mirror of
https://github.com/ollama/ollama.git
synced 2025-05-16 22:44:25 +02:00
| .. | ||
| new_runner.sh | ||
| new_runner_benchmark_test.go | ||
| README.md | ||
| server_benchmark_test.go | ||
Benchmark
Performance benchmarking for Ollama.
Prerequisites
- Ollama server running locally (
127.0.0.1:11434) - Desired models pre-downloaded (e.g.,
llama3.2:1b)
Run Benchmark
# Run all tests
go test -bench=. -timeout 30m ./...
New Runner Benchmark
go test -bench=Runner
or to test multiple models:
# run this from within the benchmark directory
# requires: llama3.2:1b, llama3.1:8b, llama3.3:70b
sh new_runner.sh