Hi, I already posted this at Beginners category, but I am seeking more help here. I am currently trying to perform full fine tuning on the ai-forever/mGPT model (1.3B parameters) using a single A100 GPU (40GB VRAM) on Google Colab. However when running the training is very slow: ~0.06 it/s. Here …

SFTTrainer training very slow on GPU. Is this training speed expected?

domce20 February 8, 2025, 6:52pm 4

1 Like

Topic		Replies	Views
SFTTrainer training very slow. Is this training speed expected? Beginners	4	178	February 8, 2025
SFTTrainer too slow during the build (or ingestion) phase 🤗Transformers	0	92	November 27, 2024
SFTTrainer takes up so much ram that it breaks an A100 GPU 🤗Transformers	0	184	July 6, 2024
SFTTrainer Doubling Speed on a Single GPU with DeepSpeed: Proposal for an Update to the Official Documentation and Verification Report DeepSpeed	1	35	March 7, 2025
Very slow training (>5mins per batch) - code review request Research	2	618	October 11, 2023