added TRRT_LLM dockerfile reference for pods

Author: Murat Magomedov

Diff for: pod-image/Dockerfile

@@ -0,0 +1,18 @@
+# Start with Runpod Python 3.10 base image
+FROM runpod/pytorch:2.2.1-py3.10-cuda12.1.1-devel-ubuntu22.04
+
+# Avoid prompts from apt
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Install system dependencies and Python
+RUN apt-get update -y && \
+    apt-get install -y git libopenmpi-dev && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/*
+
+# Update Torch
+RUN pip install torch==2.4.0 torchvision==0.19.0
+
+# Install TensorRT-LLM
+RUN pip install tensorrt_llm==0.13.0 -U --extra-index-url https://pypi.nvidia.com
+

Diff for: pod-image/README.md

@@ -0,0 +1,3 @@
+Ready to use TensorRT-LLM image for a Pod, just pull image from: *muratmagomedov/runpod-trrt:v0.13.0*
+Test: `python3 -c "import tensorrt_llm"`
+To build yourself: `docker build -t <username/image> --platform linux/amd64 .`