Closed xuyizhan closed 7 months ago
Hi there,
[Confgiuration] xFasterTransformer version: 1.4.0 Docker version: Docker version 26.0.0, build 2ae903e Kernel: 5.15.0
[Steps] docker pull intel/xfastertransformer:latest In docker container xfastertransformer run bash run_benchmark.sh -m llama-2-7b -d bf16 -s 1 -bs 1 -in 1024 -out 128 -i 10
bash run_benchmark.sh -m llama-2-7b -d bf16 -s 1 -bs 1 -in 1024 -out 128 -i 10
[Description] The new behaviour of LlamaTokenizer (with self.legacy = False) requires the protobuf library but it was not found in your environment.
self.legacy = False
pip install protobuf will resolve the issue, please help to install necessary library and update the docker image, thanks.
pip install protobuf
thx~ yizhan, we will fix it at the next release. @Duyi-Wang
protobuf will be added into requirements.txt by this PR #284 and in next docker release.
protobuf
Hi there,
[Confgiuration] xFasterTransformer version: 1.4.0 Docker version: Docker version 26.0.0, build 2ae903e Kernel: 5.15.0
[Steps] docker pull intel/xfastertransformer:latest In docker container xfastertransformer run
bash run_benchmark.sh -m llama-2-7b -d bf16 -s 1 -bs 1 -in 1024 -out 128 -i 10
[Description] The new behaviour of LlamaTokenizer (with
self.legacy = False
) requires the protobuf library but it was not found in your environment.pip install protobuf
will resolve the issue, please help to install necessary library and update the docker image, thanks.