embedding-server / .env.example
Faysal4200's picture
Upload 45 files
ec855e6 verified
Raw
History Blame Contribute Delete
1.06 kB
# .env.example - copy to .env and edit
# -----------------------------
# Security / Access
# -----------------------------
# Static API key used to guard endpoints (REQUIRED )
EMBED_API_KEY=MY_SECURE_KEY
# Private hugging face token (REQUIRED)
HUGGING_FACE_TOKEN=your_free_token_with_repo_read_access
# -----------------------------
# Model & Device Settings
# -----------------------------
# HuggingFace model identifier
MODEL_NAME=Faysal4200/bge-m3-private
# Target device: "cuda" for GPU, "cpu" for CPU, or leave empty for auto-detection
DEVICE=
# How many texts to process in one model forward pass
BATCH_SIZE=8
# Maximum input sequence length (BGE-M3 supports up to 8192)
MAX_LENGTH=1024
# -----------------------------
# Server & Reliability
# -----------------------------
RETRY_ATTEMPTS=3
RETRY_BACKOFF_SECONDS=2.0
HOST=0.0.0.0
PORT=7860
WORKERS=1
# -----------------------------
# Logging (Premium Style)
# -----------------------------
# Log level: DEBUG, INFO, WARNING, ERROR, CRITICAL
LOG_LEVEL=INFO