Why the output looks very bad

[xienan0326]

trafficstars

transformers 4.28.1 fschat 0.2.3

i use vicuna-7b-delta-v1.1 python fastchat/model/apply_delta.py --base-model ../llama-7b-hf-v1.1/ --target vicuna-7b --delta vicuna-7b-delta-v1.1/ python fastchat/serve/cli.py --model_path ./vicuna-7b

ll ./vicuna-7b total 13161628 drwxr-xr-x 2 root root 279 Apr 22 10:24 ./ drwxr-xr-x 11 root root 267 Apr 22 10:08 ../ -rw-r--r-- 1 root root 20 Apr 22 10:24 added_tokens.json -rw-r--r-- 1 root root 545 Apr 23 07:02 config.json -rw-r--r-- 1 root root 132 Apr 23 07:02 generation_config.json -rw-r--r-- 1 root root 9976634558 Apr 23 07:02 pytorch_model-00001-of-00002.bin -rw-r--r-- 1 root root 3500315539 Apr 23 07:02 pytorch_model-00002-of-00002.bin -rw-r--r-- 1 root root 26788 Apr 23 07:02 pytorch_model.bin.index.json -rw-r--r-- 1 root root 411 Apr 23 07:02 special_tokens_map.json -rw-r--r-- 1 root root 499723 Apr 23 07:02 tokenizer.model -rw-r--r-- 1 root root 727 Apr 23 07:02 tokenizer_config.json

[xienan0326]

llama-7b-hf-v1.1

llama-7b-hf-v1.1 mean use transformers 4.28.1 trans2hf llama-7b

[hrwise-nlp]

I found that my output is mostly "<s> </s>", do not know why.

[zhisbug]

1. use the latest vicuan weight v1.1
2. update your transformer to latest version
3. update your fastchat to latest version

Will solve the problem.