Tyler Titsworth
Community@tylertitsworth · Vancouver, WA
MLOps Engineer
Agent Skills by Tyler Titsworth
Showing 28 vetted skills indexed across 1 GitHub repositories.
tylertitsworthopenai-api
Coordinate OpenAI API usage across multiple backends with a consistent interface.
tylertitsworthmodel-formats
Compare ML model serialization formats for deployment and conversion workflows.
tylertitsworthllm-evaluation
Runs lm-evaluation-harness benchmarks for LLMs across multiple configurable backends and tasks.
tylertitsworthray-core
Orchestrate distributed Python applications with Ray Core tasks, actors, and object store.
tylertitsworthaws-efa
Diagnose and configure AWS EFA for distributed GPU training on EKS.
tylertitsworthlogql
Filters and analyzes Loki logs using LogQL queries with parsers and aggregations.
tylertitsworthkueue
Automate Kubernetes workload admission and scheduling with queue-based quota management.
tylertitsworthhuggingface-transformers
Manage HuggingFace models and datasets in Python workflows.
tylertitsworthvllm
Configure and run vLLM for high-throughput LLM inference on GPUs.
tylertitsworthfsdp
Distribute PyTorch model training across multiple GPUs using FSDP sharding strategies.
tylertitsworthverl
Coordinate scalable RLHF training for large language models with PPO, GRPO, DAPO, or RLOO.
tylertitsworthray-serve
Deploy scalable model serving on Kubernetes with Ray Serve.
tylertitsworthoperator-sdk
Scaffold Kubernetes operators in Go, Helm, or Ansible.
tylertitsworthaws-fsx
Explain FSx for Lustre architecture and deployment on Kubernetes with S3.
tylertitsworthtorch-compile
Apply torch.compile to PyTorch models for accelerated training and inference.
tylertitsworthflash-attention
Configure FlashAttention variants and alternative backends for CUDA GPUs.
tylertitsworthflyte-kuberay
Provision ephemeral Ray clusters per Flyte task execution via KubeRay.
tylertitsworthray-data
Build scalable ML data pipelines with Ray Data streaming execution.
tylertitsworthpytorch
Build and train PyTorch models with nn.Module, DDP, FSDP, mixed precision, and serialization to ONNX.
tylertitsworthmegatron-lm
Configure GPU cluster parallelism and manage Megatron-LM training workflows.
tylertitsworthollama
Manage local LLM models via CLI and OpenAI-compatible REST API.
tylertitsworthflyte-deployment
Deploy and operate Flyte on Kubernetes using Helm charts.
tylertitsworthkuberay
Deploy and manage distributed Ray clusters on Kubernetes with the KubeRay operator.
tylertitsworthpromql
Write, interpret, and optimize PromQL queries for Prometheus metrics analysis.