AI가 움직이고 미래가 업데이트됩니다.Beam-1.0-501B-A23B-NVFP4
Beam is an open model built by Reflection. Beam takes text as input and generates text as output. This checkpoint contains the NVFP4-quantized instruction-tuned model. The Beam release also includes BF16 and FP8 checkpoints. Beam uses a sparse mixture-of-experts (MoE) architecture with 501B total parameters and 23B active parameters. It supports a context window of 1M tokens and was developed predominantly for English. Beam models are designed to deliver strong performance for areas such as agentic tool use, coding and general reasoning with an emphasis on token efficient usage. This checkpoint was created using post-training quantization (PTQ) with NVIDIA Model Optimizer (ModelOpt) . We’re grateful to Jinze Xue, Matthew Radzihovsky, Zhiyu Cheng and Trenton Starkey at NVIDIA for their valuable guidance and support in developing and refining our quantization recipes.
번역 준비 중입니다. 원문을 표시합니다.
Beam is an open model built by Reflection. Beam takes text as input and generates text as output. This checkpoint contains the NVFP4-quantized instruction-tuned model. The Beam release also includes BF16 and FP8 checkpoints. Beam uses a sparse mixture-of-experts (MoE) architecture with 501B total parameters and 23B active parameters. It supports a context window of 1M tokens and was developed predominantly for English. Beam models are designed to deliver strong performance for areas such as agentic tool use, coding and general reasoning with an emphasis on token efficient usage. This checkpoint was created using post-training quantization (PTQ) with NVIDIA Model Optimizer (ModelOpt) . We’re grateful to Jinze Xue, Matthew Radzihovsky, Zhiyu Cheng and Trenton Starkey at NVIDIA for their valuable guidance and support in developing and refining our quantization recipes.