You are allowed to use AI to help make the movie, but you are not allowed to use AI actors or writers.

You are allowed to use AI to help make the movie, but you are not allowed to use AI actors or writers.

Now actors and writers must be human. As the Academy released its rules for the 99th Academy Awards, the organization announced that any movies with “AI-generated actors” or “AI-written screenplays” would not be eligible for acting or writing awards (but otherwise still eligible). So what do you do, exactly, in a time when we can … Read more

Code Execution for Parsing, Analyzing, Visualizing, and Debugging Agent Reasoning Traces using the lambda/hermes-agent-reasoning-traces dataset

Code Execution for Parsing, Analyzing, Visualizing, and Debugging Agent Reasoning Traces using the lambda/hermes-agent-reasoning-traces dataset

In this lesson, we examine the lambda/hermes-agent-reasoning-traces dataset understanding how agent-based models think, use tools, and generate responses across multi-curve conversations. We start by loading and examining the dataset, examining its structure, categories, and dialog format to get a clear view of the available information. We then developed simple parsers to extract important components such … Read more

New NVIDIA Research Shows Predictive Code Release on NeMo RL Achieves 1.8× Faster Generation Release on 8B and 2.5× End-to-End Speedup on 235B Designs

New NVIDIA Research Shows Predictive Code Release on NeMo RL Achieves 1.8× Faster Generation Release on 8B and 2.5× End-to-End Speedup on 235B Designs

If you’ve been using reinforcement learning (RL) in a mathematical reasoning language model, code generation, or any realizable task, you’ve probably stared at the progress bar while your GPU cluster fired up generating output. A team of researchers from NVIDIA proposes fine-tuning by integrating predictive modeling into the RL training loop itself, and doing it … Read more