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

Update license

Upload pytorch_model.bin

Update slack link

Update decode method in tokenizer

Add support for parallel quantization on Mac

Remove assert in load_cpu_kernel

Sync with chatglm-6b

Remove pytorch_model.bin.index.json

Update slack link

Add pytorch_model.bin.index.json

Add assertion when loading cpu and cuda kernel fails

Add assertion when loading cpu and cuda kernel fails

Merge branch 'dev'

Update CPU kernel loading method

Fix gmask

Add empty_init option

Update README.md

Fix eos token in tokenizer

Update dependency

Fix attention score on mps

Fix logit processor

Merge branch 'slim' of https://huggingface.co/THUDM/chatglm-6b-int4 into slim

Fix embedding quantization

Upload pytorch_model.bin

Slim embedding

Fix bugs when compiling cpu kernels

Drop icetk dependency

Fix position ids expand

Synchronize with chatglm 6b repo

Fix parallel cpu kernel

Fix bugs in quantization when loading kernels

Fix Chinese punctuation

Update README.md

Update README.md

Update README.md

init commmit

initial commit