Files
qwen-test/.env.example

16 lines
345 B
Plaintext

# Model Configuration
MODEL_ID=Qwen/Qwen3.5-9B
CACHE_DIR=./cache
TORCH_DTYPE=float16
# Hardware Configuration
CUDA_VISIBLE_DEVICES=0
# Benchmark Configuration
WARMUP_RUNS=3
BENCHMARK_RUNS=10
MAX_LENGTH=8192
# Optional: Proxy for model download (if needed)
# HTTP_PROXY=http://proxy.company.com:8080
# HTTPS_PROXY=http://proxy.company.com:8080