CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-27 08:21:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

rpc : add backend registry / device interfaces (#9812)

Browse Source 
* rpc : add backend registry / device interfaces

* llama : add llama_supports_rpc API

* ggml_backend_rpc_start_rpc_server -> ggml_backend_rpc_start_server
This commit is contained in:
Diego Devesa
2024-10-10 20:14:55 +02:00
committed by GitHub
parent cf8e0a3bb9
commit 0e9f760eb1
8 changed files with 247 additions and 88 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
common/arg.cpp
View File

@@ -1353,15 +1353,15 @@ gpt_params_context gpt_params_parser_init(gpt_params & params, llama_example ex,
params.image.emplace_back(value);
}
).set_examples({LLAMA_EXAMPLE_LLAVA}));
#ifdef GGML_USE_RPC
add_opt(llama_arg(
{"--rpc"}, "SERVERS",
"comma separated list of RPC servers",
[](gpt_params & params, const std::string & value) {
params.rpc_servers = value;
}
).set_env("LLAMA_ARG_RPC"));
#endif
if (llama_supports_rpc()) {
add_opt(llama_arg(
{"--rpc"}, "SERVERS",
"comma separated list of RPC servers",
[](gpt_params & params, const std::string & value) {
params.rpc_servers = value;
}
).set_env("LLAMA_ARG_RPC"));
}
add_opt(llama_arg(
{"--mlock"},
"force system to keep model in RAM rather than swapping or compressing",