Skip to content

perplexity : fix MSVC build after #5020 #5043

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Jan 20, 2024
Merged

perplexity : fix MSVC build after #5020 #5043

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e15c616
perplexity : fix MSVC build after #5020
cebtenzzre Jan 19, 2024
32a392f
try a differerent fix
cebtenzzre Jan 19, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 6 additions & 6 deletions examples/perplexity/perplexity.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -458,23 +458,24 @@ static bool decode_helper(llama_context * ctx, llama_batch & batch, std::vector<
return true;
}

#define K_TOKEN_CHUNK 4

static void compute_logprobs(const float * batch_logits, int n_vocab, std::vector<std::thread>& workers,
const std::vector<std::pair<size_t, llama_token>>& eval_pairs, std::vector<float>& eval_results) {
constexpr int k_token_chunk = 4;
if (eval_results.size() != eval_pairs.size()) {
eval_results.resize(eval_pairs.size());
}
if (eval_pairs.empty()) return;

size_t max_threads = std::min((eval_pairs.size() + k_token_chunk - 1)/k_token_chunk, workers.size());
size_t max_threads = std::min((eval_pairs.size() + K_TOKEN_CHUNK - 1)/K_TOKEN_CHUNK, workers.size());

std::atomic<int> counter(0);
auto compute = [&counter, &eval_pairs, &eval_results, batch_logits, n_vocab] () {
float local_logprobs[k_token_chunk];
float local_logprobs[K_TOKEN_CHUNK];
while (true) {
size_t first = counter.fetch_add(k_token_chunk, std::memory_order_relaxed);
size_t first = counter.fetch_add(K_TOKEN_CHUNK, std::memory_order_relaxed);
if (first >= eval_results.size()) break;
size_t last = std::min(first + k_token_chunk, eval_results.size());
size_t last = std::min(first + K_TOKEN_CHUNK, eval_results.size());
for (size_t i = first; i < last; ++i) {
auto logits = batch_logits + eval_pairs[i].first * n_vocab;
float max_logit = logits[0];
Expand All @@ -497,7 +498,6 @@ static void compute_logprobs(const float * batch_logits, int n_vocab, std::vecto
for (size_t it = 0; it < max_threads; ++it) {
workers[it].join();
}

}

static void hellaswag_score(llama_context * ctx, const gpt_params & params) {
Expand Down