paulcx

Testing Dataset

14

Could you provide a small training data sets even just for running the codes as the dataset is quite difficult and time-consuming to get. Thx

How does this algorithm apply on Chinese text?

2

Any hint? It will be grateful if you could provide some examples.

Something wrong with the features exporting

5

end preprocessing (18L, 1L, 208L, 208L, 208L) Traceback (most recent call last): File "main.py", line 62, in test_detect(test_loader, nod_net, get_pbb, bbox_result_path,config1,n_gpu=config_submit['n_gpu']) File "/home/xx/py/DSB2017-master/test_detect.py", line 55, in test_detect **output = split_comber.combine(output,nzhw=nzhw)**...

是否考虑多卡并行训练？

1

目前只能在一张卡训练。

overfit and why?

3

So far all my attempts, with different models, sizes, and datasets have led to one issue: the evaluation loss keeps increasing. see my log 

llm finetuning is overfitting?

5

So far all my attempts, with different models (bloom, gpt), sizes, accelerate framework, datasets have led to one issue: the evaluation loss keeps increasing. plz see my log (deepspeed)

Support Yi-VL-6B/34B

5

The Yi-VL adopts llava but with silightly different in weights and inference. see [disscusion](https://huggingface.co/01-ai/Yi-VL-34B/discussions/3) hf repo: https://huggingface.co/01-ai/Yi-VL-6B https://huggingface.co/01-ai/Yi-VL-34B

Long waiting time for shard loading

5

### System Info ghcr.io/huggingface/text-generation-inference:sha-7766fee ### Information - [X] Docker - [ ] The CLI directly ### Tasks - [X] An officially supported command - [ ] My own modifications ###...

support docker deployment

8

Could you please add support for docker deployment to streamline setting up and running the project?

Support ORPO

1

Is it considered a new feature ORPO？ ORPO, a technique that replaces SFT+DPO/PPO was released recently. I saw @_philschmid's post regarding it yesterday. Gave ORPO a shot with phi-2 and...