mirror of
https://github.com/trustgraph-ai/trustgraph.git
synced 2026-07-06 11:52:10 +02:00
74cc8a4685
git-subtree-dir: ai-context/trustgraph-templates git-subtree-split: 42a5fd1b678f32be378062e30451e2052ccb95dd
77 lines
2.5 KiB
Jsonnet
77 lines
2.5 KiB
Jsonnet
local base = import "base/base.jsonnet";
|
|
local images = import "values/images.jsonnet";
|
|
|
|
{
|
|
|
|
with:: function(key, value)
|
|
self + {
|
|
["vllm-service-" + key]:: value,
|
|
},
|
|
|
|
"vllm-service-model":: "teknium/OpenHermes-2.5-Mistral-7B",
|
|
"vllm-service-cpus":: "8.0",
|
|
"vllm-service-memory":: "16G",
|
|
|
|
"vllm-service" +: {
|
|
|
|
create:: function(engine)
|
|
|
|
local vol = engine.volume("vllm-storage").with_size("20G");
|
|
|
|
local container =
|
|
engine.container("vllm-service")
|
|
.with_image(images["vllm-service-intel-xpu"])
|
|
.with_command([
|
|
"python",
|
|
"-m",
|
|
"vllm.entrypoints.openai.api_server",
|
|
"--model",
|
|
$["vllm-service-model"],
|
|
"--dtype=float16",
|
|
"--device=xpu",
|
|
"--enforce-eager",
|
|
"--port",
|
|
"8899",
|
|
"--block-size",
|
|
"64",
|
|
"--gpu-memory-util",
|
|
"0.85",
|
|
"--trust-remote-code",
|
|
"--disable-sliding-window",
|
|
])
|
|
.with_environment({
|
|
HF_TOKEN: $["hf-token"],
|
|
VLLM_USE_V1: "1",
|
|
W_LONG_MAX_MODEL_LEN: "1",
|
|
VLLM_WORKER_MULTIPROC_METHOD: "spawn",
|
|
})
|
|
.with_privileged(true)
|
|
.with_device("/dev/dri", "/dev/dri")
|
|
.with_ipc("host")
|
|
.with_capability("SYS_NICE")
|
|
.with_limits(
|
|
$["vllm-service-cpus"], $["vllm-service-memory"]
|
|
)
|
|
.with_reservations(
|
|
$["vllm-service-cpus"], $["vllm-service-memory"]
|
|
)
|
|
.with_port(8899, 8899, "vllm")
|
|
.with_volume_mount(vol, "/data");
|
|
|
|
local containerSet = engine.containers(
|
|
"vllm-service", [ container ]
|
|
);
|
|
|
|
local service =
|
|
engine.service(containerSet)
|
|
.with_port(8899, 8899, "vllm");
|
|
|
|
engine.resources([
|
|
vol,
|
|
containerSet,
|
|
service,
|
|
])
|
|
|
|
},
|
|
|
|
}
|