Skip to content

Corrected steps for TRTLLM deployment over Triton #746

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Corrected steps for TRTLLM deployment over Triton #746

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 12 additions & 10 deletions docs/llama.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,13 +1,23 @@
## End to end workflow to run llama 7b

0. Make sure that you have initialized the TRT-LLM submodule:
0. Make sure that you have downloaded and initialized the TRT-LLM submodule:

```bash
git clone https://github.com/triton-inference-server/tensorrtllm_backend.git
git lfs install
git submodule update --init --recursive
```
1. Start the Triton Server Docker container:

1. (Optional) Download the LLaMa model from HuggingFace:
```bash
# Replace <yy.mm> with the version of Triton you want to use.
# The command below assumes the the current directory is the
# TRT-LLM backend root git repository.

`docker run --rm -ti -v ${PWD}:/mnt -w /mnt -v ${HOME}/.cache/huggingface:/root/.cache/huggingface --gpus all nvcr.io/nvidia/tritonserver:\<yy.mm\>-trtllm-python-py3 bash`


2. (Optional) Download the LLaMa model from HuggingFace:

```bash
huggingface-cli login
Expand All @@ -19,14 +29,6 @@ huggingface-cli download meta-llama/Llama-2-7b-hf
>
> Make sure that you have access to https://huggingface.co/meta-llama/Llama-2-7b-hf.

2. Start the Triton Server Docker container:

```bash
# Replace <yy.mm> with the version of Triton you want to use.
# The command below assumes the the current directory is the
# TRT-LLM backend root git repository.

docker run --rm -ti -v `pwd`:/mnt -w /mnt -v ~/.cache/huggingface:~/.cache/huggingface --gpus all nvcr.io/nvidia/tritonserver:\<yy.mm\>-trtllm-python-py3 bash
```

3. Build the engine:
Expand Down