~ 2x faster inference than qwen 3.5 122b
~ 7x faster inference than gpt-oss 120b
probably most important: training datasets and training recipe available (!)
in other words this is an open source llm release (not just open weights!)
tosh•2h ago
~ 2x faster inference than qwen 3.5 122b
~ 7x faster inference than gpt-oss 120b
probably most important: training datasets and training recipe available (!)
in other words this is an open source llm release (not just open weights!)
IamTC•2h ago