Faster Text Generation with Self-Speculative Decoding
•
62
None defined yet.
OneFlow: Concurrent Mixed-Modal and Interleaved Generation with Edit Flows
CWM: An Open-Weights LLM for Research on Code Generation with World Models