{"id":15697134,"url":"https://github.com/mddct/neural-lm-deprecated","last_synced_at":"2025-10-14T14:30:48.171Z","repository":{"id":37761950,"uuid":"499444241","full_name":"Mddct/neural-lm-deprecated","owner":"Mddct","description":"focus on  fusion on speech recognition","archived":true,"fork":false,"pushed_at":"2023-03-27T11:17:03.000Z","size":76,"stargazers_count":5,"open_issues_count":0,"forks_count":1,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-01-31T07:34:50.648Z","etag":null,"topics":[],"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/Mddct.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":"2022-06-03T08:57:21.000Z","updated_at":"2023-03-27T11:17:24.000Z","dependencies_parsed_at":"2025-01-31T07:32:03.341Z","dependency_job_id":"1cd1b757-89b9-46cf-a40c-188c6c3069bc","html_url":"https://github.com/Mddct/neural-lm-deprecated","commit_stats":null,"previous_names":["mddct/neural-lm-deprecated"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Mddct/neural-lm-deprecated","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Mddct%2Fneural-lm-deprecated","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Mddct%2Fneural-lm-deprecated/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Mddct%2Fneural-lm-deprecated/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Mddct%2Fneural-lm-deprecated/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Mddct","download_url":"https://codeload.github.com/Mddct/neural-lm-deprecated/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Mddct%2Fneural-lm-deprecated/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279019119,"owners_count":26086680,"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","status":"online","status_checked_at":"2025-10-14T02:00:06.444Z","response_time":60,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":[],"created_at":"2024-10-03T19:12:42.139Z","updated_at":"2025-10-14T14:30:48.164Z","avatar_url":"https://github.com/Mddct.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# (deprecated, will reimplement by jax) under development may not work until whole pipeline done\n\n# neural-lm\nfocus on fusion on speech recognition\n\n# Note \n\u003e When a language model is used wide beam searches often yield\n\u003e incomplete transcripts. With narrow beams, the problem is less\n\u003e visible due to implicit hypothesis pruning.\n\nSee if it appears in ctc+lm fusion\n\n# TODO\n- [x] adaptive softmax for large voca (because pytorch offical implementation can't work with torchscript)\n- [ ] onnx support and torchscript\n- [x] gru\n- [x] rnn tie embedding\n- [ ] gru fusion on wenet runtime ctc prefix beam search\n- [ ] transformer-xl with cache\n- [ ] transformer-xl with cache to fusion \n- [ ] mwer training when lm fusion \n- [ ] etc\n\n# reference\n- [Deep Speech: Scaling up end-to-end speech recognition](https://arxiv.org/pdf/1412.5567.pdf) \n- [END-TO-END ATTENTION-BASED LARGE VOCABULARY SPEECH RECOGNITION](https://arxiv.org/pdf/1508.04395.pdf)\n- [On Using Monolingual Corpora in Neural Machine Translation](https://arxiv.org/pdf/1503.03535.pdf)\n- [First-Pass Large Vocabulary Continuous Speech Recognition using Bi-Directional Recurrent DNNs](https://arxiv.org/pdf/1408.2873.pdf)\n- [Towards better decoding and language model integration in sequence to sequence models](https://arxiv.org/pdf/1612.02695.pdf)\n- [END-TO-END ATTENTION-BASED LARGE VOCABULARY SPEECH RECOGNITION](https://arxiv.org/pdf/1508.04395.pdf)\n- [Efficient softmax approximation for GPUs](https://arxiv.org/pdf/1609.04309.pdf)\n- [Using the Output Embedding to Improve Language Models](https://arxiv.org/abs/1608.05859)\n- [Minimum Word Error Rate Training with Language Model Fusion for End-to-End Speech Recognition](https://arxiv.org/pdf/2106.02302.pdf)\n- etc\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmddct%2Fneural-lm-deprecated","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmddct%2Fneural-lm-deprecated","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmddct%2Fneural-lm-deprecated/lists"}