Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 14 hours
Course Outline
AI Sovereignty and Local LLM Deployment
- Mitigating risks associated with cloud LLMs, including data retention, potential training on user inputs, and jurisdictional concerns.
- Exploring Ollama’s architecture, comprising its model server, registry, and OpenAI-compatible API.
- Evaluating Ollama against alternatives such as vLLM, llama.cpp, and Text Generation Inference.
- Understanding licensing terms for models like Llama, Mistral, Qwen, and Gemma.
Installation and Hardware Configuration
- Setting up Ollama on Linux with support for CUDA and ROCm.
- Configuring CPU-only fallbacks and optimizing for AVX/AVX2 instruction sets.
- Implementing Docker deployments with persistent volume mapping.
- Establishing multi-GPU environments and defining VRAM allocation strategies.
Model Management
- Fetching models from the Ollama registry (e.g., using the 'ollama pull llama3' command).
- Importing GGUF format models from HuggingFace and TheBloke.
- Assessing trade-offs between quantization levels such as Q4_K_M, Q5_K_M, and Q8_0.
- Managing model switching and understanding limits on concurrent model loading.
Creating Custom Modelfiles
- Mastery of Modelfile syntax, including FROM, PARAMETER, SYSTEM, and TEMPLATE directives.
- Fine-tuning parameters like temperature, top_p, and repeat_penalty.
- Engineering system prompts to enforce role-specific behaviors.
- Building and publishing custom models to a local registry.
API Integration
- Utilizing the OpenAI-compatible /v1/chat/completions endpoint.
- Handling streaming responses and enforcing JSON output mode.
- Integrating with frameworks such as LangChain, LlamaIndex, and custom applications.
- Implementing authentication and rate limiting via reverse proxies.
Performance Optimization
- Optimizing context window sizes and managing KV caches.
- Implementing batch inference and parallel request handling.
- Optimizing CPU thread allocation with NUMA awareness.
- Monitoring GPU utilization and managing memory pressure.
Security and Compliance
- Isolating network traffic for model serving endpoints.
- Implementing pipelines for input filtering and output moderation.
- Generating audit logs for prompts and completions.
- Verifying model provenance and hash integrity.
Requirements
- Intermediate proficiency in Linux administration and container management.
- A high-level understanding of machine learning concepts and transformer architectures.
- Working knowledge of REST APIs and JSON data structures.
Target Audience
- AI engineers and developers looking to decouple from cloud LLM APIs.
- Organizations with strict data sensitivity policies that preclude the use of cloud-based models.
- Government and defense sectors requiring air-gapped language model capabilities.
