added openai support

README.md

@@ -119,6 +119,26 @@ python Web-LLM.py
 
 The LLM settings can be modified in `llm_config.py`. You must specify your model name in the configuration for the researcher to function. The default configuration is optimized for research tasks with the specified Phi-3 model.
 
+
+## OpenAI Configuration
+
+To use OpenAI models:
+
+1. Set your OpenAI API key either:
+   - In `llm_config.py`: Add your API key to `LLM_CONFIG_OPENAI["api_key"]`
+   - Or as an environment variable: `export OPENAI_API_KEY='your-api-key'`
+
+2. Change the LLM_TYPE in `llm_config.py`:
+   ```python
+   LLM_TYPE = "openai"  # Change this line
+   ```
+
+3. Optional: Modify the model name in `LLM_CONFIG_OPENAI`:
+   ```python
+   "model_name": "gpt-4o-mini"  # Or another OpenAI model
+   ```
+
+
 ## Current Status
 This is a prototype that demonstrates functional automated research capabilities. While still in development, it successfully performs structured research tasks. Currently tested and working well with the phi3:3.8b-mini-128k-instruct model when the context is set as advised previously.
 

llm_config.py

@@ -1,29 +1,29 @@
 # llm_config.py
 
-LLM_TYPE = "ollama"  # Options: 'llama_cpp', 'ollama'
+LLM_TYPE = "openai"  # Options: 'llama_cpp', 'ollama', 'openai'
 
 # LLM settings for llama_cpp
-MODEL_PATH = "/home/james/llama.cpp/models/gemma-2-9b-it-Q6_K.gguf" # Replace with your llama.cpp models filepath
+MODEL_PATH = "/home/james/llama.cpp/models/gemma-2-9b-it-Q6_K.gguf"
 
 LLM_CONFIG_LLAMA_CPP = {
     "llm_type": "llama_cpp",
     "model_path": MODEL_PATH,
-    "n_ctx": 20000,  # context size
-    "n_gpu_layers": 0,  # number of layers to offload to GPU (-1 for all, 0 for none)
-    "n_threads": 8,  # number of threads to use
-    "temperature": 0.7,  # temperature for sampling
-    "top_p": 0.9,  # top p for sampling
-    "top_k": 40,  # top k for sampling
-    "repeat_penalty": 1.1,  # repeat penalty
-    "max_tokens": 1024,  # max tokens to generate
-    "stop": ["User:", "\n\n"]  # stop sequences
+    "n_ctx": 20000,
+    "n_gpu_layers": 0,
+    "n_threads": 8,
+    "temperature": 0.7,
+    "top_p": 0.9,
+    "top_k": 40,
+    "repeat_penalty": 1.1,
+    "max_tokens": 1024,
+    "stop": ["User:", "\n\n"]
 }
 
 # LLM settings for Ollama
 LLM_CONFIG_OLLAMA = {
     "llm_type": "ollama",
-    "base_url": "http://localhost:11434",  # default Ollama server URL
-    "model_name": "custom-phi3-32k-Q4_K_M",  # Replace with your Ollama model name
+    "base_url": "http://localhost:11434",
+    "model_name": "custom-phi3-32k-Q4_K_M",
     "temperature": 0.7,
     "top_p": 0.9,
     "n_ctx": 55000,
@@ -31,10 +31,24 @@ LLM_CONFIG_OLLAMA = {
     "stop": ["User:", "\n\n"]
 }
 
+# New: LLM settings for OpenAI
+LLM_CONFIG_OPENAI = {
+    "llm_type": "openai",
+    "model_name": "gpt-4o-mini",  
+    "api_key": "",
+    "temperature": 0.7,
+    "top_p": 0.9,
+    "max_tokens": 4096,
+    "stop": ["User:", "\n\n"],
+    "context_length": 128000  # GPT-4 Turbo context window
+}
+
 def get_llm_config():
     if LLM_TYPE == "llama_cpp":
         return LLM_CONFIG_LLAMA_CPP
     elif LLM_TYPE == "ollama":
         return LLM_CONFIG_OLLAMA
+    elif LLM_TYPE == "openai":
+        return LLM_CONFIG_OPENAI
     else:
         raise ValueError(f"Invalid LLM_TYPE: {LLM_TYPE}")

llm_wrapper.py

@@ -1,20 +1,60 @@
 from llama_cpp import Llama
 import requests
 import json
+import os
 from llm_config import get_llm_config
+from openai import OpenAI
 
 class LLMWrapper:
     def __init__(self):
         self.llm_config = get_llm_config()
         self.llm_type = self.llm_config.get('llm_type', 'llama_cpp')
+        
         if self.llm_type == 'llama_cpp':
             self.llm = self._initialize_llama_cpp()
         elif self.llm_type == 'ollama':
             self.base_url = self.llm_config.get('base_url', 'http://localhost:11434')
             self.model_name = self.llm_config.get('model_name', 'your_model_name')
+        elif self.llm_type == 'openai':
+            self._initialize_openai()
         else:
             raise ValueError(f"Unsupported LLM type: {self.llm_type}")
 
+    def _initialize_openai(self):
+        """Initialize OpenAI client"""
+        api_key = self.llm_config.get('api_key') or os.getenv('OPENAI_API_KEY')
+        if not api_key:
+            raise ValueError("OpenAI API key not found. Set it in config or OPENAI_API_KEY environment variable")
+        self.client = OpenAI(api_key=api_key)
+        self.model_name = self.llm_config.get('model_name', 'gpt-4o-mini')
+
+    def generate(self, prompt, **kwargs):
+        if self.llm_type == 'llama_cpp':
+            llama_kwargs = self._prepare_llama_kwargs(kwargs)
+            response = self.llm(prompt, **llama_kwargs)
+            return response['choices'][0]['text'].strip()
+        elif self.llm_type == 'ollama':
+            return self._ollama_generate(prompt, **kwargs)
+        elif self.llm_type == 'openai':
+            return self._openai_generate(prompt, **kwargs)
+        else:
+            raise ValueError(f"Unsupported LLM type: {self.llm_type}")
+
+    def _openai_generate(self, prompt, **kwargs):
+        """Generate text using OpenAI API"""
+        try:
+            response = self.client.chat.completions.create(
+                model=self.model_name,
+                messages=[{"role": "user", "content": prompt}],
+                temperature=kwargs.get('temperature', self.llm_config.get('temperature', 0.7)),
+                max_tokens=kwargs.get('max_tokens', self.llm_config.get('max_tokens', 4096)),
+                top_p=kwargs.get('top_p', self.llm_config.get('top_p', 0.9)),
+                stop=kwargs.get('stop', self.llm_config.get('stop', None))
+            )
+            return response.choices[0].message.content.strip()
+        except Exception as e:
+            raise Exception(f"OpenAI API error: {str(e)}")
+
     def _initialize_llama_cpp(self):
         return Llama(
             model_path=self.llm_config.get('model_path'),
@@ -24,16 +64,6 @@ class LLMWrapper:
             verbose=False
         )
 
-    def generate(self, prompt, **kwargs):
-        if self.llm_type == 'llama_cpp':
-            llama_kwargs = self._prepare_llama_kwargs(kwargs)
-            response = self.llm(prompt, **llama_kwargs)
-            return response['choices'][0]['text'].strip()
-        elif self.llm_type == 'ollama':
-            return self._ollama_generate(prompt, **kwargs)
-        else:
-            raise ValueError(f"Unsupported LLM type: {self.llm_type}")
-
     def _ollama_generate(self, prompt, **kwargs):
         url = f"{self.base_url}/api/generate"
         data = {

requirements.txt

@@ -9,3 +9,4 @@ keyboard
 curses-windows; sys_platform == 'win32'
 tqdm
 urllib3
+openai