InferX Catalog | IxCore 8B Serve

IxCore 8B Serve

Compact 8B serving template for internal inference workloads.

inferx text text2text text

Log in to deploy

Metadata

Provider

inferx

Modality

text

API type

text2text

Source

huggingface / inferx/ixcore-8b

Created

2026-09-25 03:01:38 UTC

Updated

2026-09-25 03:01:43 UTC

Catalog version

1

Visibility

Published

Specifications

Parameters

8.00B

MoE

No

Max model length

1024

Image

vllm/vllm-openai:v0.16.0

Default Deploy Config

GPU count

1

vRAM

8192 MB

Summary

1xGPU 8192 MB

Recommended Use Cases

chat

Model Spec

{
    "image": "vllm/vllm-openai:v0.16.0",
    "commands": [
        "--model",
        "/tmp/ixmodel",
        "--trust-remote-code",
        "--gpu-memory-utilization",
        "0.05",
        "--max-model-len",
        "1024"
    ],
    "resources": {
        "GPU": {
            "Count": 1,
            "vRam": 8192
        }
    },
    "envs": [],
    "policy": {
        "Obj": {
            "nvidia_replica": 0,
            "min_replica": 0,
            "max_replica": 1,
            "standby_per_node": 1,
            "parallel": 4,
            "queue_len": 100,
            "queue_timeout": 300.0,
            "scalein_timeout": 600.0,
            "scaleout_policy": {
                "WaitQueueRatio": {
                    "wait_ratio": 0.01
                }
            },
            "runtime_config": {
                "graph_sync": false
            }
        }
    },
    "sample_query": {
        "body": {
            "prompt": "hi",
            "max_tokens": 8
        },
        "path": "/",
        "prompt": "hi",
        "apiType": "text2text",
        "dataUrl": "",
        "prompts": [],
        "raw_query": false,
        "max_tokens": 8,
        "loadingTimeout": 90
    }
}