Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,71 @@
+---
+license: other
+pipeline_tag: text-generation
+tags:
+- qwen
+- qwen2
+- lora
+- vllm
+- open-webui
+- korean
+- coding
+---
+
+# 7bcustom-model
+
+This is a public deployment package for a local DGX AI Factory coding assistant runtime.
+
+## Model
+
+- Public name: `7bcustom-model`
+- Runtime served name: `dgx-stable-current`
+- Base family: Qwen2 7B Instruct class
+- Runtime: vLLM OpenAI-compatible API
+- Open-WebUI compatible: yes
+
+## Deployment status
+
+This public release is based on the locally validated stable deployment.
+
+```text
+average_score: 97.75
+pass_70_plus: 20/20
+strong_85_plus: 20/20
+critical_fail_count: 0
+decision: DEPLOY_CANDIDATE
+```
+
+## Runtime policy
+
+The local production runtime uses router/template safeguards for deterministic operational answers:
+
+- Linux guarded prompt
+- vLLM medium prompt
+- CUDA check template
+- LoRA/stable/rejected policy template
+
+## vLLM example
+
+```bash
+python -m vllm.entrypoints.openai.api_server \
+  --model ./ \
+  --served-model-name 7bcustom-model \
+  --dtype float16 \
+  --host 0.0.0.0 \
+  --port 8000 \
+  --max-model-len 1536 \
+  --gpu-memory-utilization 0.50 \
+  --max-num-seqs 8
+```
+
+## Open-WebUI
+
+```text
+Base URL: http://<host>:8000/v1
+Model   : 7bcustom-model
+API Key : dummy
+```
+
+## Notes
+
+This repository is intended as a public model/runtime release record. Local absolute paths, private operational logs, and preservation tarballs are not required for public usage.