sheiy

support maven system dependency

8

### Describe your motivation support maven system scope dependency ```xml taobao sdk-java auto_1709828590347-20240314 system ${project.basedir}/lib/taobao-sdk-java-auto_1709828590347-20240314.jar ``` ```xml org.springframework.boot spring-boot-maven-plugin true ``` ### Describe the solution you'd like _No response_ ###...

TritonModelException: inference request batch-size must be <= 128 for 'cls_pp'

3

- 系统环境/System Environment： - 版本号/Version：Paddle： - PaddleOCR： 问题相关组件/Related components： - 运行指令/Command Code： ``` FROM registry.baidubce.com/paddlepaddle/fastdeploy:1.0.7-gpu-cuda11.4-trt8.5-21.10 COPY ./models-gpu.tar /ocr_serving/ WORKDIR /ocr_serving RUN tar -xf models-gpu.tar RUN rm models-gpu.tar EXPOSE 8000 CMD...

Qwen1.5-72B-Chat-GPTQ-Int4 能否在4*16G V100上跑起来？

4

`python -m vllm.entrypoints.openai.api_server --served-model-name Qwen1.5-72B-Chat --model Qwen1.5-72B-Chat-GPTQ-Int4 --dtype auto --tensor-parallel-size 4 --max-model-len 1024 --gpu-memory-utilization 0.99 --quantization gptq` 启动会报内存不足 ``` INFO 04-27 23:45:33 api_server.py:151] vLLM API server version 0.4.1 INFO 04-27...

fix: vite 使用@ 编译会报错

11

**这个 PR 做了什么?** (简要描述所做更改) vite 使用@ 编译会报错 **这个 PR 是什么类型?** (至少选择一个) - [X] 错误修复(Bugfix) issue: fix # - [ ] 新功能(Feature) - [ ] 代码重构(Refactor) - [ ] TypeScript 类型定义修改(Typings)...

support China

### Self Checks - [x] I have searched for existing issues [search for existing issues](https://github.com/langgenius/dify/issues), including closed ones. - [x] I confirm that I am using English to submit this...