Commit Graph

Select branches

Hide Pull Requests

master

Mono Color

• c34eb21f4f corrige crash com sdxl master Glauber Ferreira 2025-06-28 08:02:12 -0300
• 10facffd01 fix bosta Glauber Ferreira 2025-06-27 21:48:00 -0300
• 10c6501bd0
  fix missing argument in prototype of stbi_write_jpg (#613) vmobilis 2025-03-09 07:30:10 +0300
• 10feacf031
  fix: correct img2img time (#616) vmobilis 2025-03-09 07:29:08 +0300
• 655f8a5169
  fix: clang complains about needless braces (#618) vmobilis 2025-03-09 07:26:41 +0300
• d7c7a34712
  fix: ModelLoader::load_tensors duplicated check (#623) idostyle 2025-03-09 05:23:23 +0100
• 81556f3136
  chore: silence some warnings about precision loss (#620) vmobilis 2025-03-09 07:22:39 +0300
• 3fb275a67b
  fix: suport sdxl embedddings (#621) stduhpf 2025-03-09 05:21:23 +0100
• 30b3ac8e62 fix: avoid potential dangling pointer problem leejet 2025-03-01 16:58:26 +0800
• 195d170136 sync: update ggml leejet 2025-03-01 12:09:55 +0800
• f50a7f66aa
  fix: fix race condition causing inconsistent value for decoder_only (#609) stduhpf 2025-03-01 04:49:06 +0100
• 85e9a12988
  fix: preprocess tensor names in tensor types map (#607) stduhpf 2025-03-01 04:48:04 +0100
• fbd42b6fc1
  fix: fix embeddings with quantized models (#601) stduhpf 2025-03-01 04:45:39 +0100
• 19d876ee30
  feat: implement DDIM with the "trailing" timestep spacing and TCD (#568) yslai 2025-02-22 05:34:22 -0800
• f27f2b2aa2
  docs: add missing --mask and --guidance options to print_usage (#572) lalala 2025-02-22 21:32:37 +0800
• 99609761dc
  docs: fix typo in readme (#574) piallai 2025-02-22 14:30:28 +0100
• 69c73789fe
  fix: force binary mask for inpaint models (#589) stduhpf 2025-02-22 14:29:57 +0100
• 838beb9b5e
  chore: add global SYCL compile flags (#597) Meng, Hengyu 2025-02-22 21:23:58 +0800
• f23b803a6b
  fix:: unapply current loras properly (#590) stduhpf 2025-02-22 14:22:22 +0100
• 1be2491dcf
  feat: partial LyCORIS support (tucker decomposition for LoCon + LoHa + LoKr) (#577) stduhpf 2025-02-22 14:19:26 +0100
• 3753223982
  fix: make get_files_from_dir works with absolute path (#598) Matti Pulkkinen 2025-02-22 15:16:50 +0200
• 59ca2b0f16
  chore: bump MUSA SDK version to rc3.1.1 (#599) R0CKSTAR 2025-02-22 21:14:26 +0800
• d46ed5e184
  feat: support JPEG compression (#583) vmobilis 2025-02-05 11:18:02 +0300
• 2535ad5a43
  chore: fix cuda on github action (#580) ag2s20150909 2025-02-05 16:15:41 +0800
• e500d95abd
  fix: fix rank 1 loras (#575) stduhpf 2025-02-05 09:13:17 +0100
• a3cbdf6dcb
  chore: SD_USE_CUBLAS => SD_USE_CUDA for MUSA backend (#578) R0CKSTAR 2025-02-05 16:11:26 +0800
• 5eb15ef4d0
  docs: add CLI-GUI to list (#546) piallai 2025-01-18 06:16:54 +0100
• d9b5942d98
  feat: add sdxl v-pred suppport (#536) stduhpf 2025-01-18 06:15:54 +0100
• 587a37b2e2
  fix: avoid sd2((non inpaint) crash on v-pred check (#537) stduhpf 2025-01-18 06:13:34 +0100
• 4fe83d52cf
  chore: fix CUDA on GitHub Action (#567) ag2s20150909 2025-01-18 13:12:26 +0800
• b70aaa672a
  chore: fix amd rocm build (#571) null-define 2025-01-18 13:11:39 +0800
• 27edb765a5
  chore: fix CI windows release artifacts (#532) idostyle 2025-01-18 06:09:22 +0100
• dcf91f9e0f chore: change SD_CUBLAS/SD_USE_CUBLAS to SD_CUDA/SD_USE_CUDA leejet 2024-12-28 13:27:51 +0800
• 348a54e34a
  feat: use pretty-progress for tensor loading (#516) stduhpf 2024-12-28 06:14:52 +0100
• d50473dc49
  feat: support 16 channel tae (taesd/taef1) (#527) stduhpf 2024-12-28 06:13:48 +0100
• b5cc1422da
  fix: fix typo for skip layers parameters (#492) piallai 2024-12-28 06:12:08 +0100
• 5cc74d1f09
  feat: support Moore Threads GPU (#529) R0CKSTAR 2024-12-28 13:08:36 +0800
• 0d9d6659a7
  fix: fix metal build (#513) stduhpf 2024-12-28 06:06:17 +0100
• 8f4ab9add3
  feat: support Inpaint models (#511) stduhpf 2024-12-28 06:04:49 +0100
• cc92a6a1b3
  feat: support more LoRA models (#520) stduhpf 2024-12-28 05:56:44 +0100
• 9578fdcc46 chore: remove rocm5.5 build temporarily leejet 2024-11-30 14:26:29 +0800
• 9148b980be
  feat: remove type restrictions (#489) stduhpf 2024-11-30 07:22:15 +0100
• 7ce63e740c
  feat: flexible model architecture for dit models (Flux & SD3) (#490) stduhpf 2024-11-30 07:18:53 +0100
• 4570715727 fix: use ggml_nn_attention in vae leejet 2024-11-24 18:21:31 +0800
• 53b415f787
  fix: remove default variables in c headers (#478) stduhpf 2024-11-24 11:10:25 +0100
• c3eeb669cd sync: update ggml leejet 2024-11-23 13:29:32 +0800
• b5f4932696 refactor: add some sd vesion helper functions leejet 2024-11-23 13:02:44 +0800
• 1c168d98a5
  fix: repair flash attention support (#386) Erik Scholz 2024-11-23 05:39:08 +0100
• ea9b647080
  docs: update readme, add python bindings (#423) William Murray 2024-11-23 14:52:33 +1100
• 2b1bc06477
  feat: add PhotoMaker Version 2 support (#358) bssrdf 2024-11-22 22:50:14 -0500
• b99cbfe4dc
  docs: update README.md (#452) Flavio Bizzarri 2024-11-23 04:46:50 +0100
• 8c7719fe9a
  fix: typo in clip-g encoder arg (#472) Plamen Minev 2024-11-23 05:46:00 +0200
• 8f94efafa3
  feat: add support for loading F8_E5M2 weights (#460) LostRuins Concedo 2024-11-23 11:45:11 +0800
• 07585448ad
  docs: update readme (#462) fszontagh 2024-11-23 04:42:12 +0100
• 6ea812256e
  feat: add flux 1 lite 8B (freepik) support (#474) stduhpf 2024-11-23 04:41:30 +0100
• 9b1d90bc23
  fix: improve clip text_projection support (#397) stduhpf 2024-11-23 04:19:27 +0100
• 65fa646684
  feat: add sd3.5 medium and skip layer guidance support (#451) stduhpf 2024-11-23 04:15:31 +0100
• ac54e00760
  feat: add sd3.5 support (#445) leejet 2024-10-24 21:58:03 +0800
• 14206fd488
  fix: fix clip tokenizer (#383) stduhpf 2024-09-02 16:31:46 +0200
• e410aeb534
  sync: update ggml to fix large image generation with SYCL backend (#380) zhentaoyu 2024-09-02 22:29:35 +0800
• 58d54738e2 docs: add star history leejet 2024-08-28 00:27:54 +0800
• 4f87b232c2 docs: add Vulkan build command leejet 2024-08-28 00:24:53 +0800
• e71ddcedad
  fix: improve VAE tiling (#372) Erik Scholz 2024-08-27 18:21:12 +0200
• f4c937cb94
  fix: add some missing cli args to usage (#363) stduhpf 2024-08-27 18:17:46 +0200
• 0362cc4874
  fix: fix some typos (#361) Daniele 2024-08-27 16:15:37 +0000
• 6c88ad3fd6
  fix: resolve naming conflict while llama.cpp and sd.cpp both build (#351) Yu Xing 2024-08-27 09:14:41 -0700
• dc0882cdc9
  feat: add exponential scheduler (#346) Daniele 2024-08-27 16:13:35 +0000
• d00c94844d
  feat: add ipndm and ipndm_v samplers (#344) Daniele 2024-08-27 16:03:41 +0000
• 2d4a2f7982
  feat: add GITS scheduler (#343) Daniele 2024-08-27 16:02:17 +0000
• 353ee93e2d
  fix: add enum type to sd_type_t (#293) Tim Miller 2024-08-28 00:57:24 +0900
• 2027b16fda
  feat: add vulkan backend support (#291) soham 2024-08-27 21:26:09 +0530
• 8847114abf fix: fix issue when applying lora leejet 2024-08-25 22:39:39 +0800
• 5c561eab31 feat: do not convert more flux tensors leejet 2024-08-25 16:01:36 +0800
• f5997a1951 fix: do not force using f32 for some flux layers leejet 2024-08-25 14:07:22 +0800
• 1bdc767aaf feat: force using f32 for some layers leejet 2024-08-25 13:53:16 +0800
• 79c9fe9556 feat: do not convert some tensors leejet 2024-08-25 13:37:37 +0800
• 28a614769a docs: update docs/flux.md leejet 2024-08-25 13:11:34 +0800
• c837c5d9cc style: format code leejet 2024-08-25 00:19:37 +0800
• d08d7fa632 docs: update README.md leejet 2024-08-24 14:38:44 +0800
• 64d231f384
  feat: add flux support (#356) leejet 2024-08-24 14:29:52 +0800
• 697d000f49
  feat: add SYCL Backend Support for Intel GPUs (#330) zhentaoyu 2024-08-10 13:42:50 +0800
• 5b8d16aa68 docs: reorganize README.md leejet 2024-08-03 12:06:34 +0800
• 3d854f7917 sync: update ggml submodule url leejet 2024-08-03 11:42:12 +0800
• 4a6e36edc5 sync: update ggml leejet 2024-07-28 18:30:35 +0800
• 73c2176648
  feat: add sd3 support (#298) leejet 2024-07-28 15:44:08 +0800
• 9c51d8787f
  chore: fix cuda CI (#286) Phu Tran 2024-06-12 08:13:24 -0700
• f9f0d4685b fix: sample_k_diffusion should be static leejet 2024-06-10 23:04:02 +0800
• 8d2050a5cf sync: update ggml leejet 2024-06-10 22:59:36 +0800
• 08f5b41956 refector: make the sampling module more independent leejet 2024-06-10 22:42:15 +0800
• b6daf5c55b
  fix: use PRI64 instead of %i for some log (#269) Eugene 2024-06-01 11:01:58 +0500
• be6cd1a4bf sync: update ggml leejet 2024-06-01 13:44:09 +0800
• e1384defca
  perf: make crc32 100x faster on x86-64 (#278) Justine Tunney 2024-05-31 21:58:30 -0700
• 814280343c
  chore: update artifact actions (#267) Phu Tran 2024-05-31 21:33:13 -0700
• 1d2af5ca3f fix: set n_dims of tensor storage to 1 when it's 0 leejet 2024-05-14 23:06:52 +0800
• ce1bcc74a6
  feat: add AYS(Align Your Steps) scheduler (#241) Grauho 2024-04-29 11:21:32 -0400
• 760cfaa618
  fix: ignore tensors with the particular dim while loading (#233) Eugene 2024-04-29 20:04:27 +0500
• 6d16f6853e
  fix: correct upscale progressbar (#232) Eugene 2024-04-29 19:59:46 +0500
• 036ba9e6d8 feat: enable controlnet and photo maker for img2img mode leejet 2024-04-14 16:36:08 +0800
• ec82d5279a refector: remove some useless code leejet 2024-04-14 14:04:52 +0800
• afea457eda
  fix: support more SDXL LoRA names (#216) bssrdf 2024-04-06 05:12:03 -0400