Updated the inference endpoints

HF got rid of previous endpoints. I updated with the new serverless endpoints.

app.py

@@ -15,25 +15,35 @@ load_dotenv()
 #Comment_test_11_09_2024
 
 
-# initialize the client
-client = OpenAI(
-  base_url="https://api-inference.huggingface.co/v1",
-  api_key=os.environ.get('HUGGINGFACEHUB_API_TOKEN')#"hf_xxx" # Replace with your token
-) 
-
 
 
 
-#Create supported models
 model_links ={
-    "Meta-Llama-3-8B":"meta-llama/Meta-Llama-3-8B-Instruct", 
-    "Mistral-7B":"mistralai/Mistral-7B-Instruct-v0.2",
-    "Gemma-7B":"google/gemma-1.1-7b-it",
-    "Gemma-2B":"google/gemma-1.1-2b-it",
-    "Zephyr-7B-β":"HuggingFaceH4/zephyr-7b-beta",
-    #"Meta-Llama-3.1-8B":"meta-llama/Meta-Llama-3.1-8B-Instruct", #TODO: Update when/if Serverless Inference available
-
-}
+      "Gemma-3-27B-it":{
+                      "inf_point":"https://router.huggingface.co/nebius/v1",
+                      "link":"google/gemma-3-27b-it-fast",
+                      },
+      "Meta-Llama-3.1-8B":{
+                      "inf_point":"https://router.huggingface.co/nebius/v1",
+                      "link":"meta-llama/Meta-Llama-3.1-8B-Instruct-fast",
+                      },
+      "Mistral-7B":{
+                      "inf_point":"https://router.huggingface.co/together/v1",
+                      "link":"mistralai/Mistral-7B-Instruct-v0.3",
+                      },
+      "Gemma-2-27B-it":{
+                      "inf_point":"https://router.huggingface.co/nebius/v1",
+                      "link":"google/gemma-2-27b-it-fast",
+                      },
+      "Gemma-2-2B-it":{
+                      "inf_point":"https://router.huggingface.co/nebius/v1",
+                      "link":"google/gemma-2-2b-it-fast",
+                      },
+      "Zephyr-7B-β":{
+                      "inf_point":"https://router.huggingface.co/hf-inference/models/HuggingFaceH4/zephyr-7b-beta/v1",
+                      "link":"HuggingFaceH4/zephyr-7b-beta",
+                      },
+  }
 
 #Pull info about the model to display
 model_info ={
@@ -41,41 +51,46 @@ model_info ={
         {'description':"""The Mistral model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
             \nIt was created by the [**Mistral AI**](https://mistral.ai/news/announcing-mistral-7b/) team as has over  **7 billion parameters.** \n""",
         'logo':'https://cdn-avatars.huggingface.co/v1/production/uploads/62dac1c7a8ead43d20e3e17a/wrLf5yaGC6ng4XME70w6Z.png'},
-    "Gemma-7B":        
+    "Gemma-2-27B-it":        
+        {'description':"""The Gemma model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over  **27 billion parameters.** \n""",
+        'logo':'https://pbs.twimg.com/media/GG3sJg7X0AEaNIq.jpg'},
+    "Gemma-3-27B-it":        
         {'description':"""The Gemma model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
-            \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over  **7 billion parameters.** \n""",
+            \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over  **27 billion parameters.** \n""",
+        'logo':'https://pbs.twimg.com/media/GG3sJg7X0AEaNIq.jpg'},
+    "Gemma-2-2B-it":        
+        {'description':"""The Gemma model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over  **2 billion parameters.** \n""",
         'logo':'https://pbs.twimg.com/media/GG3sJg7X0AEaNIq.jpg'},
-    "Gemma-2B":        
-    {'description':"""The Gemma model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
-        \nIt was created by the [**Google's AI Team**](https://blog.google/technology/developers/gemma-open-models/) team as has over  **2 billion parameters.** \n""",
-    'logo':'https://pbs.twimg.com/media/GG3sJg7X0AEaNIq.jpg'},
     "Zephyr-7B":        
-    {'description':"""The Zephyr model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
-        \nFrom Huggingface: \n\
-        Zephyr is a series of language models that are trained to act as helpful assistants. \
-        [Zephyr 7B Gemma](https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1)\
-        is the third model in the series, and is a fine-tuned version of google/gemma-7b \
-        that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO)\n""",
-    'logo':'https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1/resolve/main/thumbnail.png'},
+        {'description':"""The Zephyr model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nFrom Huggingface: \n\
+            Zephyr is a series of language models that are trained to act as helpful assistants. \
+            [Zephyr 7B Gemma](https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1)\
+            is the third model in the series, and is a fine-tuned version of google/gemma-7b \
+            that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO)\n""",
+        'logo':'https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1/resolve/main/thumbnail.png'},
     "Zephyr-7B-β":        
-    {'description':"""The Zephyr model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
-        \nFrom Huggingface: \n\
-        Zephyr is a series of language models that are trained to act as helpful assistants. \
-        [Zephyr-7B-β](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta)\
-        is the second model in the series, and is a fine-tuned version of mistralai/Mistral-7B-v0.1 \
-        that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO)\n""",
-    'logo':'https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha/resolve/main/thumbnail.png'},
+        {'description':"""The Zephyr model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nFrom Huggingface: \n\
+            Zephyr is a series of language models that are trained to act as helpful assistants. \
+            [Zephyr-7B-β](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta)\
+            is the second model in the series, and is a fine-tuned version of mistralai/Mistral-7B-v0.1 \
+            that was trained on on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO)\n""",
+        'logo':'https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha/resolve/main/thumbnail.png'},
     "Meta-Llama-3-8B":
-    {'description':"""The Llama (3) model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
-        \nIt was created by the [**Meta's AI**](https://llama.meta.com/) team and has over  **8 billion parameters.** \n""",
-    'logo':'Llama_logo.png'},
+        {'description':"""The Llama (3) model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nIt was created by the [**Meta's AI**](https://llama.meta.com/) team and has over  **8 billion parameters.** \n""",
+        'logo':'Llama_logo.png'},
     "Meta-Llama-3.1-8B":
-    {'description':"""The Llama (3.1) model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
-        \nIt was created by the [**Meta's AI**](https://llama.meta.com/) team and has over  **8 billion parameters.** \n""",
-    'logo':'Llama3_1_logo.png'},
+        {'description':"""The Llama (3.1) model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nIt was created by the [**Meta's AI**](https://llama.meta.com/) team and has over  **8 billion parameters.** \n""",
+        'logo':'Llama3_1_logo.png'},
 }
 
 
+
 #Random dog images for error message
 random_dog = ["0f476473-2d8b-415e-b944-483768418a95.jpg",
               "1bd75c81-f1d7-4e55-9310-a27595fa8762.jpg",
@@ -143,6 +158,12 @@ if st.session_state.prev_option != selected_model:
 #Pull in the model we want to use
 repo_id = model_links[selected_model]
 
+# initialize the client
+client = OpenAI(
+  base_url=model_links[selected_model]["inf_point"],#"https://api-inference.huggingface.co/v1",
+  api_key=os.environ.get('HUGGINGFACEHUB_API_TOKEN')#"hf_xxx" # Replace with your token
+) 
+
 
 st.subheader(f'AI - {selected_model}')
 # st.title(f'ChatBot Using {selected_model}')
@@ -178,7 +199,7 @@ if prompt := st.chat_input(f"Hi I'm {selected_model}, ask me a question"):
 
         try:
             stream = client.chat.completions.create(
-                model=model_links[selected_model],
+                model=model_links[selected_model]["link"],
                 messages=[
                     {"role": m["role"], "content": m["content"]}
                     for m in st.session_state.messages