Yue Ying
Yue Ying
fix typo
a package is missing
I really can't understand the proof of _Per-decision Importance Sampling_ in section 5.9 In my opinion, _roi(t:t+k-1)*R(t+k)_ depends on _S(t), A(t),...., S(t+k-1), A(t+k-1)_ and _roi(t+k:T-1)_ depends on _S(t+k), A(t+k),...., S(T-1),...
type of --pretrain should not be bool you can‘t change its value through command
逻辑能力有待加强

### Description  ### To Reproduce _No response_ ### Installation None ### Environment  ### Additional Information _No response_
bge-m3的稀疏向量应该是通过分词得到的吧,未来有可能尝试引入SPLADE方法吗?目前还没看到开源的中文SPLADE模型,不知道效果咋样
### Model description I try to deploy this SPLADE model [https://huggingface.co/naver/neuclir22-splade-zh/blob/main/config.json](url) whose positional encoding type is relative-key. But the error message says this type of positional encoding is not supported....
[https://github.com/QwenLM/Qwen-Agent/blob/main/examples/function_calling.py](url) 请问在使用function call的时候应该怎么设置system message和few-shot?可以给一些指导或者例子吗
Hi, thank you for your great work. I'm new to Prometheus and Grafana and I'd like to use sglang with them. After I set up according to the [document](https://sgl-project.github.io/references/production_metrics.html#../examples/monitoring/grafana.json), the...