Axian 1.0
An open-source agent foundation model from ByteDance, optimized for agent tasks. It supports multi-step planning, tool calling and environment interaction.
Timeline (last 90 days)
ByteDance released an open-source Agent foundation model (recorded as Axian 1.0 in the source material), offering two scales, 35B and 122B.
The ByteDance team said the training pipeline for this Agent foundation model was completed with only two stages of supervised fine-tuning plus one stage of reinforcement learning, and the training code and about 18,000 data entries were released together.
All Facts (2)
The ByteDance team said the training pipeline for this Agent foundation model was completed with only two stages of supervised fine-tuning plus one stage of reinforcement learning, and the training code and about 18,000 data entries were released together.
50%UnverifiedByteDance released an open-source Agent foundation model (recorded as Axian 1.0 in the source material), offering two scales, 35B and 122B.
50%