huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-04-24 16:32:12 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat(backend): create llama_context_params with default factory

Browse Source
This commit is contained in:
Morgan Funtowicz 2024-11-28 23:57:08 +01:00
parent b1ebc8f73b
commit dc6435e3a5
2 changed files with 29 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
backends/llamacpp/csrc/ffi.hpp
View File

@ -43,6 +43,15 @@ namespace huggingface::tgi::backends::llamacpp {
return std::shared_ptr<llama_model>(model, llama_model_deleter); return std::shared_ptr<llama_model>(model, llama_model_deleter);
}; };
auto get_llama_context_params = [](size_t num_threads) {
auto params = llama_context_default_params();
params.n_threads = num_threads;
params.n_threads_batch = num_threads;
params.flash_attn = true;
params.no_perf = false;
return params;
};
/** /**
* llama.cpp backend specific exception mapped from `backend_exception_t` to throw at the FFI level and * llama.cpp backend specific exception mapped from `backend_exception_t` to throw at the FFI level and
* allow automatic implementation of Result<_, Exception> from C++ to Rust * allow automatic implementation of Result<_, Exception> from C++ to Rust
@ -64,7 +73,7 @@ namespace huggingface::tgi::backends::llamacpp {
* @param num_threads The number of threads the worker is allowed to spawn accross for its threadpool * @param num_threads The number of threads the worker is allowed to spawn accross for its threadpool
*/ */
explicit llama_cpp_worker_frontend_t(llama_model *model, int32_t num_threads): explicit llama_cpp_worker_frontend_t(llama_model *model, int32_t num_threads):
model_{ make_shared_llama_model(model) }, worker_(model_, {.n_ubatch = 1, .n_threads = num_threads, .no_perf = true}) {} model_{ make_shared_llama_model(model) }, worker_(model_, get_llama_context_params(num_threads)) {}
/** /**
* Generate a new set of tokens from the provided `input_tokens`, streaming each individual token generated * Generate a new set of tokens from the provided `input_tokens`, streaming each individual token generated

25
backends/llamacpp/offline/main.cpp
View File

@ -27,24 +27,37 @@ int main(int argc, char **argv) {
llama_model_deleter llama_model_deleter
); );
auto prompt = "My name is Morgan"; auto prompt = std::string("My name is Morgan");
auto tokens = std::vector<llama_token>(16); auto tokens = std::vector<llama_token>(128);
const auto nb_tokens = llama_tokenize(model.get(), prompt, sizeof(prompt), tokens.data(), tokens.size(), true, const auto nb_tokens = llama_tokenize(model.get(), prompt.c_str(), prompt.size(), tokens.data(), tokens.size(),
true,
false); false);
tokens.resize(nb_tokens); tokens.resize(nb_tokens);
auto backend = worker_t(std::move(model), {.n_batch = 1, .n_threads = 4}); llama_numa_init(ggml_numa_strategy::GGML_NUMA_STRATEGY_DISTRIBUTE);
auto backend = worker_t(model, llama_context_default_params());
fmt::println("Tokenized: {}", tokens); fmt::println("Tokenized: {}", tokens);
// generate // generate
auto generated_tokens = std::vector<llama_token>(32); auto generated_tokens = std::vector<llama_token>(32);
const auto n_generated_tokens = backend.generate( const auto n_generated_tokens = backend.generate(
{{.max_new_tokens = 32}, {.top_k = 40}, tokens}, {{.max_new_tokens = 32}, {.top_k = 40, .top_p = 0.95, .temperature = 0.8},
tokens},
[&generated_tokens](llama_token new_token_id, float_t logit, bool is_eos, size_t step) -> bool { [&generated_tokens](llama_token new_token_id, float_t logit, bool is_eos, size_t step) -> bool {
generated_tokens.emplace(generated_tokens.begin() + (step - 1), new_token_id); generated_tokens.emplace(generated_tokens.begin() + (step - 1), new_token_id);
return false; return false;
} }
); );
generated_tokens.resize(n_generated_tokens.value()); generated_tokens.resize(n_generated_tokens.value());
fmt::println("Generated {} tokens", generated_tokens);
std::string decoded = std::string(256, 'a');
const size_t length = llama_detokenize(model.get(),
generated_tokens.data(),
generated_tokens.size(),
decoded.data(),
decoded.size(),
false, false);
decoded.resize(std::min(length, decoded.size()));
fmt::println("Generated tokens: {}", generated_tokens);
fmt::println("Generated text: {}", decoded);
} }