Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

LLM Training on Llama2 70B (64 x Trainium2-16)

11.2Iteration Time (s)

AXLearn

Updated 1mo ago

Evaluation Results

Method	Links
AXLearn 2025.07		11.2	25	-

SOTA Paper

AXLearn

AXLearn: Modular, Hardware-Agnostic Large Model Training

Dataset

Llama2

Follow for update

@wizwand_team Discord

Related Benchmarks

LLM Training on Llama2 7B LLM Training on Llama2-70B (64 x H100-8)LLM Training on Llama2-7B tpu-v5p-512 LLM Training on Llama2 70B (tpu-v5p-1024)LLM Training on Qwen-3 30B-A3B (tpu-v5p-1024)

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs