mrq 218d0e29fd ugh (batchmean actually expects batch=seq_len, and not the actual batch)		2024-12-07 12:39:01 -06:00
..
emb	fixes	2024-11-10 20:37:50 -06:00
engines	ACTUALLY actually fix KD-loss (the -inf in the logits was caused by cringecode)	2024-12-07 12:31:54 -06:00
models	ACTUALLY actually fix KD-loss (the -inf in the logits was caused by cringecode)	2024-12-07 12:31:54 -06:00
utils	agnostified KD	2024-12-06 23:53:46 -06:00
__init__.py	Rewrite init	2023-08-02 21:53:35 +00:00
__main__.py	actually fixed knowledge distillation because of errant -inf logits causing problems and needed to be filtered (and splitting text language / output audio language because it helps)	2024-12-06 21:55:20 -06:00
config.py	ACTUALLY do KD-loss because of an oversight with masked_select outputting 1D tensors that get softmax'd in total	2024-12-07 09:52:51 -06:00
data.py	oops	2024-12-04 21:24:22 -06:00
demo.py	touch ups in docs	2024-12-02 19:10:42 -06:00
export.py	cringe code to convert to LlamaForCausalLM-happy weights + tokenizer dict (still need to write logic to actually use these weights for proper inferencing)	2024-12-03 10:18:58 -06:00
inference.py	actually fixed knowledge distillation because of errant -inf logits causing problems and needed to be filtered (and splitting text language / output audio language because it helps)	2024-12-06 21:55:20 -06:00
plot.py	very, very naive layerskip speculative sampling (it just checks if the current layer's state is good enough)	2024-11-02 11:49:05 -05:00
samplers.py	cleaned up classifier-free guidance logit processing (in order to try and cope with a bad nar-len model)	2024-11-19 10:30:05 -06:00
train.py	ugh (batchmean actually expects batch=seq_len, and not the actual batch)	2024-12-07 12:39:01 -06:00
webui.py	actually fixed knowledge distillation because of errant -inf logits causing problems and needed to be filtered (and splitting text language / output audio language because it helps)	2024-12-06 21:55:20 -06:00