TGI does not reference model weights

Beginners
1

My server’s proxy does not allow me to go to Hugging Face. So, I downloaded Mistral 7B weights from GitHub to another computer, sftpd it over to the server, and untarred the contents,

$ tar -tvf mistral-7B-Instruct-v0.3.tar
-rw-rw---- nobody/nogroup 14496078512 2024-05-09 10:47 consolidated.safetensors
-rwxrwxrwx nobody/nogroup         202 2024-05-20 07:09 params.json
-rwxrwxrwx nobody/nogroup      587404 2024-05-20 07:09 tokenizer.model.v3

, to $HF_HOME/hub/models--mistralai-Mistral-7B-Instruct-v0.3/. However, when I

docker run -d     
	--name=tgi-mistral-7b   
	--env HF_HUB_OFFLINE=1   
	--env HUGGING_FACE_HUB_TOKEN=$HUGGING_FACE_HUB_TOKEN     
	--env http_proxy=$http_proxy     
	--env https_proxy=$https_proxy     
	--env MAX_BATCH_TOTAL_TOKENS=32000     
	--env MAX_BATCH_PREFILL_TOKENS=16000     
	--env MAX_TOTAL_TOKENS=32000     
    --gpus all     
	--shm-size 1g     
	-p 8080:80     
	-v $volume:/artifactory.my_company.com/ghcr.io/huggingface/text-generation-inference:1.4.5 
	--model-id mistralai/Mistral-7B-Instruct-v0.3

, I get

huggingface_hub.utils._errors.EntryNotFoundError: No .bin weights found for model mistralai/Mistral-7B-Instruct-v0.3 and revision None.

How do I place/structure these weights/files so that TGI can reach them?