Rayan

> Can you try with XGBoost 1.0? That was the last version I tested this against. As I mentioned on main post, [XGBoost 1.4.2](https://github.com/dmlc/xgboost/tree/v1.4.2) works with same compile option.

good commit

Do you have some plan to fixed this issue. for example, you may ignore namespace or support limited namespace.

> https://github.com/uber/petastorm/pull/754 this CR is a work-in-progress without implementing new dataset interface

drop innounp.exe to bin/innounp.exe  which downloading via this meta ```json { "version": "1.73", "url": "https://www.rathlev-home.de/tools/download/innounp-1.zip", "hash": "db670ce576bb89467358b6104be528a006b2e850a215703aadc485592cfc96ad", "bin": "innounp.exe", } ```

@cam-ai Which Docker image do you employ?I test oneflowinc/oneflow:0.9.1.dev20240202-cuda12.1 works fine ```shell docker run -it --rm --runtime=nvidia --privileged --network host oneflowinc/oneflow:0.9.1.dev20240202-cuda12.1 python3 -m oneflow --doctor ``` It appears that this...

@kiranzo as ISSUE [Validate_schema keyword not supported yet](https://github.com/uber/petastorm/issues/758#issuecomment-1785925528) describe ```shell pip install pyarrow==10.0.1 ```

> So when can we upgrade the pyarrow version used in the project? 🤦 We replace petastorm with torch.utils.data.DataLoader, distributed FileReader can borrow from [parquet_dataset](https://github.com/alibaba/TorchEasyRec/blob/master/tzrec/datasets/parquet_dataset.py ) or use [ray.data.read_parquet](ray.data.read_parquet)

> Hello, Thank you for submitting this issue. Please refer to our basic use: https://docs.aws.amazon.com/sdk-for-cpp/v1/developer-guide/basic-use.html > > SDK was initially designed around global context pointers (such as memory allocator, loggers,...

openjdk version "11.0.24" 2024-07-16 OpenJDK Runtime Environment (build 11.0.24+8-post-Ubuntu-1ubuntu320.04) still failed on master 707e4614f31c7747bdca0778128f502d72227d67