Open sommerper opened 1 year ago
The docker image now includes a version for ARM.
Would you be able to test and see if it works?
docker pull ghcr.io/cogentapps/chat-with-gpt:release
docker run -v $(pwd)/data:/app/data -p 3000:3000 ghcr.io/cogentapps/chat-with-gpt:release
I tried the above but the docker image is still amd64, even if you
$ docker pull --platform linux/arm64/v8 ghcr.io/cogentapps/chat-with-gpt:release
you still have
$ docker inspect ghcr.io/cogentapps/chat-with-gpt:release | grep Arch
"Architecture": "amd64",
I pulled the latest and it still throws the error.
~/docker/chatwithgpt
❯ docker-compose pull
[+] Running 14/14
⠿ cogentapps Pulled 50.6s
⠿ 63b65145d645 Already exists 0.0s
⠿ bc59db64d63d Pull complete 23.6s
⠿ c8fd6d18c4a0 Pull complete 24.0s
⠿ eb563838426b Pull complete 24.1s
⠿ 06c6bcf0f0eb Pull complete 24.4s
⠿ d58b902867ad Pull complete 24.6s
⠿ f0e7e688716c Pull complete 24.8s
⠿ ebede05b6017 Pull complete 24.9s
⠿ 511deccdd15b Pull complete 41.1s
⠿ 55364b240ecf Pull complete 48.9s
⠿ c363fe49e713 Pull complete 49.0s
⠿ e0e18c2c0d2b Pull complete 49.2s
⠿ 29d6d12e1f51 Pull complete 49.4s
~/docker/chatwithgpt took 50s
❯ docker-compose up
[+] Running 1/1
⠿ Container chatwithgpt-cogentapps-1 Recreated 0.1s
Attaching to chatwithgpt-cogentapps-1
chatwithgpt-cogentapps-1 | exec /usr/local/bin/docker-entrypoint.sh: exec format error
chatwithgpt-cogentapps-1 exited with code 1
My docker-compose.yml
:
❯ cat docker-compose.yml
version: '3.3'
services:
cogentapps:
volumes:
- './data:/app/data'
ports:
- '3487:3000'
image: 'ghcr.io/cogentapps/chat-with-gpt:release'%
yeah same issue for me too.
The requested image's platform (linux/amd64) does not match the detected host platform (linux/arm64/v8) and no specific platform was requested
same here, can't be running on Raspberry Pi4 exec /usr/local/bin/docker-entrypoint.sh: exec format error
PR #109 demonstrates building the docker locally which could work as a solution. Just tried it and it is working well on a Pi4 running 64-bit Raspberry Pi OS.
The GitHub Action that I was using to deploy the ARM image was taking hours to run (and often failing), probably due to virtualization, so I removed it temporarily. I haven't had an opportunity to write a new one.
PRs for a new GitHub action are welcome, if anyone wants to take a stab at solving this.
Preferably a new workflow, separate from the existing docker build workflow, that builds and deploys in a relatively short period of time.
arm support is very much appreciated.
currently getting:
2023-05-16T23:37:26.962765142Z exec /opt/nvidia/nvidia_entrypoint.sh: exec format error
2023-05-16T23:37:32.788802356Z exec /opt/nvidia/nvidia_entrypoint.sh: exec format error
2023-05-16T23:37:38.597547827Z exec /opt/nvidia/nvidia_entrypoint.sh: exec format error
2023-05-16T23:37:44.453824339Z exec /opt/nvidia/nvidia_entrypoint.sh: exec format error
Hi!
Wanted to test this out on my small home server but I couldn't get it up and running.
I get the error:
I've seen this before when the image wasn't ARM compatible.
Just FYI.