wanzhixiao
wanzhixiao
AttributeError: 'MSS' object has no attribute 'get_pixes' in module readScreen line 113 sct.get_pixes(mon)
as title, i didnt understand either 1 or 0 denoting an edge or a non-edge? what is means? valid.txt: Each line represents an edge or a non-edge, which contains four...
A doubt
Why GAMN good at long-term forecasting, but not so obvious for short-term forecasting. it is because Temporal attention? Looking forward to your answer,thanks.
使用Paddle Serving部署深度模型,看https://github.com/PaddlePaddle/Serving/tree/develop/examples/Pipeline/simple_web_service 例子,使用命令 python3 -m paddle_serving_server.serve --model uci_housing_model --thread 10 --port 9292 其中uci_housing_model这个文件目录是怎么生成的呢
epoch [7] |███████████████████ | 23/30 loss 1.50 eval ORG recall 1.00 precision 1.00 f1 1.00 PER recall 1.00 precision 1.00 f1 1.00 -------------------------------------------------- epoch [7] |████████████████████ | 24/30 loss 0.59...
cpu部署
``` model = FlagModel(fmodel_path, query_instruction_for_retrieval="为这个句子生成表示以用于检索相关文章:", use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation ``` 怎么把模型移到cpu上呢? use_fp16=True 或者False都会使用GPU
关于性能问题
服务端用bge-small做文本embedding, 任务是短文本匹配,使用gpu推理,单样本(20字符长度文本对)做推理,没有用batch, 平均耗时约20ms。但是接口调用方的请求比较快,每秒钟有20个请求,每个请求有数百条匹配文本,请求数越多,模型的平均推理耗时越长。想问下有没有可以提升模型推理速度的方法?看issue中提到的几个方法 1. embedding降维,用pca或者加一层全连接,把768/512降低到 128维; 2. 转onnx部署 3. 用batch的形式做embedding 4. 使用hugging face的text-embedding-inference 的docker形式部署 方法4应该是比较好的,bge的微调后的FlagModel模型,好像有不能用4的方式部署了?有什么方法解决吗
### System Info CUDA Version: 12.0 , with A10 GPU CentOS: 7.9 ### Information - [X] Docker - [ ] The CLI directly ### Tasks - [X] An officially supported...
Hi, i want to know what is the function of this project? it is a bullet website like yotube?