GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Lumina-ERP
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
Lumina-ERP
Model Tree
Lumina-ERP/lumina-tsql-8b
Input Modalities
Output Modalities
Supported Functionality
FP8 dynamic quantization of Lumina-ERP/lumina-tsql-8b (weights FP8, activations dynamic per token, lm_head kept in bf16). Roughly half the memory of the bf16 release; serves comfortably on a single 24 GB GPU with vLLM.
bash
vllm serve Lumina-ERP/lumina-tsql-8b-FP8-dynamic --max-model-len 16384
Use the system prompt shipped with the main repo (system_prompt.txt), temperature 0, and {"chat_template_kwargs": {"enable_thinking": false}}. See the main model card for benchmarks, training method, schema-constrained decoding tooling, and limitations. Built by Lumina ERP.