ToanDo

XTC in DeepSpeed Compression does not work

1

Hi all, Thanks for great works. I ran some experiments with Deepspeed compression using configs in model_compression/bert. I got some issues: - Size of output model when using DeepSpeedExamples/model_compression/bert/bash_script/XTC/quant_1bit.sh config...

- Add more fine-tuning/generation parameters (gradient_accumulation_steps, save_total_limit, eval_steps, max_grad_norm,...) in config files.

Bloom code

2

Thank you for your great work. I am very interested in Bloom int8 models. Could you please share the code and checkpoints for Int8 Bloom models ?