using LLM for sentiment analysis
This commit is contained in:
38
model-runner.env
Normal file
38
model-runner.env
Normal file
@@ -0,0 +1,38 @@
|
||||
# Docker Model Runner Environment Configuration
|
||||
# Copy values to your main .env file or use with --env-file
|
||||
|
||||
# AMD GPU Configuration
|
||||
HSA_OVERRIDE_GFX_VERSION=11.0.0
|
||||
GPU_LAYERS=35
|
||||
THREADS=8
|
||||
BATCH_SIZE=512
|
||||
CONTEXT_SIZE=4096
|
||||
|
||||
# API Configuration
|
||||
MODEL_RUNNER_PORT=11434
|
||||
LLAMA_CPP_PORT=8000
|
||||
METRICS_PORT=9090
|
||||
|
||||
# Model Configuration
|
||||
DEFAULT_MODEL=ai/smollm2:135M-Q4_K_M
|
||||
MODEL_CACHE_DIR=/app/data/models
|
||||
MODEL_CONFIG_DIR=/app/data/config
|
||||
|
||||
# Network Configuration
|
||||
MODEL_RUNNER_NETWORK=model-runner-network
|
||||
MODEL_RUNNER_HOST=0.0.0.0
|
||||
|
||||
# Performance Tuning
|
||||
MAX_CONCURRENT_REQUESTS=10
|
||||
REQUEST_TIMEOUT=300
|
||||
KEEP_ALIVE=300
|
||||
|
||||
# Logging
|
||||
LOG_LEVEL=info
|
||||
LOG_FORMAT=json
|
||||
|
||||
# Health Check
|
||||
HEALTH_CHECK_INTERVAL=30s
|
||||
HEALTH_CHECK_TIMEOUT=10s
|
||||
HEALTH_CHECK_RETRIES=3
|
||||
HEALTH_CHECK_START_PERIOD=40s
|
Reference in New Issue
Block a user