which version about llama-2 ? #71
Description
i use huggyllama/llama-7b, but error
python instruct is python mllm/models/shikra/apply_delta.py --base /data/private/projects/base_model/llama-7b --target output/shikra-7b --delta /data/private/projects/base_model/shikra-7b-delta-v1
log following,
OSError: Unable to load weights from pytorch checkpoint file for '/data/private/projects/base_model/shikra-7b-delta-v1/pytorch_model-00001-of-00002.bin' at '/data/private/projects/base_model/shikra-7b-delta-v1/pytorch_model-00001-of-00002.bin'. If you tried to load a PyTorch model from a TF 2.0 checkpoint, please set from_tf=True.