We have an extensive collection of digitized archive records. Sometimes the OCR technologies that digitize these physical records make mistakes and produce some sections of garbled text. If you were to train an LLM on these records, how would you handle such OCR errors affecting training and inference?
Why did you leave that company?
Two arbitarty rectangle overlap area
How to train a PCA algorithm
How many ways to climb to a ladder of n height, if you can take either one or two steps. (hint: dynamic programming)
About kalman filter
C++ question, difference between shallow copying and deep copying
What makes me unique? How your past experience and skills fit to this position? Do you need a visa? How your experience with problem X,Y,Z?
They asked me about my ML background and the exercise was about regular expressions in Python.
En la prueba técnica del cliente te piden desarrollar un API que consuma un modelo de ML. Es obligatorio entregar: - Continuous Integration y Continuous Deliver - Testing - Endpoints en FastAPI - Despliegue en Google Cloud Entregan muy poco código hecho y un CSV. Además, piden implementar todo el ciclo de vida de un modelo de ML en 5 días
Viewing 251 - 260 interview questions