Guoheng Sun

When using Accelerate for data parallel inference, using different numbers of GPUs results in different results

2

Hi, @haileyschoelkopf Thank you for your awsome open-source work. We have been evaluating using `lm-eval` and noticed that when using `accelerate` for data parallel inference, the number of GPUs utilized...

The loss curve exhibits a stair-step pattern of descent.

4

I used the following setting to train my own dataset with lora, but I found that the loss curve exhibits a stair-step pattern of descent. It appears that the loss...

感谢你们贡献如此优秀的开源项目。 在[训练数据demo](https://github.com/PKU-YuanGroup/ChatLaw/blob/main/data/demo_data_%E6%B3%95%E5%BE%8B%E5%92%A8%E8%AF%A2.jsonl)中，meta_instruction中的指令并不通顺（你一个名叫） `"meta_instruction": "你一个名叫ChatLAW` 请问这是疏忽还是有意为之，这会影响模型的效果吗？

Thank you for your contribution. I was trying to access the source data of GitHub, but suddenly https://the-eye.eu/public/AI/pile_preliminary_components/ is no longer accessible. A few days ago, I was able to...

I saw in [Domino](https://arxiv.org/pdf/2409.15241) that the code would be released here. Could you let me know when will the code be released to the public?