Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Inference Efficiency on 30k Context Length (Llama-3.1-8B)

15.8Inference Throughput (QPS)

Finetuning

0.724.6358.5512.465Mar 11, 2025
Updated 3d ago

Evaluation Results

MethodLinks
2025.03
15.8-
2025.03
12.8-
2025.03
11.6-
2025.03
1.3-