11月12月1月2月3月4月5月6月7月8月9月10月
周一周三周五
一年内 258504 次贡献
更少更多
- frostbyte_neo 推送了仓库 github-featured/rasahq--rasa 的 e2e-tests-with-response-keys 分支比较 2 提交 »e0fd6b6e3e Reproduce action prediction issues.83909d65bd Make NLU evaluation in e2e testing consider full retrieval intents.
- frostbyte_neo 推送了仓库 github-featured/nvidia-nemo--speech 的 boxiangw/auto-mcore-fsdp2-exp 分支比较 6 提交 »6063baf017 Apply isort and black reformattingdb1f20143f Add notes for cfsdp194ad6bb6d Apply isort and black reformatting7a9667880e styl95578c041c Add CP loss non_zero_loss_rank
- frostbyte_neo 推送了仓库 github-featured/nvidia-nemo--speech 的 duplex-s2s-new-debugdata 分支84107131fd data debugging service
- frostbyte_neo 推送了仓库 github-featured/nvidia-nemo--speech 的 pagaray/update_PT_24.03 分支e4c4c4ed83 Merge branch 'r2.0.0rc1' into pagaray/update_PT_24.03
- frostbyte_neo 推送了仓库 github-featured/rasahq--rasa 的 e2e-rec-at-20 分支18802e38d9 Merge branch 'main' into e2e-rec-at-20
- frostbyte_neo 推送了仓库 github-featured/rasahq--rasa 的 e2e-md 分支649fd49d89 Merge branch 'e2e' into e2e-md
- frostbyte_neo 推送了仓库 github-featured/nvidia-nemo--speech 的 duplex-realtime-inference-rebase 分支比较 10 提交 »5be67a378b remove stale function_head code74f2632b52 remove Nemotron LLM cache workaround063fc7ff95 sort will be from longest to shortest, following convention for offline modelsabe5ae2388 apply isort and black reformattinge41ccc5d87 change _maybe_apply_forced_turn_taking call signature and make it shared between duplexstt and inference pipeline model wrapper; add tests
- frostbyte_neo 推送了仓库 github-featured/nvidia-nemo--speech 的 boxiangw/add-fsdp2-sharding-strategy 分支比较 2 提交 »066396f23d Apply isort and black reformatting5011e2156a Merge branch 'main' into boxiangw/add-fsdp2-sharding-strategy
- frostbyte_neo 推送了仓库 github-featured/nvidia-nemo--speech 的 pagaray/reduce_storage_usage_in_favor_of_local_storage_part3 分支比较 2 提交 »8c83de6cd4 /tmp path needs to be cleaned & use output folder per PR id22e55c18d5 revert making test optional
