bugfix

api/__init__.py

@@ -742,7 +742,7 @@ class GPT4AllAPI():
             self.generating = True            
             self.process.generate(self.discussion_messages, self.current_message, message_id, n_predict = self.config['n_predict'])
             self.process.started_queue.get()
-            while(self.process.is_generating.value):  # Simulating other commands being issued
+            while(self.process.is_generating.value or not self.process.generation_queue.empty()):  # Simulating other commands being issued
                 try:
                     chunk, tok, message_type = self.process.generation_queue.get(False, 2)
                     if chunk!="":

docs/youtube/script_personalities.md

@@ -99,3 +99,45 @@ Now, in the scripts folder, there are two files. The first one is install.py, wh
 The second file, processor.py, is where the personality's magic happens. It's got a class called Processor that inherits from PAPScript and does some important stuff. The star of the show is the run_workflow method. It takes the current prompt, the accumulated messages so far, and a callback function that's triggered at each step of the workflow.
 
 In this art personality, we've set it up to generate an image description based on the user's prompt. Then it uses the stable diffusion generator to bring that description to life as an actual image. Finally, it prepares and sends the image as a cool markdown text to the UI.
+
+Alright, let's get started with this cool personality! First, open up the UI and head over to the settings tab. Now, look for the section called "Personalities Zoo." It's where all the personalities are organized by language and category.
+
+By default, the personality is set to "GPT for All," but we're going to change that. Find the "Art" category, and within it, select the "GPT for Art" personality. Give it a moment to sink in.
+
+Now, this is important—click on "Apply Changes" to make sure the new settings take effect. Since this is the first time we're running this personality, it might take a bit longer. It needs to install some stuff, like the stable diffusion system and the model itself. So, just hang tight and be patient while everything gets set up.
+
+Once the installation is complete, you'll be all set to explore the artistic wonders of this personality.
+
+Just a heads up, if you decide to reboot the application now, all the updates to your settings will be lost. So, make sure you save those changes before you go ahead.
+
+
+Alright, let's head back to the discussions tab and start a fresh discussion. This is where we can really put the tool to the test!
+
+To challenge its creative abilities, let's ask it to imagine something cool without providing too much information. We want to see how it can come up with an image all on its own. Give it some freedom to work its magic and let's see what it creates.
+
+Here is the output generated by our artbot. We can see the description of the artwork as well as the artwork itself.
+
+Since this is a discussion, we can go on and ask for changes. For example let's ask it to change this to a day vew.
+
+As we can see, the model did the requested changes. 
+
+
+Keep in mind that this tool is still in its early stages of development, and there's plenty of room for improvement. One way to enhance its performance is by adjusting the default sampler to an Euler sampler, which can potentially yield even better results. Additionally, you have the flexibility to explore a wide range of models available on Hugging Face repositories. With thousands of models at your disposal, you can experiment and choose the one that aligns best with your specific needs and preferences. By making these adjustments, you can take this tool to new heights and unlock its full potential.
+
+Please note that all the generated images bear a watermark with the GPT4All signature, serving as a clear indication that they were created by AI using the stable diffusion WatermarkEncoder. This step is crucial to promote responsible AI usage and ensure that each generated work is properly identified as an AI creation.
+
+It's important to emphasize that this tool is intended for appreciating art, fostering creative exploration, and sparking new ideas. It is not meant for malicious purposes or spreading misinformation. We firmly stand against such misuse.
+
+By incorporating the watermark and setting these ethical guidelines, we aim to promote transparency and accountability in AI-generated content, while encouraging the responsible and respectful use of this tool. Let's embrace the power of AI to ignite imagination and appreciate the beauty of art in a responsible manner.
+
+Now, let's dive into the structure of the personality folder once it's installed. Take a peek at what's inside, and you'll notice a couple of new files that have appeared.
+
+First up, we have the .install file. It's a simple empty file that serves as a marker to indicate that the personality has been successfully installed. Whenever the personality is loaded, the install script checks for the presence of this file. If it's missing, the script knows that the personality needs to be installed again. So, if you ever remove this file, don't worry! The next time you use this personality, it will reinstall itself automatically.
+
+Next on the list is the config_local.yaml file. This nifty little file holds all the local configuration settings. You have the power to customize various aspects such as the maximum size of the generated image description text. Just tinker with the settings to your liking. Additionally, you can specify the model you want to use. Simply download a stable diffusion model and place it in the project's root directory under "shared/sd_models." Then, reference the model's name in the config_local.yaml file. Cool, right?
+
+Oh, and let's not forget! You can even set the number of images to generate for each prompt. Fancy playing around with multiple images? This is your chance. And if you want consistency, you can set a specific seed value for the sampling noise generation. A seed value of -1 means randomness, while any other value ensures that all generated images will be the same for each prompt.
+
+Feel free to explore and make tweaks according to your preferences. It's your playground now!
+
+