server: add cURL support to server Dockerfiles

[elepedus]

Adding cURL to support --hf-repo and --hf-file params when running through Docker.

[phymbert]

Thanks, I think you can do all images in one PR for server

[elepedus]

Roger that. I'll add to this PR :)

[elepedus]

I'm having some issues testing some of the platform-specific images locally e.g. CUDA/Rocm/Vulkan. Will our CI pipeline catch issues if I just submit a PR?

e.g the CUDA images don't find the CUDA libraries (is it because I'm on a Mac?)

docker run -p 8000:8000 local/llama.cpp:full-cuda-curl -s -m ./gemma-2b-it-q4_k_m.gguf --hf-repo lmstudio-ai/gemma-2b-it-GGUF --hf-file gemma-2b-it-q4_k_m.gguf --port 8000 --host 0.0.0.0 -n 512 
./server: error while loading shared libraries: libcuda.so.1: cannot open shared object file: No such file or directory

e.g.2: the ROCm images error out when installing packages due to GPG signature issues

 => CACHED [5/8] WORKDIR /app                                                                                                                                                                         0.0s
 => [6/8] COPY . .                                                                                                                                                                                    0.0s
 => ERROR [7/8] RUN apt-get update &&     apt-get install -y libcurl4-openssl-dev                                                                                                                     2.3s
------
 > [7/8] RUN apt-get update &&     apt-get install -y libcurl4-openssl-dev:
0.325 Get:1 http://archive.ubuntu.com/ubuntu jammy InRelease [270 kB]
0.325 Get:2 http://security.ubuntu.com/ubuntu jammy-security InRelease [110 kB]
0.424 Get:3 http://archive.ubuntu.com/ubuntu jammy-updates InRelease [119 kB]
0.448 Get:4 http://archive.ubuntu.com/ubuntu jammy-backports InRelease [109 kB]
0.618 Get:5 https://repo.radeon.com/amdgpu/5.6/ubuntu jammy InRelease [5427 B]
0.723 Get:6 https://repo.radeon.com/rocm/apt/5.6 jammy InRelease [2615 B]
0.736 Err:2 http://security.ubuntu.com/ubuntu jammy-security InRelease
0.736   At least one invalid signature was encountered.
1.048 Err:1 http://archive.ubuntu.com/ubuntu jammy InRelease
1.048   At least one invalid signature was encountered.
1.353 Err:3 http://archive.ubuntu.com/ubuntu jammy-updates InRelease
1.353   At least one invalid signature was encountered.
1.656 Err:4 http://archive.ubuntu.com/ubuntu jammy-backports InRelease
1.656   At least one invalid signature was encountered.
1.961 Err:5 https://repo.radeon.com/amdgpu/5.6/ubuntu jammy InRelease
1.961   At least one invalid signature was encountered.
2.264 Err:6 https://repo.radeon.com/rocm/apt/5.6 jammy InRelease
2.264   At least one invalid signature was encountered.
2.290 Reading package lists...
2.333 W: GPG error: http://security.ubuntu.com/ubuntu jammy-security InRelease: At least one invalid signature was encountered.
2.333 E: The repository 'http://security.ubuntu.com/ubuntu jammy-security InRelease' is not signed.
2.333 W: GPG error: http://archive.ubuntu.com/ubuntu jammy InRelease: At least one invalid signature was encountered.
2.333 E: The repository 'http://archive.ubuntu.com/ubuntu jammy InRelease' is not signed.
2.333 W: GPG error: http://archive.ubuntu.com/ubuntu jammy-updates InRelease: At least one invalid signature was encountered.
2.333 E: The repository 'http://archive.ubuntu.com/ubuntu jammy-updates InRelease' is not signed.
2.333 W: GPG error: http://archive.ubuntu.com/ubuntu jammy-backports InRelease: At least one invalid signature was encountered.
2.333 E: The repository 'http://archive.ubuntu.com/ubuntu jammy-backports InRelease' is not signed.
2.333 W: GPG error: https://repo.radeon.com/amdgpu/5.6/ubuntu jammy InRelease: At least one invalid signature was encountered.
2.333 E: The repository 'https://repo.radeon.com/amdgpu/5.6/ubuntu jammy InRelease' is not signed.
2.333 W: GPG error: https://repo.radeon.com/rocm/apt/5.6 jammy InRelease: At least one invalid signature was encountered.
2.333 E: The repository 'https://repo.radeon.com/rocm/apt/5.6 jammy InRelease' is not signed.

e.g.3: Intel server image downloads the model but fails to load it (probably because I'm on a Mac)

|ID|       Device Type|                                         Name|capability|units      |group   |group  |Global mem size|
|--|------------------|---------------------------------------------|----------|-----------|--------|-------|---------------|
| 0|    [opencl:cpu:0]|                       VirtualApple @ 2.50GHz|       3.0|         12|    8192|     64|     8221675520|
| 1|    [opencl:acc:0]|               Intel(R) FPGA Emulation Device|       1.2|         12|67108864|     64|     8221675520|
ggml_backend_sycl_set_mul_device_mode: true
llama_model_load: error loading model: DeviceList is empty. -30 (PI_ERROR_INVALID_VALUE)
llama_load_model_from_file: failed to load model
llama_init_from_gpt_params: error: failed to load model './gemma-2b-it-q4_k_m.gguf'

e.g.4: Vulcan can't install SDK due to GPG key errors

ERROR: failed to solve: process "/bin/sh -c wget -qO - https://packages.lunarg.com/lunarg-signing-key-pub.asc | apt-key add - &&     wget -qO /etc/apt/sources.list.d/lunarg-vulkan-jammy.list https://packages.lunarg.com/vulkan/lunarg-vulkan-jammy.list &&     apt update -y &&     apt-get install -y vulkan-sdk" did not complete successfully: exit code: 100

[phymbert]

You cannot test CUDA image without nvidia container runtime. It's OK.
I worry more about the EditorConfig issue:
https://github.com/ggerganov/llama.cpp/actions/runs/8553404533/job/23436492415?pr=6474

Probably linked with:

• CI maintenance: Update checkout, setup-python and upload-artifact to latest versions #6456 (comment)