Farshad Ghodsian
Sr. Technical Product Manager - AI Infrastructure & MLOps @ AMD
Actions
Farshad is a Senior Technical Product Manager on the AI Infrastructure and MLOps team at AMD. He is currently helping to lead AI accelerator software products at AMD, with a focus on AI cluster management and Kubernetes-based AI deployment solutions for both large-scale inference and training workloads. He is responsible for helping to define the product strategy and roadmap for AMD Instinct GPUs, assisting large datacenter-scale customers with deploying AMD Instinct GPUs at scale, and documenting and presenting AMD's enterprise software offerings to large audiences. He is passionate about learning and teaching others and if you let him, he will prob talk about Machine Learning and GenAI for days on end.
Area of Expertise
Powering your Generative AI Workloads with AMD and Open-Source ROCm
Presented at AI_dev: Open Source GenAI & ML Summit Europe in Paris, France - June, 2024
View Recording: https://www.youtube.com/watch?v=k2g_lC0fI-k
In the generative AI ecosystem today, there is a strong emphasis on expensive AI hardware and proprietary CUDA implementations. While CUDA has undeniably played a crucial role in the success of generative AI, I’d like to share my experience with running generative AI workloads and applications on cost-effective AMD hardware and the open-source ROCm software stack. This alternative approach aims to provide users with greater flexibility and options, allowing them to apply their generative AI solutions across a wider range of hardware and software choices than ever before.
Learn how to run your favourite open source large language and image generation models using ROCm, how far ROCm has come from previous versions and what features are currently supported, including PyTorch, HuggingFace Transformers, BitsandBytes, Flash Attention, vLLM and TorchTune, and how more affordable workstation and server class AMD GPUs compare to their Nvidia counterparts in terms of performance and inference speed. You will also see several demos of ROCm in action and some tips and things to watch out for when working with AMD GPUs.
GenOps: Building a MLOps Platform to Support GenAI Workloads with Open-Source and Kubeflow
Presented at Cassandra Summit + AI.dev 2023 in San Jose, California - Dec 2023
View Recording: https://www.youtube.com/watch?v=w8a7Pu7n5Nc
Taking a deep dive into how we have built an end-to-end MLOps platform on GKE (Google Kubernetes Engine) using Open-Sourced technologies like Kubeflow, MLFlow, Spark on Kubernetes, and other open-sourced tools and how we are using it to support Generative AI models (specifically LLMs) in the Cloud. Will also walkthrough some learnings, tips and a demo on how you can leverage the same open-sourced tooling to run your models.
GenOps: Building a MLOps Platform to Support GenAI Workloads with Kubeflow on Google Cloud
Presented at DevFestYYC in Calgary, Alberta, Canada - Nov 2023
Taking a deep dive into how we have built an end-to-end MLOps platform on Google Kubernetes Engine (GKE) using Kubeflow, MLFlow, Spark on Kubernetes, and other open-sourced tools to support all aspects of the machine learning lifecycle. Will also walkthrough some learnings, tips and a demo on how we are leveraging this platform to run Generative AI models (specifically LLMs) in the Cloud.
Prompt Hacking and How to Safeguard Your LLM with Nvidia NeMo Guardrails
Presented at Bell Cloud Day Conference in Toronto & Montreal - June 2024
Presenting different methods bad actors use to circumvent traditional LLM prompt guards via prompt injection and prompt hacking and how to safeguard your LLM from these techniques using NeMo Guardrails from Nvidia. Will walk through how to setup NeMo Guardrails, how to implement various guards and rails and demo them in action using the NeMo Guardrails server.
AMD Instinct Kubernetes & Virtualization Tooling
Presented at AMD AI Infrastructure Summit 2024 in Sonoma, California - Nov 2024
Providing an in-depth overview of AMD's software and platform offerings for AMD Instinct GPUs on Kubernetes and in virtualized environments. This presentation will also cover key upcoming features of the GPU Operator and Device Metrics Exporter for Kubernetes including a demo of GPU partitioning and various virtualization strategies for GPU Passthrough and SR-IOV.
AMD AI Infrastructure Summit 2024
AMD Instinct Kubernetes & Virtualization Tooling
AI_dev: Open Source GenAI & ML Summit Europe Sessionize Event
Prompt Hacking and How to Safeguard Your LLM with Nvidia NeMo Guardrails
Presenting different methods bad actors use to circumvent traditional LLM prompt guards via prompt injection and prompt hacking and how to safeguard your LLM from these techniques using NeMo Guardrails from Nvidia. Will walk through how to setup NeMo Guardrails, how to implement various guards and rails and demo them in action using the NeMo Guardrails server on our Kubeflow GenOps Platform.
Cassandra Summit + AI.dev 2023 Sessionize Event
ᐳᐅ!DEVFESTYYC | More Festival! More Google! More Fun! Sessionize Event
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top