doesn't seem to be running

README.md

@@ -143,6 +143,50 @@ Judging will be conducted by representatives from sponsor partners and the Huggi
 * **LlamaIndex Docs**: [https://llamaindex.ai/docs](https://llamaindex.ai/docs)
 * **Mistral Model Hub**: [https://huggingface.co/mistral-ai/mistral-small](https://huggingface.co/mistral-ai/mistral-small)
 
+## Free Credits!
+
+**Modal Labs Compute Credits** (\$250 per participant)
+  Monitor your GPU/CPU credit usage by logging into your Modal account and navigating to **Dashboard → Billing**:
+  [https://modal.com/dashboard](https://modal.com/dashboard) ([huggingface.co][1], [modal.com][2])
+
+**Hugging Face API Credits** (\$25 per participant)
+  View your remaining credits and invoices on the Hugging Face billing dashboard:
+  [https://huggingface.co/settings/billing](https://huggingface.co/settings/billing) ([huggingface.co][1], [huggingface.co][3])
+
+**Nebius AI Cloud Credits** (\$25 to first 3,300 participants)
+  Check your Nebius “Grants and promocodes” balance and detailed billing reports at:
+  [https://nebius.com/services/billing](https://nebius.com/services/billing) ([huggingface.co][1], [nebius.com][4])
+
+**Anthropic Claude API Credits** (\$25 to first 1,000 participants)
+  Track your Claude usage and remaining credits in the Anthropic Console under **Settings → Billing**:
+  [https://console.anthropic.com/settings/billing](https://console.anthropic.com/settings/billing) ([huggingface.co][1])
+
+**OpenAI API Credits** (\$25 to first 1,000 participants)
+  Monitor your API calls, token usage, and spend on the OpenAI Usage dashboard:
+  [https://platform.openai.com/account/usage](https://platform.openai.com/account/usage) ([huggingface.co][1], [platform.openai.com][5])
+
+**Hyperbolic Labs API Credits** (\$15 to first 1,000 participants)
+  After logging in at the Hyperbolic AI Dashboard, go to **Settings → Billing** to view your credit balance and transaction history:
+  [https://app.hyperbolic.xyz](https://app.hyperbolic.xyz) ([huggingface.co][1], [docs.hyperbolic.xyz][6], [hyperbolic.xyz][7])
+
+**Mistral AI API Credits** (\$25 to first 500 participants)
+  Sign in at the Mistral Console and navigate to **Workspace → Billing** to activate and monitor your credits:
+  [https://console.mistral.ai](https://console.mistral.ai) ([huggingface.co][1], [docs.mistral.ai][8])
+
+**SambaNova AI Cloud Credits** (\$25 to first 250 participants)
+  Log in to SambaNova Cloud and check your **Billing & Usage** in the plans section:
+  [https://cloud.sambanova.ai/plans/billing](https://cloud.sambanova.ai/plans/billing) ([huggingface.co][1], [cloud.sambanova.ai][9])
+
+[1]: https://huggingface.co/Agents-MCP-Hackathon "Agents-MCP-Hackathon (Agents-MCP-Hackathon)"
+[2]: https://modal.com/ "Modal: High-performance AI infrastructure"
+[3]: https://huggingface.co/docs/hub/billing?utm_source=chatgpt.com "Billing - Hugging Face"
+[4]: https://nebius.com/services/billing?utm_source=chatgpt.com "Billing - Nebius"
+[5]: https://platform.openai.com/account/usage?utm_source=chatgpt.com "Account Usage - OpenAI Platform"
+[6]: https://docs.hyperbolic.xyz/docs/getting-started?utm_source=chatgpt.com "Hyperbolic API"
+[7]: https://hyperbolic.xyz/blog/how-to-set-up-your-account-on-hyperbolic?utm_source=chatgpt.com "How to Set Up Your Account on Hyperbolic"
+[8]: https://docs.mistral.ai/getting-started/quickstart/?utm_source=chatgpt.com "Quickstart | Mistral AI Large Language Models"
+[9]: https://cloud.sambanova.ai/plans/billing?utm_source=chatgpt.com "Billing - SambaNova Cloud"
+
 # About the Author
 
 **Graham Paasch** is an AI realist passionate about the coming AI revolution.

requirements.txt

@@ -23,3 +23,5 @@ requests  # For MCP endpoint testing
 
 # system requirement for audio I/O
 ffmpeg-python
+
+psutil  # For system resource detection

src/app.py

@@ -1,10 +1,117 @@
 import os
+from pathlib import Path
+from huggingface_hub import snapshot_download
 import gradio as gr
 from llama_index.core import Settings, ServiceContext
 from llama_index.embeddings.huggingface import HuggingFaceEmbedding
 from llama_index.llms.llama_cpp import LlamaCPP
 from parse_tabular import create_symptom_index
 import json
+import torch
+import psutil
+import subprocess
+from typing import Tuple, Dict
+
+# Model options mapped to their requirements
+MODEL_OPTIONS = {
+    "tiny": {
+        "name": "TinyLlama-1.1B-Chat-v1.0.Q4_K_M.gguf",
+        "repo": "TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF",
+        "vram_req": 2,  # GB
+        "ram_req": 4    # GB
+    },
+    "small": {
+        "name": "phi-2.Q4_K_M.gguf",
+        "repo": "TheBloke/phi-2-GGUF",
+        "vram_req": 4,
+        "ram_req": 8
+    },
+    "medium": {
+        "name": "mistral-7b-instruct-v0.1.Q4_K_M.gguf",
+        "repo": "TheBloke/Mistral-7B-Instruct-v0.1-GGUF",
+        "vram_req": 6,
+        "ram_req": 16
+    }
+}
+
+def get_system_specs() -> Dict[str, float]:
+    """Get system specifications."""
+    # Get RAM
+    ram_gb = psutil.virtual_memory().total / (1024**3)
+    
+    # Get GPU info if available
+    gpu_vram_gb = 0
+    if torch.cuda.is_available():
+        try:
+            # Query GPU memory in bytes and convert to GB
+            gpu_vram_gb = torch.cuda.get_device_properties(0).total_memory / (1024**3)
+        except Exception as e:
+            print(f"Warning: Could not get GPU memory: {e}")
+    
+    return {
+        "ram_gb": ram_gb,
+        "gpu_vram_gb": gpu_vram_gb
+    }
+
+def select_best_model() -> Tuple[str, str]:
+    """Select the best model based on system specifications."""
+    specs = get_system_specs()
+    print(f"\nSystem specifications:")
+    print(f"RAM: {specs['ram_gb']:.1f} GB")
+    print(f"GPU VRAM: {specs['gpu_vram_gb']:.1f} GB")
+    
+    # Prioritize GPU if available
+    if specs['gpu_vram_gb'] >= 4:  # You have 6GB, so this should work
+        model_tier = "small"  # phi-2 should work well on RTX 2060
+    elif specs['ram_gb'] >= 8:
+        model_tier = "small"
+    else:
+        model_tier = "tiny"
+    
+    selected = MODEL_OPTIONS[model_tier]
+    print(f"\nSelected model tier: {model_tier}")
+    print(f"Model: {selected['name']}")
+    
+    return selected['name'], selected['repo']
+
+# Set up model paths
+MODEL_NAME, REPO_ID = select_best_model()
+BASE_DIR = os.path.dirname(os.path.dirname(__file__))
+MODEL_DIR = os.path.join(BASE_DIR, "models")
+MODEL_PATH = os.path.join(MODEL_DIR, MODEL_NAME)
+
+def ensure_model():
+    # Create models/ directory if missing
+    os.makedirs(MODEL_DIR, exist_ok=True)
+    
+    # Download model if it's not already there
+    model_path = os.path.join(MODEL_DIR, MODEL_NAME)
+    if not os.path.isfile(model_path):
+        print(f"Downloading model from {REPO_ID}...")
+        # Download to a subdirectory to avoid file conflicts
+        download_dir = os.path.join(MODEL_DIR, "download_cache")
+        snapshot_download(
+            repo_id=REPO_ID,
+            repo_type="model",
+            local_dir=download_dir,
+            local_dir_use_symlinks=False
+        )
+        
+        # Move the specific model file we want to models/
+        src_path = os.path.join(download_dir, MODEL_NAME)
+        if os.path.exists(src_path):
+            import shutil
+            shutil.move(src_path, model_path)
+            print(f"Moved model to {model_path}")
+        else:
+            raise ValueError(f"Downloaded files but couldn't find {MODEL_NAME}")
+    else:
+        print(f"Model already exists at {model_path}")
+    
+    return model_path
+
+# Ensure model is downloaded
+model_path = ensure_model()
 
 # Configure embeddings globally
 Settings.embed_model = HuggingFaceEmbedding(
@@ -13,8 +120,7 @@ Settings.embed_model = HuggingFaceEmbedding(
 
 # Configure local LLM with LlamaCPP
 llm = LlamaCPP(
-    model_url="https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF/resolve/main/mistral-7b-instruct-v0.1.Q4_K_M.gguf",
-    model_path="models/mistral-7b-instruct-v0.1.Q4_K_M.gguf",
+    model_path=model_path,
     temperature=0.7,
     max_new_tokens=256,
     context_window=2048