GPT-OSS-20B
AWS shows SFT fixes GPT-OSS-20B's English-only reasoning, GRPO adds less
GPT-OSS-20B tends to reason in English even when told to use French, German, Spanish or Italian, and AWS says supervised fine-tuning on Amazon SageMaker HyperPod fixed…
d-Matrix pairs Corsair with Nvidia H100 for low-latency attention-FFN split
d-Matrix says it has built a heterogeneous inference pipeline in which an Nvidia H100 handles attention while its Corsair accelerators run the expert computation of a…