You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardexpand all lines: README.md
+46-49
Original file line number
Diff line number
Diff line change
@@ -127,56 +127,53 @@ The majority of scripts is licensed under ASL 2.0 (including codes from Diffuser
127
127
128
128
## Change History
129
129
130
-
### 8 Apr. 2021, 2021/4/8:
131
-
132
-
- Added support for training with weighted captions. Thanks to AI-Casanova for the great contribution!
133
-
- Please refer to the PR for details: [PR #336](https://github.com/kohya-ss/sd-scripts/pull/336)
134
-
- Specify the `--weighted_captions` option. It is available for all training scripts except Textual Inversion and XTI.
135
-
- This option is also applicable to token strings of the DreamBooth method.
136
-
- The syntax for weighted captions is almost the same as the Web UI, and you can use things like `(abc)`, `[abc]`, and `(abc:1.23)`. Nesting is also possible.
137
-
- If you include a comma in the parentheses, the parentheses will not be properly matched in the prompt shuffle/dropout, so do not include a comma in the parentheses.
- There may be bugs because I changed a lot. If you cannot revert the script to the previous version when a problem occurs, please wait for the update for a while.
148
-
149
-
- Added a feature to upload model and state to HuggingFace. Thanks to ddPn08 for the contribution! [PR #348](https://github.com/kohya-ss/sd-scripts/pull/348)
150
-
- When `--huggingface_repo_id` is specified, the model is uploaded to HuggingFace at the same time as saving the model.
151
-
- Please note that the access token is handled with caution. Please refer to the [HuggingFace documentation](https://huggingface.co/docs/hub/security-tokens).
152
-
- For example, specify other arguments as follows.
153
-
-`--huggingface_repo_id "your-hf-name/your-model" --huggingface_path_in_repo "path" --huggingface_repo_type model --huggingface_repo_visibility private --huggingface_token hf_YourAccessTokenHere`
154
-
- If `public` is specified for `--huggingface_repo_visibility`, the repository will be public. If the option is omitted or `private` (or anything other than `public`) is specified, it will be private.
155
-
- If you specify `--save_state` and `--save_state_to_huggingface`, the state will also be uploaded.
156
-
- If you specify `--resume` and `--resume_from_huggingface`, the state will be downloaded from HuggingFace and resumed.
157
-
- In this case, the `--resume` option is `--resume {repo_id}/{path_in_repo}:{revision}:{repo_type}`. For example: `--resume_from_huggingface --resume your-hf-name/your-model/path/test-000002-state:main:model`
158
-
- If you specify `--async_upload`, the upload will be done asynchronously.
159
-
- Added the documentation for applying LoRA to generate with the standard pipeline of Diffusers. [training LoRA](./train_network_README-ja.md#diffusersのpipelineで生成する) (Japanese only)
160
-
- Support for Attention Couple and regional LoRA in `gen_img_diffusers.py`.
161
-
- If you use ` AND ` to separate the prompts, each sub-prompt is sequentially applied to LoRA. `--mask_path` is treated as a mask image. The number of sub-prompts and the number of LoRA must match.
- プロンプトを` AND `で区切ると各サブプロンプトが順にLoRAに適用されます。`--mask_path` がマスク画像として扱われます。サブプロンプトの数とLoRAの数は一致している必要があります。
130
+
### Naming of LoRA
179
131
132
+
The LoRA supported by `train_network.py` has been named to avoid confusion. The documentation has been updated. The following are the names of LoRA types in this repository.
133
+
134
+
1.__LoRA-LierLa__ : (LoRA for __Li__ n __e__ a __r____La__ yers)
135
+
136
+
LoRA for Linear layers and Conv2d layers with 1x1 kernel
137
+
138
+
2.__LoRA-C3Lier__ : (LoRA for __C__ olutional layers with __3__ x3 Kernel and __Li__ n __e__ a __r__ layers)
139
+
140
+
In addition to 1., LoRA for Conv2d layers with 3x3 kernel
141
+
142
+
LoRA-LierLa is the default LoRA type for `train_network.py` (without `conv_dim` network arg). LoRA-LierLa can be used with [our extension](https://github.com/kohya-ss/sd-webui-additional-networks) for AUTOMATIC1111's Web UI, or with the built-in LoRA feature of the Web UI.
143
+
144
+
To use LoRA-C3Liar with Web UI, please use our extension.
- Added support for DyLoRA in `train_network.py`. Please refer to [here](./train_network_README.md) for details (currently only in Japanese).
165
+
- Added support for caching latents to disk in each training script. Please specify __both__`--cache_latents` and `--cache_latents_to_disk` options.
166
+
- The files are saved in the same folder as the images with the extension `.npz`. If you specify the `--flip_aug` option, the files with `_flip.npz` will also be saved.
167
+
- Multi-GPU training has not been tested.
168
+
- This feature is not tested with all combinations of datasets and training scripts, so there may be bugs.
169
+
- Added workaround for an error that occurs when training with `fp16` or `bf16` in `fine_tune.py`.
0 commit comments