dearwolf/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-05-20 10:35:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 21

Remove some debug logging

Browse source
This commit is contained in:
Ettore Di Giacinto 2025-05-16 18:28:18 +02:00
parent b81896a297
commit b35483742c
1 changed files with 11 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

28
backend/cpp/llama/grpc-server.cpp
View file

@ -4084,32 +4084,31 @@ public:
} }
grpc::Status Predict(ServerContext* context, const backend::PredictOptions* request, backend::Reply* reply) { grpc::Status Predict(ServerContext* context, const backend::PredictOptions* request, backend::Reply* reply) {
std::cout << "[DEBUG] Starting Predict request processing" << std::endl; json data = parse_options(true, request);
json data = parse_options(true, request);
std::cout << "[DEBUG] Parsed request options" << std::endl;
data["stream"] = false; data["stream"] = false;
//Raise error if embeddings is set to true //Raise error if embeddings is set to true
if (ctx_server.params_base.embedding) { if (ctx_server.params_base.embedding) {
std::cout << "[DEBUG] Error: Embedding mode not supported in streaming" << std::endl; return grpc::Status(grpc::StatusCode::INVALID_ARGUMENT, "Embedding is not supported in Predict mode");
return grpc::Status(grpc::StatusCode::INVALID_ARGUMENT, "Embedding is not supported in streaming mode");
} }
std::cout << "[PREDICT] Received result: " << data.dump(2) << std::endl;
auto completion_id = gen_chatcmplid(); auto completion_id = gen_chatcmplid();
std::cout << "[DEBUG] Generated completion ID: " << completion_id << std::endl;
std::unordered_set<int> task_ids; std::unordered_set<int> task_ids;
try { try {
std::vector<server_task> tasks; std::vector<server_task> tasks;
const auto & prompt = data.at("prompt"); const auto & prompt = data.at("prompt");
const auto type = SERVER_TASK_TYPE_COMPLETION; const auto type = SERVER_TASK_TYPE_COMPLETION;
std::cout << "[DEBUG] Processing prompt of type: " << type << std::endl; // TODO: this log can become very long, put it behind a flag or think about a more compact format
//SRV_DBG("Prompt: %s\n", prompt.is_string() ? prompt.get<std::string>().c_str() : prompt.dump(2).c_str());
std::vector<raw_buffer> files; std::vector<raw_buffer> files;
const auto &images_data = data.find("image_data"); const auto &images_data = data.find("image_data");
// std::cout << "[PREDICT] Images data: " << images_data->dump(2) << std::endl;
if (images_data != data.end() && images_data->is_array()) if (images_data != data.end() && images_data->is_array())
{ {
std::cout << "[DEBUG] Found " << images_data->size() << " images to process" << std::endl; std::cout << "[PREDICT] Processing " << images_data->size() << " images" << std::endl;
for (const auto &img : *images_data) for (const auto &img : *images_data)
{ {
std::cout << "[PREDICT] Processing image" << std::endl; std::cout << "[PREDICT] Processing image" << std::endl;
@ -4117,19 +4116,17 @@ public:
files.push_back(decoded_data); files.push_back(decoded_data);
} }
} }
// process files // process files
mtmd::bitmaps bitmaps; mtmd::bitmaps bitmaps;
const bool has_mtmd = ctx_server.mctx != nullptr; const bool has_mtmd = ctx_server.mctx != nullptr;
{ {
if (!has_mtmd && !files.empty()) { if (!has_mtmd && !files.empty()) {
std::cout << "[DEBUG] Error: Server does not support multimodal processing" << std::endl;
throw std::runtime_error("This server does not support multimodal"); throw std::runtime_error("This server does not support multimodal");
} }
for (auto & file : files) { for (auto & file : files) {
std::cout << "[DEBUG] Processing image file of size: " << file.size() << " bytes" << std::endl;
mtmd::bitmap bmp(mtmd_helper_bitmap_init_from_buf(file.data(), file.size())); mtmd::bitmap bmp(mtmd_helper_bitmap_init_from_buf(file.data(), file.size()));
if (!bmp.ptr) { if (!bmp.ptr) {
std::cout << "[DEBUG] Error: Failed to load image" << std::endl;
throw std::runtime_error("Failed to load image"); throw std::runtime_error("Failed to load image");
} }
// calculate bitmap hash (for KV caching) // calculate bitmap hash (for KV caching)
@ -4142,12 +4139,11 @@ public:
// process prompt // process prompt
std::vector<server_tokens> inputs; std::vector<server_tokens> inputs;
if (!prompt.is_string()) { if (!prompt.is_string()) {
std::cout << "[DEBUG] Error: Prompt must be a string" << std::endl; std::cout << "[PREDICT] Prompt must be a string" << std::endl;
throw std::runtime_error("prompt must be a string"); throw std::runtime_error("prompt must be a string");
} }
if (has_mtmd) { if (has_mtmd) {
std::cout << "[DEBUG] Processing multimodal input" << std::endl;
// multimodal // multimodal
std::string prompt_str = prompt.get<std::string>(); std::string prompt_str = prompt.get<std::string>();
mtmd_input_text inp_txt = { mtmd_input_text inp_txt = {
@ -4163,14 +4159,13 @@ public:
bitmaps_c_ptr.data(), bitmaps_c_ptr.data(),
bitmaps_c_ptr.size()); bitmaps_c_ptr.size());
if (tokenized != 0) { if (tokenized != 0) {
std::cout << "[DEBUG] Error: Failed to tokenize multimodal prompt" << std::endl; std::cout << "[PREDICT] Failed to tokenize prompt" << std::endl;
throw std::runtime_error("Failed to tokenize prompt"); throw std::runtime_error("Failed to tokenize prompt");
} }
server_tokens tmp(chunks, true); server_tokens tmp(chunks, true);
inputs.push_back(std::move(tmp)); inputs.push_back(std::move(tmp));
} else { } else {
std::cout << "[DEBUG] Processing standard text input" << std::endl;
// non-multimodal version // non-multimodal version
auto tokenized_prompts = tokenize_input_prompts(ctx_server.vocab, prompt, true, true); auto tokenized_prompts = tokenize_input_prompts(ctx_server.vocab, prompt, true, true);
for (auto & p : tokenized_prompts) { for (auto & p : tokenized_prompts) {
@ -4179,7 +4174,6 @@ public:
} }
} }
std::cout << "[DEBUG] Created " << inputs.size() << " input tasks" << std::endl;
tasks.reserve(inputs.size()); tasks.reserve(inputs.size());
for (size_t i = 0; i < inputs.size(); i++) { for (size_t i = 0; i < inputs.size(); i++) {
server_task task = server_task(type); server_task task = server_task(type);