File size: 909 Bytes
5afb7b3 f9490c0 5afb7b3 f9490c0 5afb7b3 f9490c0 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | # Copy to .env.local and fill in. Without these, the app runs in mock mode.
# Modal-hosted OpenAI-compatible endpoint (vLLM serving Qwen2.5-14B + LoRA).
# Leave empty to force mock mode for the LLM calls.
MODAL_URL=
# Modal proxy auth (Modal-Key and Modal-Secret headers).
MODAL_KEY=
MODAL_SECRET=
# Force mock mode for the LLM even if MODAL_URL is set.
ORACLES_FORCE_MOCK=0
# Override the served-model alias. Defaults to "oracle-wizard-lora" which
# the deployed vLLM serves as Qwen2.5-14B + the custom humor LoRA. Set
# to "llm" to request the bare base model.
# ORACLES_LLM_MODEL=oracle-wizard-lora
# Disable LLM-call trace recording. Tracing is default-on; remove or set
# to 0 to keep it on. Set to 1 to opt out.
# ORACLES_TRACE_DISABLE=0
# Visual mode: "lean" (default; minimal PNGs for fast load) or "full"
# (parallax banner, scene landscapes, all decorative assets).
# ORACLES_VISUAL_MODE=lean
|