Junyang Lin comments

Results 173 comments of


                                            Junyang Lin

如何把qwen1.5部署为api形式？你们的其他项目qwen-agent，想要调api形式的qwen1.5

I'll promote this adaptation of Qwen1.5 to domestic GPUs, but it takes time as it is not really under my control... I am not an expert in this. oai api,...

想问一下微调之后，Qwen1.5的base版本，{'input_ids': [None, 32, 25, 220, 108386, 35946, 108850, 102815, 94444, 20412, 99459, 55502, 39953, 73296, 104673, 99901, 24156, 198, 33, 25, 69372, 98749, 51154, 25177, 99720, 102289, 108617, 198, 22043, 264, 3239, 362, 323, 264, 21085, 425, 11, 8253, 3425, 279, 21085, 5610, 458, 4226, 311, 279, 3239, 553, 8241, 264, 19639, 315, 2987, 364, 9454, 6, 476, 364, 2753, 4427, 198, 22043, 264, 3239, 362, 323, 264, 21085, 425, 11, 8253, 3425, 279, 21085, 5610, 458, 4226, 311, 279, 3239, 553, 8241, 264, 19639, 315, 2987, 364, 9454, 6, 476, 364, 2753, 4427], 生成的token开头变成none了，如何去解决。

and

Sync huggingface modifications of qwen Moe model

bo is our member. yes, i just noticed that this is merged into transformers. somehow it is not necessary cuz we do not have this setup for our models, but...

DATA="path_to_data",请问数据集在哪下载？

https://qwen.readthedocs.io/zh-cn/latest/training/SFT/example.html

DATA="path_to_data",请问数据集在哪下载？

no. just organize the data yourself following the doc.

pexpect.pxssh.ExceptionPxssh: Could not establish connection to host

you need to restart your container for that. btw the very recent pr https://github.com/OpenDevin/OpenDevin/pull/891 solved this problem also.

make run error

this is a similar issue to the previous one. try the latest code with this pr https://github.com/OpenDevin/OpenDevin/pull/891 . and remember to restart the docker container.

Doomsday is coming Gotta do fast

We are going to design a contribution guideline then. Now people are just rushing

What has been updated on modelscope for qwen1.5-72b-chat? 2024/03/07

We found some model repos' checkpoints in HF and MS not aligned and we fixed.

where BAAI/bge-small-en-v1.5 should be saved?

i think this is a network issue. try `export HF_ENDPOINT=https://hf-mirror.com` and see if it helps. the model should be saved in your hf cache, and then you can load the...