Drax

> I have the same issue, though I am not sure it is only related to VSCode: when I start the jupyter notebook server directly on the remote machine, and...

> when I use chrome, it will appeared in bottom, not adjust itself on the right side as normal. Got the same problem. It turns out this is affected by...

> Hi, Thanks for sharing the code. I am interested in your paper. I followed your instructions and ran the code with the default setting, but I cannot reproduce the...

> > > Hi, Thanks for sharing the code. I am interested in your paper. I followed your instructions and ran the code with the default setting, but I cannot...

Same. I really like the setting of helix but I think it needs to support more options

I feel like this is a model problem. Because I can run llama-13b but can't run merged llama-7b. And I don't know how to solve this.

I solved this. Just add `legacy=True` in ``` tokenizer = transformers.AutoTokenizer.from_pretrained( args.version, cache_dir=None, model_max_length=args.model_max_length, padding_side="right", use_fast=False, legacy=True ) ``` Refer to [link](https://github.com/huggingface/transformers/pull/24565#issue-1780000805)

> try to changed the version to "gradio==3.37.0" or `pip install gradio_client==0.2.7` according to [UX-Decoder/Segment-Everything-Everywhere-All-At-Once#59](https://github.com/UX-Decoder/Segment-Everything-Everywhere-All-At-Once/issues/59) and [AUTOMATIC1111/stable-diffusion-webui#11855](https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/11855) Thanks for your work! Directly using `pip install -r requirements.txt` results in ""llava'...

Same errors on refcoco dataset. ```NotImplementedError: `.as_dataset()` not implemented for ArrayRecord files. Please, use `.as_data_source()`.```

Anyway, one thing I do to solve this is add the following line: ``` builder = tfds.builder('ref_coco/refcocog_umd') builder.info.set_file_format(tfds.core.FileFormat.PARQUET, override=True, override_if_initialized=True) builder.download_and_prepare() ref_ds = tfds.load('ref_coco/refcocog_umd', split='validation') ```