mpi : add support for distributed inference via MPI (#2099)

* MPI support, first cut

* fix warnings, update README

* fixes

* wrap includes

* PR comments

* Update CMakeLists.txt

* Add GH workflow, fix test

* Add info to README

* mpi : trying to move more MPI stuff into ggml-mpi (WIP) (#2099)

* mpi : add names for layer inputs + prep ggml_mpi_graph_compute()

* mpi : move all MPI logic into ggml-mpi

Not tested yet

* mpi : various fixes - communication now works but results are wrong

* mpi : fix output tensor after MPI compute (still not working)

* mpi : fix inference

* mpi : minor

* Add OpenMPI to GH action

* [mpi] continue-on-error: true

* mpi : fix after master merge

* [mpi] Link MPI C++ libraries to fix OpenMPI

* tests : fix new llama_backend API

* [mpi] use MPI_INT32_T

* mpi : factor out recv / send in functions and reuse

* mpi : extend API to allow usage with outer backends (e.g. Metal)

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

7 files changed, 19 insertions, 7 deletions

diff --git a/examples/embd-input/embd-input-lib.cpp b/examples/embd-input/embd-input-lib.cpp
index 5fa4942..2656382 100644
--- a/examples/embd-input/embd-input-lib.cpp
+++ b/examples/embd-input/embd-input-lib.cpp
@@ -34,7 +34,7 @@ struct MyModel* create_mymodel(int argc, char ** argv) {
     }
     fprintf(stderr, "%s: seed  = %d\n", __func__, params.seed);
 
-    llama_init_backend(params.numa);
+    llama_backend_init(params.numa);
 
     llama_model * model;
     llama_context * ctx;
diff --git a/examples/embedding/embedding.cpp b/examples/embedding/embedding.cpp
index 03e801c..5192d6d 100644
--- a/examples/embedding/embedding.cpp
+++ b/examples/embedding/embedding.cpp
@@ -35,7 +35,7 @@ int main(int argc, char ** argv) {
         params.prompt = gpt_random_prompt(rng);
     }
 
-    llama_init_backend(params.numa);
+    llama_backend_init(params.numa);
 
     llama_model * model;
     llama_context * ctx;
@@ -93,5 +93,7 @@ int main(int argc, char ** argv) {
     llama_free(ctx);
     llama_free_model(model);
 
+    llama_backend_free();
+
     return 0;
 }
diff --git a/examples/main/main.cpp b/examples/main/main.cpp
index 0f6391a..07d8fc6 100644
--- a/examples/main/main.cpp
+++ b/examples/main/main.cpp
@@ -105,7 +105,7 @@ int main(int argc, char ** argv) {
         params.prompt = gpt_random_prompt(rng);
     }
 
-    llama_init_backend(params.numa);
+    llama_backend_init(params.numa);
 
     llama_model * model;
     llama_context * ctx;
@@ -671,5 +671,7 @@ int main(int argc, char ** argv) {
     llama_free(ctx);
     llama_free_model(model);
 
+    llama_backend_free();
+
     return 0;
 }
diff --git a/examples/perplexity/perplexity.cpp b/examples/perplexity/perplexity.cpp
index fd4b03c..7e120ff 100644
--- a/examples/perplexity/perplexity.cpp
+++ b/examples/perplexity/perplexity.cpp
@@ -147,7 +147,7 @@ int main(int argc, char ** argv) {
         params.prompt = gpt_random_prompt(rng);
     }
 
-    llama_init_backend(params.numa);
+    llama_backend_init(params.numa);
 
     llama_model * model;
     llama_context * ctx;
@@ -172,5 +172,7 @@ int main(int argc, char ** argv) {
     llama_free(ctx);
     llama_free_model(model);
 
+    llama_backend_free();
+
     return 0;
 }
diff --git a/examples/quantize/quantize.cpp b/examples/quantize/quantize.cpp
index 1eb0f75..797d2f0 100644
--- a/examples/quantize/quantize.cpp
+++ b/examples/quantize/quantize.cpp
@@ -180,7 +180,7 @@ int main(int argc, char ** argv) {
         usage(argv[0]);
     }
 
-    llama_init_backend(false);
+    llama_backend_init(false);
 
     // parse command line arguments
     const std::string fname_inp = argv[arg_idx];
@@ -257,5 +257,7 @@ int main(int argc, char ** argv) {
         printf("%s:    total time = %8.2f ms\n", __func__, (t_main_end_us - t_main_start_us)/1000.0);
     }
 
+    llama_backend_free();
+
     return 0;
 }
diff --git a/examples/server/server.cpp b/examples/server/server.cpp
index 2cbfc00..296c5d6 100644
--- a/examples/server/server.cpp
+++ b/examples/server/server.cpp
@@ -1079,7 +1079,7 @@ int main(int argc, char **argv)
         params.model_alias = params.model;
     }
 
-    llama_init_backend(params.numa);
+    llama_backend_init(params.numa);
 
     LOG_INFO("build info", {{"build", BUILD_NUMBER},
                             {"commit", BUILD_COMMIT}});
@@ -1309,5 +1309,7 @@ int main(int argc, char **argv)
         return 1;
     }
 
+    llama_backend_free();
+
     return 0;
 }
diff --git a/examples/simple/simple.cpp b/examples/simple/simple.cpp
index 2d913ce..aa2c435 100644
--- a/examples/simple/simple.cpp
+++ b/examples/simple/simple.cpp
@@ -66,7 +66,7 @@ int main(int argc, char ** argv)
     // Init LLM :
     //---------------------------------
 
-    llama_init_backend(params.numa);
+    llama_backend_init(params.numa);
 
     llama_model * model;
     llama_context * ctx;
@@ -173,6 +173,8 @@ int main(int argc, char ** argv)
     llama_free( ctx );
     llama_free_model( model );
 
+    llama_backend_free();
+
     return 0;
 }