Switch to Llama-3.1-8B-Instruct with Nebius provider using chat.completions.create method

config/models.yaml

@@ -1,10 +1,10 @@
 models:
-  # Qwen2.5-7B-Instruct with HF Inference Provider
-  - name: "Qwen2.5-7B-Instruct"
-    provider: "hf-inference"
-    model_id: "Qwen/Qwen2.5-7B-Instruct"
+  # Llama-3.1-8B-Instruct with Nebius Provider
+  - name: "Llama-3.1-8B-Instruct"
+    provider: "nebius"
+    model_id: "meta-llama/Llama-3.1-8B-Instruct"
     params:
       max_new_tokens: 256
       temperature: 0.1
       top_p: 0.9
-    description: "Qwen2.5-7B-Instruct - Instruction-following model for text generation"
+    description: "Llama-3.1-8B-Instruct - Meta's instruction-following model via Nebius"

src/models_registry.py

@@ -86,7 +86,23 @@ class HuggingFaceInference:
             )
             
             # Use different methods based on provider capabilities
-            if provider == "together":
+            if provider == "nebius":
+                # Nebius provider uses chat.completions.create
+                completion = client.chat.completions.create(
+                    model=model_id,
+                    messages=[
+                        {
+                            "role": "user",
+                            "content": prompt
+                        }
+                    ],
+                    max_tokens=params.get('max_new_tokens', 128),
+                    temperature=params.get('temperature', 0.1),
+                    top_p=params.get('top_p', 0.9)
+                )
+                # Extract the content from the response
+                return completion.choices[0].message.content
+            elif provider == "together":
                 # Together provider uses chat_completion for conversational models
                 result = client.chat_completion(
                     messages=[{"role": "user", "content": prompt}],
@@ -184,7 +200,7 @@ class ModelInterface:
             return self._generate_mock_sql(model_config, prompt)
         
         try:
-            if model_config.provider in ["huggingface", "hf-inference", "together"]:
+            if model_config.provider in ["huggingface", "hf-inference", "together", "nebius"]:
                 print(f"🤗 Using {model_config.provider} Inference API for {model_config.name}")
                 return self.hf_interface.generate(
                     model_config.model_id,