Jinkai

> embedding-reader I see, that would be great. By the way, in the paper "_DataComp: In search of the next generation of multimodal datasets_" I see you guys are using...

tried setting ENDPOINT via os.environ and the error becomes different : > 23/07/15 05:03:41 INFO DAGScheduler: Job 2 finished: count at :0, took 0.282508 s count: 5177571 Starting the downloading...

> did you fix this ? seems like an issue with fsspec nope, later on, I tried solve this issue by download to NAS first and then start from there.

same here, tried execute img2dataset command line and near the end (by monitoring network ,almost no receival) ,the process just would not exit: USER PID %CPU %MEM VSZ RSS TTY...

Sorry to get back to you late, but I was able to bypass this issue by modifying the datacomp source code as follows: ``` diff --git a/eval_utils/retr_eval.py b/eval_utils/retr_eval.py index 3c19917..647edf7...

> 已开源 在哪个repo呀， 目前尝试执行 downstream finetune时候报错： Traceback (most recent call last): File "downstream/run_cls_distributed_gpt3.py", line 26, in from models.distributed_gpt3 import DistributedGPT3_Cls ModuleNotFoundError: No module named 'models.distributed_gpt3'

> evalcap 弱问楼主，你 from dataset import create_dataset, create_sampler, create_loader 这一行能执行么，也是在run_caption_distributed_gpt3.py 里的，我找不到这几个 helper function 的定义。