 -06:00
gpt_music2.py	some stuff	2022-07-13 21:26:25 -06:00
instrument_quantizer.py	iq checkin	2022-07-20 10:19:32 -06:00
m2v_code_to_mel.py	some stuff	2022-05-27 11:40:31 -06:00
mel2vec_codes_gpt.py	m2v stuff	2022-05-20 11:01:17 -06:00
music_quantizer.py	bunch of new stuff	2022-06-04 22:23:08 -06:00
music_quantizer2.py	Revert previous changes	2022-06-10 14:34:05 -06:00
tfdpc_v5.py	fix script	2022-07-19 11:17:43 -06:00
transformer_diffusion12.py	fix compatibility	2022-07-13 21:28:20 -06:00
transformer_diffusion13.py	rework tfd13 further	2022-07-20 23:28:29 -06:00
transformer_diffusion14.py	misc	2022-07-20 10:19:02 -06:00
unet_diffusion_music_codes.py	network updates	2022-06-08 09:26:59 -06:00
unet_diffusion_waveform_gen_simple.py	Simpler form of the wavegen model	2022-05-06 16:37:04 -06:00
unet_diffusion_waveform_gen.py	few things with gap filling	2022-05-06 14:33:44 -06:00
unet_diffusion_waveform_gen3.py	tfdpc_v4	2022-06-25 21:17:00 -06:00