55 Commits

Author	SHA1	Message	Date
henryruhs	94571c5676	Rename attribute to feature	2025-03-16 08:39:35 +01:00
henryruhs	b5efcbe44a	Adjust the weights in README	2025-03-12 23:02:26 +01:00
henryruhs	5550c89f43	Fix AAD naming, Attribute and Embedding	2025-03-12 16:30:59 +01:00
henryruhs	2af4f2c8ab	Extend config	2025-03-12 16:21:47 +01:00
henryruhs	afab997ffc	Rename parser to face parser	2025-03-11 14:43:10 +01:00
harisreedhar	f90fd73b54	add strategy to config	2025-03-11 14:43:10 +01:00
harisreedhar	d809f66216	changes	2025-03-11 14:43:10 +01:00
henryruhs	54923abf7f	Follow layers and sequences pattern	2025-03-11 14:43:10 +01:00
henryruhs	d2efb2fd08	Fix README.md	2025-03-11 14:43:10 +01:00
harisreedhar	64464b6f1c	fix	2025-03-11 14:43:10 +01:00
harisreedhar	a3ac4d5ddd	changes	2025-03-11 14:43:10 +01:00
harisreedhar	1659805b08	changes	2025-03-11 14:43:10 +01:00
harisreedhar	8f1f002c64	add masknet	2025-03-11 14:43:10 +01:00
harisreedhar	7cf5609c1f	changes	2025-03-11 14:43:10 +01:00
henryruhs	aedaa20d78	Remove UnetPro, make float values visible in README	2025-03-11 14:43:10 +01:00
henryruhs	5056b8df75	Variable AAD layer according to output size	2025-03-11 14:43:10 +01:00
henryruhs	430c71d031	Adjust config and namings	2025-03-11 14:43:10 +01:00
harisreedhar	176dced1f6	changes	2025-03-11 14:43:10 +01:00
henryruhs	18a605e1a3	Change the weight config order	2025-03-11 14:43:10 +01:00
harisreedhar	f2d3f8a19f	changes	2025-03-11 14:43:10 +01:00
henryruhs	d93daf9e5f	Introduce batch mode via config for equal and same batches	2025-03-11 14:43:10 +01:00
henryruhs	3a61da8bab	Introduce batch mode via config for equal and same batches	2025-03-11 14:43:10 +01:00
harisreedhar	83c20f8331	changes	2025-03-11 14:43:10 +01:00
henryruhs	80d1694e23	Replace naming warp_matrix with warp_template	2025-03-11 14:43:10 +01:00
henryruhs	589568bfb5	Add config to dataloader alignment	2025-03-11 14:43:10 +01:00
henryruhs	5b3b2abdd7	Remove annoying Tuner, Match both trainer configs a bit more	2025-03-11 14:43:10 +01:00
henryruhs	ee3fc40e83	Move output channels to config	2025-03-11 14:43:10 +01:00
henryruhs	14bbece850	Rename id_ to identity_	2025-03-11 14:43:10 +01:00
henryruhs	6eabcad1d0	Introduce new AttributeLoss class	2025-03-11 14:43:10 +01:00
henryruhs	086d9eed87	Introduce new ReconstructionLoss class	2025-03-11 14:43:09 +01:00
henryruhs	085c493e18	Rename id_embedder to embedder, Tons of naming in training step, Introduce new IdentityLoss class	2025-03-11 14:43:09 +01:00
henryruhs	dc0abff0ce	Minor typo	2025-03-11 14:43:09 +01:00
henryruhs	10ce04ed58	Introduce feature flag for Unet	2025-03-11 14:43:09 +01:00
henryruhs	83ef075b1d	Remove map_location	2025-03-11 14:43:09 +01:00
henryruhs	d153c68813	Better split_ratio	2025-03-11 14:43:09 +01:00
henryruhs	c17378f3c7	Let's call it batch ratio	2025-03-11 14:43:09 +01:00
henryruhs	84503761b9	follow the not invented here syndrome	2025-03-11 14:43:09 +01:00
harisreedhar	db44c91dd8	changes	2025-03-11 14:43:09 +01:00
henryruhs	04eaa831ea	Rename to resume_path only	2025-03-11 14:43:09 +01:00
henryruhs	f63bc788ac	Fix typo	2025-03-11 14:43:09 +01:00
henryruhs	11bb9065ba	Uniform resume checkpoint approach	2025-03-11 14:43:09 +01:00
henryruhs	5bb41ecbb2	Adjust README	2025-03-11 14:43:09 +01:00
henryruhs	777a8384c2	Extend README with TensorBoard guide	2025-03-11 14:43:09 +01:00
henryruhs	d9e10a9f7c	Use lightning over pytorch_lightning import, Configure tensorboard	2025-03-11 14:43:09 +01:00
henryruhs	026bcf0c97	Adjust licenses	2025-03-11 14:43:09 +01:00
harisreedhar	8f8dfecdbc	some fixes	2025-03-11 14:43:09 +01:00
henryruhs	f05ff6cdb1	Enforce IR version for older onnxruntime	2025-03-11 14:43:09 +01:00
henryruhs	3fe32d7832	Fix example config	2025-03-11 14:43:09 +01:00
henryruhs	860771e482	Fix weight identity	2025-03-11 14:43:09 +01:00
henryruhs	1b6e7a6ca5	Rename ArcFace Converter to Embedding Converter, Add EmbeddingDataset, Add learning rate to config	2025-03-11 14:43:06 +01:00