cohere and HF integration

app.py

@@ -1,11 +1,17 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
 import os
+import cohere
+from huggingface_hub import InferenceClient
+
+# Retrieve API keys from environment variables
+HF_API_KEY = os.getenv("HF_API_KEY")  
+COHERE_API_KEY = os.getenv("COHERE_API_KEY")  
 
-HF_API_KEY = os.getenv("HF_API_KEY")  # Retrieve API key from environment variable
+# Initialize clients
+hf_model = "meta-llama/Llama-3.2-3B-Instruct"  # Change to preferred HF model
+hf_client = InferenceClient(model=hf_model, token=HF_API_KEY)
 
-models = ["HuggingFaceH4/zephyr-7b-beta", "microsoft/Phi-4-mini-instruct", "meta-llama/Llama-3.2-3B-Instruct"]
-client = InferenceClient(model=models[2], token=HF_API_KEY)  # Pass API key to client
+cohere_client = cohere.Client(COHERE_API_KEY)
 
 
 def respond(
@@ -15,9 +21,10 @@ def respond(
     max_tokens,
     temperature,
     top_p,
+    use_cohere,  # Checkbox input
 ):
     messages = [{"role": "system", "content": system_message}]
-
+    
     for val in history:
         if val[0]:
             messages.append({"role": "user", "content": val[0]})
@@ -26,24 +33,34 @@ def respond(
 
     messages.append({"role": "user", "content": message})
 
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+    # 🔹 **Switch API based on checkbox**
+    if use_cohere:
+        response = cohere_client.chat(
+            model="command-r-plus",  
+            message=message,
+            chat_history=[{"user_name": "User", "text": h[0]} for h in history if h[0]] +
+                         [{"user_name": "Assistant", "text": h[1]} for h in history if h[1]],
+            temperature=temperature,
+            max_tokens=max_tokens,
+            p=top_p,
+        )
+        return response.text  # Cohere returns full response
 
-        response += token
-        yield response
+    else:
+        response = ""
+        for message in hf_client.chat_completion(
+            messages,
+            max_tokens=max_tokens,
+            stream=True,
+            temperature=temperature,
+            top_p=top_p,
+        ):
+            token = message.choices[0].delta.content
+            response += token
+            yield response  # Hugging Face supports streaming
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# 🔥 **Gradio UI with Checkbox**
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
@@ -57,9 +74,9 @@ demo = gr.ChatInterface(
             step=0.05,
             label="Top-p (nucleus sampling)",
         ),
+        gr.Checkbox(label="Use Cohere API", value=False),  # Checkbox to toggle API
     ],
 )
 
-
 if __name__ == "__main__":
     demo.launch()