{"id":19999792,"url":"https://github.com/ictnlp/sillm","last_synced_at":"2025-05-04T14:32:18.369Z","repository":{"id":223482776,"uuid":"760444226","full_name":"ictnlp/SiLLM","owner":"ictnlp","description":"SiLLM is a Simultaneous Machine Translation (SiMT) Framework. It utilizes a Large Language model as the translation model and employs a traditional SiMT model for policy-decision to achieve SiMT through collaboration.","archived":false,"fork":false,"pushed_at":"2024-02-22T02:15:16.000Z","size":10393,"stargazers_count":15,"open_issues_count":0,"forks_count":2,"subscribers_count":2,"default_branch":"main","last_synced_at":"2024-10-18T22:03:56.905Z","etag":null,"topics":["chatgpt","large","large-language-models","llama2","machine-translation","simultaneous-machine-translation","translation"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ictnlp.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-02-20T12:39:22.000Z","updated_at":"2024-10-16T03:39:32.000Z","dependencies_parsed_at":"2024-07-27T19:05:20.559Z","dependency_job_id":"ef586b11-a735-41d9-b19b-cfceb56a4437","html_url":"https://github.com/ictnlp/SiLLM","commit_stats":{"total_commits":11,"total_committers":1,"mean_commits":11.0,"dds":0.0,"last_synced_commit":"2c952ec5dc6e78bf6ba2481f4496b522c39c52c8"},"previous_names":["ictnlp/sillm"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ictnlp%2FSiLLM","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ictnlp%2FSiLLM/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ictnlp%2FSiLLM/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ictnlp%2FSiLLM/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ictnlp","download_url":"https://codeload.github.com/ictnlp/SiLLM/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":224393837,"owners_count":17303725,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["chatgpt","large","large-language-models","llama2","machine-translation","simultaneous-machine-translation","translation"],"created_at":"2024-11-13T05:12:49.186Z","updated_at":"2024-11-13T05:12:49.822Z","avatar_url":"https://github.com/ictnlp.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# SiLLM\n\nSource code for our paper \"SiLLM: Large Language Models for Simultaneous Machine Translation\".\n\n\u003cdiv style=\"text-align:center;\"\u003e\n  \u003cimg src=\"./model.PNG\" alt=\"替代文本\" width=\"800\" height=\"320\"\u003e\n\u003c/div\u003e\n\nThe framework of SiLLM incorporates the LLM to achieve the Simultaneous Machine Translation. It generates the translations under the guidance of the policy decided by the conventional Simultaneous Machine Translation Model.\n\nOur method is implemented based on the open-source toolkit [Alpaca-LoRA](https://github.com/tloen/alpaca-lora).\n\n## Requirements and Installation\n\n* Python version = 3.8\n\n* PyTorch version = 2.2\n\n* Install our library:\n\n```\ngit clone https://github.com/ictnlp/SiLLM.git\ncd SiLLM\npip install -r requirements.txt\n```\n\n## Quick Start\n\n### Fine-tune\n\nWe sample 100k data for fine-tuning LLM from WMT15 German-English (download [here](https://www.statmt.org/wmt15)) and MuST-C English-German (download [here](https://mt.fbk.eu/must-c/)), respectively. In the given example, we sample only 50k of data to provide the data format.\n\n\nWe perform SFT for WMT15 German-English dataset using the script:\n```\nbash finetune.sh\n```\n\n### Wait-k-SiLLM\nWe can execute the Wait-k policy with LLM by running the following script:\n```\nbash Wait-k-SiLLM.sh\n```\n\n\n### HMT-SiLLM\nWe can execute the HMT policy with LLM and get the outputs by running the following script:\n```\nbash HMT-SiLLM.sh\n```\n\n\n## Citation\n```\n@misc{guo2024sillm,\n      title={SiLLM: Large Language Models for Simultaneous Machine Translation}, \n      author={Shoutao Guo and Shaolei Zhang and Zhengrui Ma and Min Zhang and Yang Feng},\n      year={2024},\n      eprint={2402.13036},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fictnlp%2Fsillm","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fictnlp%2Fsillm","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fictnlp%2Fsillm/lists"}