storytold/FineTrainers-Conditioning
2
0
Fork 0
mirror of https://github.com/storytold/FineTrainers-Conditioning.git synced 2026-10-09 00:09:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • 6c2ba374ae Todo fix the issue with the conditioning we do not re adjust the values into the model. feature/cross_attn_patchify CrossProduct 2025-02-13 18:03:26 +00:00
  • 93e5069dc0 Simple conditioning sorta merges but checking inputs now. feature/simple-condition CrossProduct 2025-02-06 17:25:13 +00:00
  • d65958b181 Checking inputs. feature/condition_patchify CrossProduct 2025-02-06 04:45:30 +00:00
  • 3fe4de9356 Modifed the inference loop to try something. also added a copy of the configuration. CrossProduct 2025-02-04 23:03:42 +00:00
  • 28165e67d0 Adapter weights can be loaded. CrossProduct 2025-02-04 16:46:45 +00:00
  • ea3cd5c753 Full Finetune should work with this adapter removed the residual_x CrossProduct 2025-01-30 00:14:44 +00:00
  • e399684065 Make residuals none. CrossProduct 2025-01-30 00:07:39 +00:00
  • 09e3f3dd5f move residual arg. CrossProduct 2025-01-30 00:03:12 +00:00
  • be1cb10894 make residual optional. CrossProduct 2025-01-30 00:02:31 +00:00
  • 8fe0ed225a Conditioned Adapter make residual optional. CrossProduct 2025-01-29 23:49:22 +00:00
  • 41eec16c67 Renamed a few variables to reduce cog load. CrossProduct 2025-01-29 23:26:39 +00:00
  • e4a14af8d7 Add back in the pose for both channels. CrossProduct 2025-01-29 22:40:18 +00:00
  • bca44f6d7f Remove some comments. CrossProduct 2025-01-29 21:11:55 +00:00
  • 536dfc8f6a Figure out why classifier free guidence breaks the loop. I suspect it wasn't baked into pipeline should have been 0 CrossProduct 2025-01-29 06:10:29 +00:00
  • a16285141e Noted place that requires intervention. CrossProduct 2025-01-29 05:30:17 +00:00
  • 707e001f7b The encoder hidden states and encode attention mask sizes do not match up in the validation pipeline. CrossProduct 2025-01-29 05:24:06 +00:00
  • 4ab3b25c13 Make a fix to the video tensors. CrossProduct 2025-01-28 23:03:04 +00:00
  • 8ec35d58fb trying training again but broke something so going back. CrossProduct 2025-01-28 22:36:09 +00:00
  • dfa03f1ed1 Fix before trying again. CrossProduct 2025-01-27 08:30:45 +00:00
  • 80f653f336 Passed data in for validation. CrossProduct 2025-01-25 03:07:12 +00:00
  • e99d70ff46 Fixed some issues with imports and the original lora.py code. CrossProduct 2025-01-25 02:22:05 +00:00
  • 06e26ec092 test and trace the changes. CrossProduct 2025-01-24 22:58:44 +00:00
  • c6490c0d51 Reducing cogload on the framework. CrossProduct 2025-01-24 22:46:32 +00:00
  • 305a4a994d Need to pass in image ref video. CrossProduct 2025-01-24 05:21:53 +00:00
  • 440738e535 Add conditioned pipeline ... CrossProduct 2025-01-24 02:47:58 +00:00
  • 64898ef20d Test the conditioned pipeline and ensure it loads. CrossProduct 2025-01-24 02:47:23 +00:00
  • b7f26bd2e6 removed some code made things clearer to replace the pipeline inference which is needed for validation. CrossProduct 2025-01-23 23:49:24 +00:00
  • d6b3615421 Ok adapter trains need to get the pass for validation using a skeleton. CrossProduct 2025-01-22 22:07:52 +00:00
  • d631737329 Post patchify the latents. CrossProduct 2025-01-22 21:24:33 +00:00
  • 731f326e86 Added residual CrossProduct 2025-01-22 20:10:01 +00:00
  • 981767154b modify the forward pass CrossProduct 2025-01-22 06:56:22 +00:00
  • 07e755c2b3 modifcations to the loading of the weights. CrossProduct 2025-01-22 01:26:49 +00:00
  • 9fde6600dc Fixing weights copy. CrossProduct 2025-01-22 01:26:17 +00:00
  • 34996c1eb1 replaced the pipeline functions. time to test this then test the weight changes. CrossProduct 2025-01-21 02:22:35 +00:00
  • 48d2b61029 Need to test saving and loading. CrossProduct 2025-01-18 06:58:04 +00:00
  • 3c605a124e Adding in the conditioned transformer 3d model CrossProduct 2025-01-18 06:47:39 +00:00
  • fa422c4ca5 Adding conditioned transformer layer change. CrossProduct 2025-01-18 06:29:01 +00:00
  • 2720d24f62 Moved the norm back to prepare latents step. Now its time to add the adapter layer. CrossProduct 2025-01-18 04:18:25 +00:00
  • 44df9e3bce reformulate the conditioning and patchification CrossProduct 2025-01-18 03:11:09 +00:00
  • 59afd55775 Retested changes without conditioning it works. CrossProduct 2025-01-17 22:43:20 +00:00
  • b73d61ae4d saving more changes related to args. CrossProduct 2025-01-17 06:40:06 +00:00
  • 47356cd864 Testing the original code path. Need to check rectified linear flow vs flow matching. CrossProduct 2025-01-17 06:03:45 +00:00
  • cd66e6e58e fixed the training pipeline to address new changes from upstream CrossProduct 2025-01-17 01:46:35 +00:00
  • 315431be0a Recommit the conditioning code. CrossProduct 2025-01-16 05:03:44 +00:00
  • b341b79652 Adding Conditioning Blocks to the Diffusion Transformer. CrossProduct 2025-01-15 22:29:47 +00:00
  • f34b8659c3 Patchification research to condition the model CrossProduct 2025-01-14 18:14:53 +00:00
  • 9f0fdb5055 Rename lora files (#214) main Aryan 2025-01-14 01:17:42 +05:30
  • 354a6b9e25 reposition private methods (#213) Aryan 2025-01-14 00:08:04 +05:30
  • e5df80cc36 Full Finetuning for LTX possibily extended to other models. (#192) ArEnSc 2025-01-13 13:25:39 -05:00
  • eb6b0f596a There was a results_args mapping that needed to be modified. feature/full-finetuning CrossProduct 2025-01-12 05:09:42 +00:00
  • 4354fb3164 LTX uses a default frame rate of 24 FPS We need to modify the output validation framerate to match that value. Add Framerate args. Add Update video output and inference frame rate CrossProduct 2025-01-12 04:39:05 +00:00
  • a78fedd108 Merge branch 'main' into feature/full-finetuning Aryan 2025-01-11 07:17:29 +05:30
  • a3898075d1 Helpful mesages (#210) Aryan 2025-01-11 06:35:28 +05:30
  • d3b797ee35 Improve argument handling (#209) Aryan 2025-01-11 06:18:13 +05:30
  • ca957e543c remove unnecessary changes Aryan 2025-01-11 01:36:52 +01:00
  • 491b35f3dd Merge branch 'main' into feature/full-finetuning Aryan 2025-01-11 05:59:08 +05:30
  • 1f304b36d4 handle saving of final model weights correctly Aryan 2025-01-11 01:28:47 +01:00
  • 06dd96c0a3 remove unnecessary changes Aryan 2025-01-11 01:11:27 +01:00
  • d6821c3da8 refactor Aryan 2025-01-11 00:54:03 +01:00
  • 81bf590a91 Argument descriptions (#208) Aryan 2025-01-11 05:06:02 +05:30
  • 014960fd4c Merge branch 'main' into feature/full-finetuning Sayak Paul 2025-01-10 19:56:34 +05:30
  • b34091362e Fix the checkpoint dir bug in get_intermediate_ckpt_path so that checkpointing_limit can actually work. (#207) Boyang Deng 2025-01-09 23:48:32 -08:00
  • 34da4c5a6b Merge branch 'main' into feature/full-finetuning Sayak Paul 2025-01-10 07:54:05 +05:30
  • 4b21d37c9d Update pr_tests.yml to update ruff version (#205) Sayak Paul 2025-01-10 07:53:39 +05:30
  • a422f7f07b updates sayakpaul 2025-01-10 07:53:16 +05:30
  • 6d597699b3 Merge branch 'main' into feature/full-finetuning Sayak Paul 2025-01-10 07:49:06 +05:30
  • c0f38892f8 remove __class__. sayakpaul 2025-01-10 07:44:36 +05:30
  • 28b3e845d8 Merge branch 'main' into auxiliary-support-ff-2 Sayak Paul 2025-01-10 07:43:02 +05:30
  • 9523a05220 Shell script to minimally test supported models on a real dataset (#204) Sayak Paul 2025-01-10 07:22:00 +05:30
  • 2c6f5497a5 Merge branch 'main' into auxiliary-support-ff-2 sayakpaul 2025-01-09 14:26:32 +05:30
  • b9a6492ad1 Epoch loss (#201) Aryan 2025-01-09 14:22:56 +05:30
  • 162e6cd50d fix subfolder bug. sayakpaul 2025-01-09 13:25:01 +05:30
  • 5183405e61 tackle validation. sayakpaul 2025-01-09 13:17:15 +05:30
  • 8188f8a11e tackle saving and loading hooks. sayakpaul 2025-01-09 13:03:14 +05:30
  • acffc2d9fc parse subfolder when needed. sayakpaul 2025-01-09 12:54:57 +05:30
  • f311f16d98 [core] Fix loading of precomputed conditions and latents (#199) Sayak Paul 2025-01-09 12:43:03 +05:30
  • 6716ccf8c2 update (#198) Aryan 2025-01-09 11:06:34 +05:30
  • 19bba0aed3 Run linter. CrossProduct 2025-01-09 00:39:20 +00:00
  • d0ee9c3101 Switching to Full FineTuning CrossProduct 2025-01-09 00:27:15 +00:00
  • 72ec207708 Merge branch 'main' into feature/full-finetuning Sayak Paul 2025-01-08 10:49:13 +05:30
  • dbffc80e7c Precomputation folder name based on model name (#196) Aryan 2025-01-08 08:45:44 +05:30
  • bd02e6d9d2 Merge branch 'main' into feature/full-finetuning ArEnSc 2025-01-07 19:14:25 -05:00
  • 905c22edd8 Fix minor bug with function call that doesn't exist. (#195) ArEnSc 2025-01-07 18:55:27 -05:00
  • 5bd5bc4a59 Fix minor bug with function call that doesn't exist. bugfix/call-to-non-existent-function CrossProduct 2025-01-07 23:32:06 +00:00
  • cb9381b40e Suggestions Addressed Renamed to SFT Added 2 other models. Testing required. CrossProduct 2025-01-07 22:59:09 +00:00
  • f1b02964d7 Update requirements (#189) Aryan 2025-01-08 04:14:55 +05:30
  • 4cd5a8e53b Used disable grad for component on lora fine tuning enabled CrossProduct 2025-01-07 20:41:12 +00:00
  • f0db0ccb00 Change name of the flag CrossProduct 2025-01-07 20:36:03 +00:00
  • eeea82ad04 Full Finetuning for LTX possibily extended to other models. CrossProduct 2025-01-07 20:27:19 +00:00
  • 7b569daf55 simplify docs part ii (#190) Sayak Paul 2025-01-07 12:47:01 +05:30
  • 38413aa167 Allow images; Remove LLM generated prefixes; Allow JSON/JSONL; Fix bugs (#158) Aryan 2025-01-06 18:52:54 +05:30
  • eeb4dd7afa [docs] refactor docs for easier info parsing (#175) Sayak Paul 2025-01-05 20:50:30 +05:30
  • b3d0ba8e12 support model cards (#176) Sayak Paul 2025-01-05 19:21:17 +05:30
  • b5a5124cd5 add a workflow to do quality checks. (#180) Sayak Paul 2025-01-05 18:07:49 +05:30
  • a709a1cec1 fixes (#178) Sayak Paul 2025-01-04 08:57:41 +05:30
  • 071c140879 Fix scheduler bugs (#177) Sayak Paul 2025-01-03 18:52:40 +05:30
  • b8352abf70 Support CogVideoX T2V (#165) Sayak Paul 2025-01-03 10:53:16 +05:30
  • 2a1aa05191 Update README.md to fix graph paths (#171) Sayak Paul 2025-01-02 20:11:27 +05:30
  • 018953ed86 bulk update styling and formatting (#170) Sayak Paul 2025-01-02 13:20:33 +05:30
  • 191ea8b0cc [optimization] support 8bit optims from bistandbytes (#163) Sayak Paul 2024-12-30 14:03:56 +05:30