Ho sostenuto un colloquio presso TensorGo Technologies
Colloquio
3 Rounds.
First - technical interview with their AI avatar Zai regarding previous experiences and plans for ongoing research/projects
Second - technical interview with the Computer Vision team lead regarding ongoing projects & research work covering conceptual questions about model requirements, performance, and limitations
Third - HR discussion briefing about the company work culture and expectations
Domande di colloquio [1]
Domanda 1
What are the fine-tuning requirements for a multi-modal model given a specific task?
Ho presentato la mia candidatura online. La procedura ha richiesto 5 giorni. Ho sostenuto un colloquio presso TensorGo Technologies (Hyderabad) nel mese di dic 2024
Colloquio
First round: An Assignment on finetuning an LLM even though my role had nothing to do with it. You are required to make a video explaining the process and code of the assignment.
Second round: Interview with AI Lead, He did not seem well interested in having a conversation. Felt more like viva in college where its just QnA and nothing else.
He did ask some nice and thoughtful questions.
Domande di colloquio [2]
Domanda 1
Imagine there are 2 people in a 2D Image. how would you find the distance between them?
Imagine a camera watching traffic on a crossroad. The camera is writing each of the frames IDs and other information into a csv file. How will you read the csv file while its being written by the the camera.
Ho presentato la mia candidatura tramite l'università. La procedura ha richiesto una settimana. Ho sostenuto un colloquio presso TensorGo Technologies (Chennai) nel mese di giu 2024
Colloquio
round 1 assignment submission. it was easy, they asked me to build a RAG without training. Then i got shortlisted to round 2 . The round 2 is a technical round. They asked about llms and computer vison and deep learning questions. Then i got rejected after round 2.
Domande di colloquio [1]
Domanda 1
how will you finetune llms?
difference between object detection and image segmentation?