Download examples/sycl/win-run-llama2.bat from Brunobkr/llama.cpp_AlgMor24_github: direct link, hf CLI and curl.
- Browser
- Download file 471 Bytes
-
https://huggingface.co/datasets/Brunobkr/llama.cpp_AlgMor24_github/resolve/main/examples/sycl/win-run-llama2.bat
- Command line
-
hf download hf://datasets/Brunobkr/llama.cpp_AlgMor24_github/examples/sycl/win-run-llama2.bat
-
curl -L -o win-run-llama2.bat https://huggingface.co/datasets/Brunobkr/llama.cpp_AlgMor24_github/resolve/main/examples/sycl/win-run-llama2.bat
471 Bytes
| :: MIT license | |
| :: Copyright (C) 2024 Intel Corporation | |
| :: SPDX-License-Identifier: MIT | |
| set INPUT2="Building a website can be done in 10 simple steps:\nStep 1:" | |
| @call "C:\Program Files (x86)\Intel\oneAPI\setvars.bat" intel64 --force | |
| :: support malloc device memory more than 4GB. | |
| set UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1 | |
| set LOAD_MODE="--mmap" | |
| .\build\bin\llama-completion.exe -m models\llama-2-7b.Q4_0.gguf -no-cnv -p %INPUT2% -n 400 -e -ngl 99 -s 0 %LOAD_MODE% | |