diff --git a/assets/RVC/Cloud/EasyGUI.md b/assets/RVC/Cloud/EasyGUI.md index 367f46f7..e170b377 100644 --- a/assets/RVC/Cloud/EasyGUI.md +++ b/assets/RVC/Cloud/EasyGUI.md @@ -3,7 +3,7 @@ icon: chevron-right order: 5000 visibility: private --- -``Last update: Mar 5, 2024`` +``Last update: Mar 6, 2024`` ‎ *** @@ -13,21 +13,20 @@ visibility: private - Its strongest points are its active maintenance & that it holds the simplicity of Mainline. -- The Colab offers two ways to use it, with or without the UI. The latter being available **only** for Pro tier users. +- The Colab offers two ways to use it, with or without the UI. The first one being ideal for newbies. ‎ #### Pros & Cons :icon-tasklist: ==- *Learn more* !!! *The pros & cons are subjective to your necessities.* !!! ||| ✔️ **PROS** -- Very complete - Active maintenance - UI version +- Automatic model upload - Feature to save model to HF ||| ❌ **CONS** - Has less features - No [Mangio-Crepe](https://aihubdocs.github.io/en/rvc/resources/inference-settings/#pitch-extraction-algorithm) -- Manual model upload - Usage limit for free users ||| === @@ -57,13 +56,181 @@ b. Then **Log in** to your Google account. ‎ image‎ ‎ -!!! Connecting to GD is not mandatory, but highly recommended. -As inputting files through Colab is much slower. + +*** +###### ‎ +#### 4. Open UI (optional) +- +- If you are going to train models, ensure load models from drive & open tensorboard are ticked + +- Then run the cell. + +image‎ + +- Open the public url. + +image‎ + +*** +!!! +Connecting to GD is not mandatory, but highly recommended, as submitting files through Colab is much slower. +Don't close Colab until you're done using EasyGUI, or it will stop working. !!! ###### ‎ *** -## Training :icon-rocket: +## With UI +*** + +### Inference :icon-unmute: +*** +#### 1. Download model +a. Go to the **Download Model** tab & paste the [model link]() in the upper bar. + +image‎ + +b. In **Name your model** insert a name for it. Don't include spaces/special characters. Then press ``Download``. + +image‎ +*** + +#### 2. Select model +a. Return to the **Model Inference** tab & click `Refresh voice list and index path`. + +b. Then select the model in **Inferencing voice**. + +image‎ + +*** +#### 3. Input vocals +a. Click the upload box & select your vocals. Or simply drag & drop. + +image‎ + +b. Once done uploading, click `Refresh` again. +*** +#### 4. Modify settings (optional) +- If you wish, modify the [inference settings]() for better results. Unfold the **General settings** menu to see more. + +!!! +The **Index Rate** is in the **Auto-detect index path** bar, and **Pitch** is on the right of `Refresh`. +!!! + +*** +#### 5. Convert +a. Begin processing by clicking the `Convert` button on top. + +b. Once done, you'll be able to hear the results in the **Export audio** box. To download it, click the 3 dots on the right & `Download`. + +image‎ + +*** +### Training :icon-rocket: + +!!!warning +**For newbies:** if a setting isn't mentioned, leave it untouched. +!!! +#### 1. Name the model +- Go to the **Train** tab. First set a name for your model in **Enter the experiment name**. + +image‎ + +*** + +#### 2. Target sample rate +- Select your dataset's [sample rate](). + +image‎ + +*** +#### 3. Select dataset + +a. Press the upload box & select your dataset. Or simply drag & drop. + +image‎ + +b. Once it's done uploading, press **Process Data**. It'll finish when the logs say ``Process data``. + +image‎ + +*** +#### 4. Feature extraction +- Press `Feature extraction` + +*** +#### 5. Total training epochs +- Input the total amount of [epochs](https://aihubdocs.github.io/en/rvc/resources/epochs-overtraining--tensorboard/) (training cycles) for the model. + +- But since we'll use [TensorBoard](https://aihubdocs.github.io/en/rvc/resources/epochs-overtraining--tensorboard/#tensorboard), use an arbitrarily large value like `1000` + +image‎ + +*** +## Without UI +*** + +###### ‎ +*** +### Inference :icon-unmute: +*** +#### 1. Input model +- Depending on your device, the process of inputting the model varies. + + +++ :icon-device-desktop: ‎ DESKTOP + a. Upload the model to Google Drive. + ‎ + image‎ + ‎ + b. Click the folder symbol ( icon-file-directory ) & go to **drive -> MyDrive**. + + image‎ + ‎ + c. Drag & drop the .PTH to the folder **RVC -> assets -> weights**. + ‎ + image‎ + ‎ + d. In **model_name** input the name of the PTH followed by `.pth`. + ‎ + image‎ + +++ :icon-device-mobile: MOBILE + a. Upload the model's INDEX to Google Drive. + +++ +*** +###### ‎ +#### 2. Upload INDEX +a. Open the file explorer, open **MyDrive**, right-click the INDEX & click **Copy path**. + + image‎ +‎ +b. Then paste it in the **index_path** bar. + + image‎ +*** +###### ‎ +#### 3. Input vocals +a. Upload the vocals to GD & copy the path. + + image‎ +‎ +b. Paste it in the **input_path** & **output_path** bars. + + image‎ + +*** +###### ‎ +#### 4. Modify settings (optional) +- Modify the [inference settings](https://aihubdocs.github.io/en/rvc/resources/inference-settings/) for better results if you wish. + + +*** +###### ‎ +#### 5. Convert +a. Run the cell to begin processing the audio. + +b. Once done, a playable audio will + + +### Training :icon-rocket: *** ==- **STEP 1** ###### ‎ @@ -154,65 +321,10 @@ iii. If after around 2:30 hours of training you don't detect [OT](https://aih - Run out of [GPU runtime](http://aihubdocs.github.io/en/extra/glossary/#google-colab). !!! === -###### ‎ -*** -## Inference :icon-unmute: -*** -#### 1. Input model -- Depending on your device, the process of inputting the model varies. - - +++ :icon-device-desktop: ‎ DESKTOP - a. Upload the model to Google Drive. - ‎ - image‎ - ‎ - b. Click the folder symbol ( icon-file-directory ) & go to **drive -> MyDrive**. - - image‎ - ‎ - c. Drag & drop the .PTH to the folder **RVC -> assets -> weights**. - ‎ - image‎ - ‎ - d. In **model_name** input the name of the PTH followed by `.pth`. - ‎ - image‎ - +++ :icon-device-mobile: MOBILE - a. Upload the model's INDEX to Google Drive. - +++ -*** -###### ‎ -#### 2. Upload INDEX -a. Open the file explorer, open **MyDrive**, right-click the INDEX & click **Copy path**. - - image‎ -‎ -b. Then paste it in the **index_path** bar. - - image‎ -*** -###### ‎ -#### 3. Input vocals -a. Upload the vocals to GD & copy the path. - - image‎ -‎ -b. Paste it in the **input_path** & **output_path** bars. - - image‎ - -*** -###### ‎ -#### 4. Modify settings (optional) -- Modify the [inference settings](https://aihubdocs.github.io/en/rvc/resources/inference-settings/) for better results if you wish. -*** -###### ‎ -#### 5. Convert -a. Run the cell to begin processing the audio. - -b. Once done, a playable audio will *** ## Troubleshooting :icon-tools: *** + +https://paypal.me/lesantillan \ No newline at end of file diff --git a/assets/RVC/Cloud/easygui-img/2-openui.png b/assets/RVC/Cloud/easygui-img/2-openui.png new file mode 100644 index 00000000..b77c9622 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/2-openui.png differ diff --git a/assets/RVC/Cloud/easygui-img/2-publicurl.png b/assets/RVC/Cloud/easygui-img/2-publicurl.png new file mode 100644 index 00000000..654fd6d9 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/2-publicurl.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-i-modellink.png b/assets/RVC/Cloud/easygui-img/ui-i-modellink.png new file mode 100644 index 00000000..e0060e45 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-i-modellink.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-i-namemodel.png b/assets/RVC/Cloud/easygui-img/ui-i-namemodel.png new file mode 100644 index 00000000..305cf772 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-i-namemodel.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-i-output.png b/assets/RVC/Cloud/easygui-img/ui-i-output.png new file mode 100644 index 00000000..981d3647 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-i-output.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-i-refreshselect.png b/assets/RVC/Cloud/easygui-img/ui-i-refreshselect.png new file mode 100644 index 00000000..a9bfdda9 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-i-refreshselect.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-i-uploadaudio.png b/assets/RVC/Cloud/easygui-img/ui-i-uploadaudio.png new file mode 100644 index 00000000..28cfe71a Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-i-uploadaudio.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-t-endpreprocess.png b/assets/RVC/Cloud/easygui-img/ui-t-endpreprocess.png new file mode 100644 index 00000000..7522939e Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-t-endpreprocess.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-t-epochs.png b/assets/RVC/Cloud/easygui-img/ui-t-epochs.png new file mode 100644 index 00000000..d3203b42 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-t-epochs.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-t-feature.png b/assets/RVC/Cloud/easygui-img/ui-t-feature.png new file mode 100644 index 00000000..92e827b2 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-t-feature.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-t-inputds.png b/assets/RVC/Cloud/easygui-img/ui-t-inputds.png new file mode 100644 index 00000000..6b302bad Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-t-inputds.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-t-name.png b/assets/RVC/Cloud/easygui-img/ui-t-name.png new file mode 100644 index 00000000..07ba4c01 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-t-name.png differ diff --git a/assets/RVC/Cloud/easygui-img/ui-t-samplerate.png b/assets/RVC/Cloud/easygui-img/ui-t-samplerate.png new file mode 100644 index 00000000..db6e91d9 Binary files /dev/null and b/assets/RVC/Cloud/easygui-img/ui-t-samplerate.png differ