With Habana’s SynapseAI 1.8.0 release support of DeepSpeed Inference, users can run inference on large language models, including BLOOM 176B.