| | Use positional bidirectional mask helper call (#22) | BoLiu, nvidia-oliver-holworthy | May 20, 2026 |
| | Add rope_theta to rope_scaling for transformers 5.4+ compatibility (#19) | nvidia-oliver-holworthy |
| | Upload LICENSE | BoLiu |
| | Update vLLM usage docs: remove config_vllm.json overwrite, relax version pin, and clarify minimal required flags (#18) | ybabakhin, nvidia-oliver-holworthy |
| | Update vllm version to 0.16.0 (#17) | BoLiu, radekosmulski-nvidia |
| | Add support for transformers 4.44 through 5.0+ (#16) | nvidia-oliver-holworthy |
| | Remove the setting of _attn_implementation from llama_bidirectional_model (#3) | nvidia-oliver-holworthy |
| | Remove unused imports (#15) | nvidia-oliver-holworthy |
| | "use_bidirectional_attention": true flag (#13) | ybabakhin, michaelfeil |
| | Update sentence_bert_config.json | ybabakhin |
| | Integrate with Sentence Transformers (#6) | ybabakhin |
| | Add vllm config and information (#11) | ybabakhin |
| | Update README.md | Mengyao00 |
| | Update README.md (#4) | Mengyao00 |
| | Update README.md | ybabakhin |
| | Update README.md | sosahi |
| | Update README.md | Mengyao00 |
| | Update README.md | Mengyao00 |
| | Update llama_bidirectional_model.py | Mengyao00 |
| | Update README.md | Mengyao00 |