Files
ggml-org--llama.cpp/examples/model-conversion/scripts/utils/perplexity-run-simple.sh
T
wehub-resource-sync 09a3d3ab17
Copilot Setup Steps / copilot-setup-steps (push) Failing after 2s
Python check requirements.txt / check-requirements (push) Has been cancelled
Python Type-Check / python type-check (push) Has been cancelled
Update Operations Documentation / update-ops-docs (push) Has been cancelled
Check Pre-Tokenizer Hashes / pre-tokenizer-hashes (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 11:57:56 +08:00

33 lines
770 B
Bash
Executable File

#!/usr/bin/env bash
set -e
QUANTIZED_MODEL="${1:-"$QUANTIZED_MODEL"}"
BUILD_DIR="${2:-"$BUILD_DIR"}"
if [ -z "$QUANTIZED_MODEL" ]; then
echo "Error: Model path must be provided either as:" >&2
echo " 1. Command line argument" >&2
echo " 2. QUANTIZED_MODEL environment variable" >&2
exit 1
fi
# Check if data/wikitext-2-raw directory exists
if [ ! -d "ppl/wikitext-2-raw" ]; then
echo "ppl/wikitext-2-raw directory does not exist. Downloading..." >&2
mkdir -p ppl
pushd ppl
./../../../scripts/get-wikitext-2.sh
popd
fi
if [ -z "$BUILD_DIR" ]; then
BUILD_DIR="../../build"
fi
cmake --build $BUILD_DIR --target llama-perplexity -j8
${BUILD_DIR}/bin/llama-perplexity -m $QUANTIZED_MODEL -f ppl/wikitext-2-raw/wiki.test.raw