diff --git a/colab/GPU.ipynb b/colab/GPU.ipynb
index d5c24d77b..238446cf1 100644
--- a/colab/GPU.ipynb
+++ b/colab/GPU.ipynb
@@ -42,8 +42,9 @@
         "#@markdown Enables Silero text-to-speech module\n",
         "extras_enable_sd = True #@param {type:\"boolean\"}\n",
         "#@markdown Enables SD picture generation\n",
-        "SD_Model = \"ckpt/anything-v4.5-vae-swapped\" #@param [ \"ckpt/anything-v4.5-vae-swapped\", \"ckpt/sd15\" ]\n",
+        "SD_Model = \"ckpt/anything-v4.5-vae-swapped\" #@param [ \"ckpt/anything-v4.5-vae-swapped\", \"philz1337/clarity\", \"ckpt/sd15\" ]\n",
         "#@markdown * ckpt/anything-v4.5-vae-swapped - anime style model\n",
+        "#@markdown * philz1337/clarity - realistic style model\n",
         "#@markdown * ckpt/sd15 - base SD 1.5\n",
         "\n",
         "import subprocess\n",
diff --git a/public/notes/content.md b/public/notes/content.md
index c80e7ba45..edbf848d2 100644
--- a/public/notes/content.md
+++ b/public/notes/content.md
@@ -416,7 +416,53 @@ _When using Pygmalion models these anchors are automatically disabled, since Pyg
 
 ## Instruct Mode
 
-_This section is under construction. Please check later._
+Instruct Mode allows you to adjust the prompting for instruction-following models, such as Alpaca, Metharme, WizardLM, etc.
+
+**This is not supported for OpenAI API.**
+
+### Instruct Mode Settings
+
+#### System Prompt
+
+Added to the beginning of each prompt. Should define the instructions for the model to follow.
+
+For example:
+
+```
+Write one reply in internet RP style for {{char}}. Be verbose and creative.
+```
+
+#### Presets
+
+Provides ready-made presets with prompts and sequences for some well-known instruct models.
+
+*Changing a preset resets your system prompt to default!*
+
+#### Input Sequence
+
+Text added before the user's input.
+
+#### Output Sequence
+
+Text added before the character's reply.
+
+#### System Sequence
+
+Text added before the system prompt.
+
+#### Stop Sequence
+
+Text that denotes the end of the reply. Will be trimmed from the output text.
+
+#### Include Names
+
+If enabled, prepend character and user names to chat history logs after inserting the sequences.
+
+*Always enabled for group chats!*
+
+#### Wrap Sequences with Newline
+
+Each sequence text will be wrapped with newline characters when inserted to the prompt. Required for Alpaca and its derivatives.
 
 ## Chat import