Serve as model id glm-5.2 (keep GLM-5.2-NVFP4-REAP-469B as alias)

Clients prompting model=glm-5.2 got 404 because the only served name was the long
checkpoint name. Make glm-5.2 the primary --served-model-name; the full name and
GLM-5.2 still resolve as aliases. Update chat.sh default + README API examples.