Dogukan Uraz Tuna – Medium

Dogukan Uraz Tuna

Dogukan Uraz Tuna
in
ITNEXT

Structured Outputs Effectively with GPT-4o Function Calling on Financial Calculations

Hi everyone, welcome to this week’s post! Today, we’re diving into an exciting new feature from OpenAI — the Structured Outputs via API…

Aug 17

Structured Outputs Effectively with GPT-4o Function Calling on Financial Calculations

Aug 17

Dogukan Uraz Tuna

Training RL Agent with Policy Gradients with Parameter-based Exploration for Custom Objective

Let’s train RL agent with neuroevolutionary strategy using Jax.

Aug 5

Training RL Agent with Policy Gradients with Parameter-based Exploration for Custom Objective

Aug 5

Dogukan Uraz Tuna
in
ITNEXT

Writing AI Agent with Rust from Scratch, Deploying to Edge with WebAssembly — OS Week 1

Let’s write AI agent with Rust from scratch, deploy to the client-side edge with WebAssembly…

Jul 29

Writing AI Agent with Rust from Scratch, Deploying to Edge with WebAssembly — OS Week 1

Jul 29

Dogukan Uraz Tuna
in
ITNEXT

Enhancing Performance with C/C++ Code Execution for Langchain Agents

For various algorithms, logics or calculations, it can be more effective to have the LLM generate code to perform the computation, and then…

Jul 18

Enhancing Performance with C/C++ Code Execution for Langchain Agents

Jul 18

Dogukan Uraz Tuna
in
ITNEXT

OpenAI’s Code Execution Runtime & Replicating Sandboxing Infrastructure

Looking into OpenAI’s code execution runtime for Assistants API and replicating its sandboxing infrastructure on Google Kubernetes Engine.

Jul 2

OpenAI’s Code Execution Runtime & Replicating Sandboxing Infrastructure

Jul 2

Dogukan Uraz Tuna

Utilizing Airflow for Planning, Scheduling, Executing and Scaling AI Agentic Workloads

My summarized thought about this convergence: Despite having such a setup, a toy agent setup, as a conclusion, I believe the Airflow engine…

May 6

Utilizing Airflow for Planning, Scheduling, Executing and Scaling AI Agentic Workloads

May 6

Dogukan Uraz Tuna

Linen Modules for Simple Long Convolutions Sequence Modeling with Log-Linear Time Complexity

Introducing the LongConv-Jax. Implementation of the “Simple Long Convolutions Sequence Modeling” paper in Jax library.

Mar 20

Linen Modules for Simple Long Convolutions Sequence Modeling with Log-Linear Time Complexity

Mar 20

Dogukan Uraz Tuna

Releasing Tri-RMSNorm: Yielding Speed-up for Training Stabilization + [Let’s Review] Root Mean…

In this blog post, releasing accelerated Tri-RMSNorm compute kernel v0.1.0 and going into the details of LayerNorm and RMSNorm processes.

Mar 14

Releasing Tri-RMSNorm: Yielding Speed-up for Training Stabilization + [Let’s Review] Root Mean…

Mar 14

Dogukan Uraz Tuna

[Let’s Review] Gradient Ascent + Jax Implementation

Introduction to gradient ascent algorithm and its use in Jax/FLAX models through GradientAscent-Jax packaged, implemented by simudt.

Mar 11

[Let’s Review] Gradient Ascent + Jax Implementation

Mar 11

Dogukan Uraz Tuna

[Let’s Review] Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient LMs

Architectural review of the paper “Griffin” by Google Deepmind and implemetation in Jax.

Mar 7

[Let’s Review] Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient LMs

Mar 7

Dogukan Uraz Tuna

Dogukan Uraz Tuna

Building AI products & large-scale platform services for AI ⚡ Interests; AI infra and AI agents x distributed systems - bento.me/dtuna

Help
Status
About
Careers
Press
Blog
Privacy
Terms
Text to speech
Teams