Merge pull request #97 from anujchoudhary-17/documentation-added

Added documenttations for 1. Setting up LM Studio 2. Setting up Ollam…

docs/setting-up-lm-studio.md

@@ -0,0 +1,36 @@
+# Setting-up LM Studio with BrowserOS
+
+## Step 1: Open LM Studio
+
+Open LM Studio, navigate to the `Developer` page and select the desired model to load.
+
+![Open LM Studio](images/setting-up-lm-studio/lmstudio-step1.png)
+
+### Step 2: Load the model
+
+In below example I’m loading `openai/gpt-oss-20b` model
+
+![Load the Studio](images/setting-up-lm-studio/lmstudio-step2.png)
+
+### Step 3: Configure in BrowserOS
+
+- Navigate to `chrome://settings/browseros-ai` or `Go to Settings -> BrowserOS AI`
+- Click `Add Provider`
+- Select `OpenAl Compatible` in `Provider Type` dropdown
+
+![Configure in BrowserOS](images/setting-up-lm-studio/lmstudio-step3.png)
+
+### Step 4: Configure the Model
+
+- Replace the Base URL with LM Studio URL, default is [`http://localhost:1234/v1/`](http://localhost:1234/v1/)
+- Set the Model ID to the one you loaded in LM studio
+- **💥 IMPORTANT:** Ensure you set the `Context Window Size` to what you configured in LM Studio.
+
+![Configure the Model](images/setting-up-lm-studio/lmstudio-step4-1.png)
+![Configure the Model](images/setting-up-lm-studio/lmstudio-step4-2.png)
+
+### Step 5: Use the model in `BrowerOS` 🐶
+
+- Select the model in BrowserOS agent and start using! 🎉
+
+![Use the model in BrowerOS](images/setting-up-lm-studio/lmstudio-step5.png)

docs/setting-up-ollama.md

@@ -0,0 +1,38 @@
+# Setting-up Ollama with BrowserOS
+
+## Steps
+
+1. Navigate to `chrome://settings/browseros-ai` to add Ollama as a provider.
+2. Get the model ID of ollama model i.e. `gpt-oss:20b`
+3. You need to start ollama with this command to prevent CORS issue `OLLAMA_ORIGINS="*" ollama serve`
+    - 💥 If you don’t want to run from CLI, we recommend using LM studio. See the guide here - [Setting-up LM Studio with BrowserOS](setting-up-lm-studio.md)
+4. Select the model in agent and start using it! 🥳
+
+## Visual Steps
+
+### **Step 1: Navigate to settings page `chrome://settings/browseros-ai`**
+
+![Navigate to settings page](images/setting-up-ollama/ollama-step1.png)
+
+
+### **Step 2: Get the ollama model ID**
+
+![Get the ollama model ID](images/setting-up-ollama/ollama-step2.png)
+
+
+### **Step 3: Start Ollama from CLI**
+
+`OLLAMA_ORIGINS="*" ollama serve` 
+
+Unfortunately, Ollama by default doesn’t allow requests from other apps without this.
+
+💥 If you don’t want to run from CLI, we recommend using LM studio. See the guide here - [Setting-up LM Studio with BrowserOS](setting-up-lm-studio.md)
+
+![Start Ollama from CLI](images/setting-up-ollama/ollama-step3.png)
+
+
+### Step 4: Use the model
+
+Select the model in `Agent` drop-down and start using it 🚀
+
+![Use the model](images/setting-up-ollama/ollama-step4.png)

docs/using-lm-studio-openais-gpt-oss.md

@@ -0,0 +1,53 @@
+# Using LM Studio + OpenAI’s gpt-oss
+
+### STEP1: Setup LMStudio and download **OpenAI GPT-OSS**
+
+1. Download LMStudio from https://lmstudio.ai/ 
+    
+2. Click on Discover in LMStudio, it should be the “🔍” icon on the left
+
+![Setup LMStudio and download OpenAI GPT-OSS](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step1.png)
+
+3. Search for gpt-oss-20b and click Download
+
+![Search for gpt-oss-20b and click Download](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step2.png)
+
+4. After download finishes, load the model. 
+
+![After download finishes, load the model.](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step3.png)
+
+NOTE: Enable the flag to choose model parameters on load
+
+![Enable the flag to choose model parameters on load](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step4.png)
+
+Set context length to 32768, depending on your hardware and load the model
+
+![Set context length to 32768, depending on your hardware and load the model](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step5.png)
+
+### STEP2: Configure BrowserOS to use LMStudio
+
+1. Add Provider from `chrome://settings/browseros-ai`
+
+![Configure BrowserOS to use LMStudio](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step6.png)
+
+2. Choose Provider Type as “OpenAI Compatible”
+
+![Choose Provider Type as “OpenAI Compatible”](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step7.png)
+
+3. Set Base URL to http://127.0.0.1:1234/v1 and model ID to openai/gpt-oss-20b and set context length to 32768 and save!
+
+![Set Base URL-1](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step8.png)
+
+![Set Base URL-2](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step9.png)
+
+4. Change the default provider to lmstudio and you are good to go!
+
+![Change the default provider to lmstudio](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step10.png)
+
+5. You can gpt-oss from Agent!
+
+![You can gpt-oss from Agent](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step11.png)
+
+If everything is setup you should see messages on the LMStudio, like below.
+
+![If everything is setup you should see messages on the LMStudio](images/using-lm-studio-openais-gpt-oss/lm-studio-openais-step12.png)