Dragon

修复了，感觉这样才有意义 ``` def CROSS(S1,S2): CROSS_BOOL=IF(S1>S2, True ,False) return (COUNT(CROSS_BOOL>0,2)==1)*CROSS_BOOL ``` buy_signal = CROSS(ma10, ma30) # 找出10日线上穿30日线的点 sell_signal = CROSS(ma30, ma10) # 找出30日线上穿30日线的点 如果使用原版，不修复 buy_signal = CROSS(ma10, ma30) # 找出10日线上穿30日线的点 sell_signal...

> Seen this problem as well. A solution that worked for me was to set the number of intra_op_num_threads to something corresponding to the number of available cores: > >...

> > > @haotian-liu I have the same question, can you share which result to present? > > > @OliverLeeXZ and @g-h-chen were you able to find the correct strategy?...

> 看下infer文件夹下的日志是否有报错，如果报错是 MKL_THREADING_LAYER=INTEL is incompatible with libgomp.so.1 library 那么需要在run.py文件前设置， import os os.environ["MKL_SERVICE_FORCE_INTEL"] = '1' os.environ["MKL_THREADING_LAYER"] = '1' > > 设置后如果卡在 [opencompass.openicl.icl_inferencer.icl_gen_inferencer] [INFO] Starting inference process... | 0/799 [00:00

same issue that pip cannot work. what should I do to use pip

> export MKL_THREADING_LAYER=GNU Hi @goosey-geo Thanks for sharing the information. I try to use export MKL_THREADING_LAYER=GNU, but I get stucked in waiting" 0%| | 0/1 [00:00

> For the base model, we recommend using perplexity (ppl) for evaluation in multi-choice questions. > > ``` > python -u run.py --datasets commonsenseqa_ppl --hf-num-gpus 1 --hf-type base --hf-path meta-llama/Meta-Llama-3-8B...

这个问题没彻底解决，最新版本的transformer （4.45） 之后的版本还是会报，大佬们

Thank you for your prompt reply! I’m also curious—how much training time does Vision‑SR1 require for the cold‑start (SFT) stage and the subsequent GRPO (RL) stage, respectively?

Thanks for your reply! Could you please give me some suggestions about how to cut off? use ctrl+c? and how to choose the stopping points. And if the 47K data...