char-fork/llama.cpp

git clone https://git.t4t.associates/char-fork/llama.cpp

Sigbjørn Skjæretdocker : include legacy llama-completion binary (#17964)b7f5f46e0

char/aliceai-t5-moe
brancheschar/aliceai-t5-moeall refs →
2.4 KiB53 linesraw
1#!/usr/bin/env bash
2set -e
3
4# Read the first argument into a variable
5arg1="$1"
6
7# Shift the arguments to remove the first one
8shift
9
10if [[ "$arg1" == '--convert' || "$arg1" == '-c' ]]; then
11    exec python3 ./convert_hf_to_gguf.py "$@"
12elif [[ "$arg1" == '--quantize' || "$arg1" == '-q' ]]; then
13    exec ./llama-quantize "$@"
14elif [[ "$arg1" == '--run' || "$arg1" == '-r' ]]; then
15    exec ./llama-cli "$@"
16elif [[ "$arg1" == '--run-legacy' || "$arg1" == '-l' ]]; then
17    exec ./llama-completion "$@"
18elif [[ "$arg1" == '--bench' || "$arg1" == '-b' ]]; then
19    exec ./llama-bench "$@"
20elif [[ "$arg1" == '--perplexity' || "$arg1" == '-p' ]]; then
21    exec ./llama-perplexity "$@"
22elif [[ "$arg1" == '--all-in-one' || "$arg1" == '-a' ]]; then
23    echo "Converting PTH to GGML..."
24    for i in $(ls $1/$2/ggml-model-f16.bin*); do
25        if [ -f "${i/f16/q4_0}" ]; then
26            echo "Skip model quantization, it already exists: ${i/f16/q4_0}"
27        else
28            echo "Converting PTH to GGML: $i into ${i/f16/q4_0}..."
29            exec ./llama-quantize "$i" "${i/f16/q4_0}" q4_0
30        fi
31    done
32elif [[ "$arg1" == '--server' || "$arg1" == '-s' ]]; then
33    exec ./llama-server "$@"
34else
35    echo "Unknown command: $arg1"
36    echo "Available commands: "
37    echo "  --run (-r): Run a model (chat) previously converted into ggml"
38    echo "              ex: -m /models/7B/ggml-model-q4_0.bin"
39    echo "  --run-legacy (-l): Run a model (legacy completion) previously converted into ggml"
40    echo "              ex: -m /models/7B/ggml-model-q4_0.bin -no-cnv -p \"Building a website can be done in 10 simple steps:\" -n 512"
41    echo "  --bench (-b): Benchmark the performance of the inference for various parameters."
42    echo "              ex: -m model.gguf"
43    echo "  --perplexity (-p): Measure the perplexity of a model over a given text."
44    echo "              ex: -m model.gguf -f file.txt"
45    echo "  --convert (-c): Convert a llama model into ggml"
46    echo "              ex: --outtype f16 \"/models/7B/\" "
47    echo "  --quantize (-q): Optimize with quantization process ggml"
48    echo "              ex: \"/models/7B/ggml-model-f16.bin\" \"/models/7B/ggml-model-q4_0.bin\" 2"
49    echo "  --all-in-one (-a): Execute --convert & --quantize"
50    echo "              ex: \"/models/\" 7B"
51    echo "  --server (-s): Run a model on the server"
52    echo "              ex: -m /models/7B/ggml-model-q4_0.bin -c 2048 -ngl 43 -mg 1 --port 8080"
53fi