pytorch/serve

add support for SeTFiT models

Open

#2,107 建立於 2023年2月5日

在 GitHub 查看
 (4 留言) (0 反應) (0 負責人)Java (3,844 star) (790 fork)batch import
good first issuetriaged

描述

🚀 The feature

Add support for setfit models which currently is unsupported and throwing Dependency installation stderr: error.

Motivation, pitch

SOTA is quickly moving to few-shot approaches and the SeTFiT library provides a fast and easy way to finetune sentence-transformers-based models with little labeled data. Would strongly appreaciate this feature to be considered.

Alternatives

No response

Additional context

No response

貢獻者指南

技術棧
pythonpytorchjava
領域
backendmachine learning
議題類型
feature
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
3
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
1-2 days
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
mostly clear
前置要求
familiarity with SetFitPyTorch Serve basicsPython package management
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
40
研究方向
Investigate how PyTorch Serve currently supports custom models by examining existing handlers in the `ts/` or `serve/` directory, particularly for models with external dependencies. Review the error 'Dependency installation stderr:' to understand how dependencies are managed. Look at issue comments for any additional context or proposed approaches. The goal is to create a handler that can load and serve SetFit models, similar to existing transformers support.