The GLUE benchmark graph shows how Microsoft and other AI models improved at language tasks, tracking progress vs. human performance. LINK SCREENSHOTGPT 333Tags: 111222 […]
Category: chat gpt
dataset format type torch vs tensorflow
Both are AI/ML tensor data formats—one as PyTorch tensors, the other as TensorFlow tensors LINK SCREENSHOTGPT 333Tags: 111222 […]
Function Calling vs General Fine-Tuning
Function calling fine-tuning needs specific data, not just general fine-tuning LINK SCREENSHOTGPT 333Tags: 111222 […]
Training vs Inference Function Calling
Function calling can also be part of training to improve robustness LINK SCREENSHOTGPT 333Tags: 111222 […]
Three ways to implement system prompt logic
System behavior can be embedded via training, simulated through instructions or examples, or set at runtime LINK SCREENSHOTGPT 333Tags: 111222 […]
Inside Custom GPT Using ReAct: How System Prompts Steer Dynamic Tool Loops
System instructions shape how GPT reacts to action responses across every stage LINK SCREENSHOTGPT 333Tags: 111222 […]
Read More… from Inside Custom GPT Using ReAct: How System Prompts Steer Dynamic Tool Loops
DALL·E vs Midjourney Differences
DALL·E excels at prompt accuracy and text; Midjourney leads in art quality, style, and control options LINK SCREENSHOTGPT 333Tags: 111222 […]
AI Agent Core Definition
AI agent = model + tools + autonomy LINK SCREENSHOTGPT 333Tags: 111222 […]
Autonomy in ReAct Loop – connection to observe step
Autonomy lets agents observe and adapt independently LINK SCREENSHOTGPT 333Tags: 111222 […]
Read More… from Autonomy in ReAct Loop – connection to observe step
ChatGPT Doesn’t Need Message Templates
It handles chat formatting internally unlike raw Hugging Face models (model is more raw but is more flexible) LINK SCREENSHOTGPT 333Tags: 111222 […]