Resolving Docker Image Pull Error for `huggingface/text-generation-inference`

[HydraTechnologies-ops]

If you're encountering an error while pulling the latest tag of the huggingface/text-generation-inference Docker image, follow these steps to resolve it:

Steps to Fix

1. Find the Specific Version:

   • Visit the Hugging Face Docker Repository and select a specific version or SHA for AMD instead of using latest.

2. Update the spec.yaml File:

   • Navigate to your spec.yaml file and locate the environment section (line 21).
   • Replace latest with the specific version or SHA. For example, update it as follows:

     environment:
     base:
      registry: ghcr.io
      image: huggingface/text-generation-inference:sha-f3c5d7d@sha256:7c169c536de65dbdb867302aedf9ad0a7472df5d7c7dfd85c080037215b18c16
      build_timestamp: "20231011102429"
      name: TGI
      supported_architectures:
      - amd64
      cuda_version: "11.8"
      description: The huggingface text-generation-interface base container
      entrypoint_script: ""

3. Clear Cache and Rebuild:

   • On the NVIDIA AI Workstation, click on "Clear cache and build" to rebuild the container with the correct image version.

This process should resolve the issue and allow you to successfully pull and use the desired Docker image.

---

[fauziyasin]

If you're encountering an error while pulling the latest tag of the huggingface/text-generation-inference Docker image, follow these steps to resolve it:

Steps to Fix

1. Find the Specific Version:

   • Visit the Hugging Face Docker Repository and select a specific version or SHA for AMD instead of using latest.

2. Update the spec.yaml File:

   • Navigate to your spec.yaml file and locate the environment section (line 21).
   • Replace latest with the specific version or SHA. For example, update it as follows: environment: base: registry: ghcr.io image: huggingface/text-generation-inference:sha-f3c5d7d@sha256:7c169c536de65dbdb867302aedf9ad0a7472df5d7c7dfd85c080037215b18c16 build_timestamp: "20231011102429" name: TGI supported_architectures:
     • amd64 cuda_version: "11.8" description: The huggingface text-generation-interface base container entrypoint_script: ""

3. Clear Cache and Rebuild:

   • On the NVIDIA AI Workstation, click on "Clear cache and build" to rebuild the container with the correct image version.

This process should resolve the issue and allow you to successfully pull and use the desired Docker image.

[arsentievalex]

is there a compatible version for linux/arm64? I'm having the below error:

1 warning found (use docker --debug to expand):
 - InvalidBaseImagePlatform: Base image ghcr.io/huggingface/text-generation-inference:sha-f3c5d7d@sha256:7c169c536de65dbdb867302aedf9ad0a7472df5d7c7dfd85c080037215b18c16 was pulled with platform "linux/amd64", expected "linux/arm64" for current build (line 1)
Containerfile:67