Michael Wyatt

@tomeras91 I can confirm that I'm able to reproduce this error. I don't think it has anything to do with `MAX_OUT_TOKES`. @RezaYazdaniAminabadi could you take a look at this?

@gtvforever you can pass multiple items to the pipeline like so: `outputs = generator(["DeepSpeed is the", "It's a DeepSpeed kind of summer"], do_sample=True, min_length=50)` If you are looking to the...

Could you share the output of `nvcc --version`? Also, does the install work with just `pip install deepspeed` or `pip install git+https://github.com/microsoft/DeepSpeed.git`?

@kisseternity Are you able to run if you install from pypi (`pip install deepspeed==0.6.5`)? I've seen a similar issue in the past and it was likely due to the symlinks...

@kisseternity thanks for sharing! I think the issue is stemming from symbolic links we create in `setup.py` - but I'm unable to reproduce this behavior in my own environment. Could...

@kisseternity could you please share what issue you are having when building from source?

The fp16 Bloom weights are now supported. Int8 models are also supported, but currently the DeepSpeed sharded int8 weights for the Bloom model will throw an error. I'm working on...

@mayank31398 @TahaBinhuraib I finally found the time to fix #69 so that it works with int8. You no longer need to download the sharded checkpoint files separately and MII will...

Hi @pohunghuang-nctu we currently do not support multi-node deployments, but this feature is on our radar as we continue development of MII.

@reymondzzzz I am able to reproduce these results. I'm not exactly sure why the outputs don't match. If I use the HuggingFace pipeline API, the problem goes away. Can you...