| feat(machine): Simplify runtime config stage 2 | 1 个月前 |
| feat(operation): Fix LI accuracy issue | 16 小时前 |
| refactor(operator): Clear clean code | 15 天前 |
| feat(python): A3 platform high-throughput performance version of glm_attention. | 1 天前 |
| fix(operator): Replace the implementation of the Attnres operator with four kernels | 19 天前 |
| feat(operator): Add quantization aware training(QAT) | 2 个月前 |
| fix(operator): Fix NaN/Inf handling in the GDR operator compare function | 14 天前 |
