Fix use_cache=False

Fix tokenizer for transformers 0.34

Update use case

Update quantized gemm kernel

Fix prefix projection

Fix tokenization space

Add prefix prompt

Update slack link

Add default model type

Update modeling_chatglm.py

Add pad_token_id

Fix arange_cpu

Update requirement

Add cpu kernel

Merge branch 'main' of https://huggingface.co/THUDM/chatglm2-6b-int4

Init commit

Upload tokenizer.model with huggingface_hub

Upload pytorch_model.bin with huggingface_hub

initial commit