Skip to content

replaced all API facing int's with int32_t #4577

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 3 commits into from
Jan 2, 2024
Merged

replaced all API facing int's with int32_t #4577

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
efe287f
replaced all API facing `int`'s with `int32_t`
MarcusDunn Dec 21, 2023
0f84180
formatting and missed `int` in `llama_token_to_piece`
MarcusDunn Dec 21, 2023
1484483
Merge branch 'ggerganov:master' into sized-ints
MarcusDunn Jan 1, 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
50 changes: 25 additions & 25 deletions llama.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7917,7 +7917,7 @@ void llama_sample_softmax(struct llama_context * ctx, llama_token_data_array * c
}
}

void llama_sample_top_k(struct llama_context * ctx, llama_token_data_array * candidates, int k, size_t min_keep) {
void llama_sample_top_k(struct llama_context * ctx, llama_token_data_array * candidates, int32_t k, size_t min_keep) {
const int64_t t_start_sample_us = ggml_time_us();

k = std::max(k, (int) min_keep);
Expand Down Expand Up @@ -8277,7 +8277,7 @@ void llama_sample_classifier_free_guidance(
}
}

llama_token llama_sample_token_mirostat(struct llama_context * ctx, llama_token_data_array * candidates, float tau, float eta, int m, float * mu) {
llama_token llama_sample_token_mirostat(struct llama_context * ctx, llama_token_data_array * candidates, float tau, float eta, int32_t m, float * mu) {
GGML_ASSERT(ctx);

auto N = float(llama_n_vocab(llama_get_model(ctx)));
Expand Down Expand Up @@ -9485,7 +9485,7 @@ struct llama_model_quantize_params llama_model_quantize_default_params() {
return result;
}

int llama_max_devices(void) {
int32_t llama_max_devices(void) {
return LLAMA_MAX_DEVICES;
}

Expand Down Expand Up @@ -9796,23 +9796,23 @@ enum llama_vocab_type llama_vocab_type(const struct llama_model * model) {
return model->vocab.type;
}

int llama_n_vocab(const struct llama_model * model) {
int32_t llama_n_vocab(const struct llama_model * model) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

return model->vocab.id_to_token.size();
}

int llama_n_ctx_train(const struct llama_model * model) {
int32_t llama_n_ctx_train(const struct llama_model * model) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

return model->hparams.n_ctx_train;
}

int llama_n_embd(const struct llama_model * model) {
int32_t llama_n_embd(const struct llama_model * model) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

return model->hparams.n_embd;
}

float llama_rope_freq_scale_train(const struct llama_model * model) {
return model->hparams.rope_freq_scale_train;
}

int llama_model_meta_val_str(const struct llama_model * model, const char * key, char * buf, size_t buf_size) {
int32_t llama_model_meta_val_str(const struct llama_model * model, const char * key, char * buf, size_t buf_size) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ssize_t

const auto & it = model->gguf_kv.find(key);
if (it == model->gguf_kv.end()) {
if (buf_size > 0) {
Expand All @@ -9823,11 +9823,11 @@ int llama_model_meta_val_str(const struct llama_model * model, const char * key,
return snprintf(buf, buf_size, "%s", it->second.c_str());
}

int llama_model_meta_count(const struct llama_model * model) {
int32_t llama_model_meta_count(const struct llama_model * model) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

return (int)model->gguf_kv.size();
}

int llama_model_meta_key_by_index(const struct llama_model * model, int i, char * buf, size_t buf_size) {
int32_t llama_model_meta_key_by_index(const struct llama_model * model, int i, char * buf, size_t buf_size) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ssize_t

if (i < 0 || i >= (int)model->gguf_kv.size()) {
if (buf_size > 0) {
buf[0] = '\0';
Expand All @@ -9839,7 +9839,7 @@ int llama_model_meta_key_by_index(const struct llama_model * model, int i, char
return snprintf(buf, buf_size, "%s", it->first.c_str());
}

int llama_model_meta_val_str_by_index(const struct llama_model * model, int i, char * buf, size_t buf_size) {
int32_t llama_model_meta_val_str_by_index(const struct llama_model * model, int32_t i, char * buf, size_t buf_size) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ssize_t

if (i < 0 || i >= (int)model->gguf_kv.size()) {
if (buf_size > 0) {
buf[0] = '\0';
Expand All @@ -9851,7 +9851,7 @@ int llama_model_meta_val_str_by_index(const struct llama_model * model, int i, c
return snprintf(buf, buf_size, "%s", it->second.c_str());
}

int llama_model_desc(const struct llama_model * model, char * buf, size_t buf_size) {
int32_t llama_model_desc(const struct llama_model * model, char * buf, size_t buf_size) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ssize_t

return snprintf(buf, buf_size, "%s %s %s",
llama_model_arch_name(model->arch).c_str(),
llama_model_type_name(model->type),
Expand All @@ -9878,7 +9878,7 @@ struct ggml_tensor * llama_get_model_tensor(struct llama_model * model, const ch
return ggml_get_tensor(model->ctx, name);
}

int llama_model_quantize(
uint32_t llama_model_quantize(
const char * fname_inp,
const char * fname_out,
const llama_model_quantize_params * params) {
Expand All @@ -9891,7 +9891,7 @@ int llama_model_quantize(
}
}

int llama_apply_lora_from_file(struct llama_context * ctx, const char * path_lora, float scale, const char * path_base_model, int n_threads) {
int32_t llama_apply_lora_from_file(struct llama_context * ctx, const char * path_lora, float scale, const char * path_base_model, int32_t n_threads) {
try {
return llama_apply_lora_from_file_internal(ctx->model, path_lora, scale, path_base_model, n_threads);
} catch (const std::exception & err) {
Expand All @@ -9900,7 +9900,7 @@ int llama_apply_lora_from_file(struct llama_context * ctx, const char * path_lor
}
}

int llama_model_apply_lora_from_file(const struct llama_model * model, const char * path_lora, float scale, const char * path_base_model, int n_threads) {
int32_t llama_model_apply_lora_from_file(const struct llama_model * model, const char * path_lora, float scale, const char * path_base_model, int32_t n_threads) {
try {
return llama_apply_lora_from_file_internal(*model, path_lora, scale, path_base_model, n_threads);
} catch (const std::exception & err) {
Expand Down Expand Up @@ -9998,7 +9998,7 @@ void llama_kv_cache_view_update(const struct llama_context * ctx, struct llama_k
}
}

int llama_get_kv_cache_token_count(const struct llama_context * ctx) {
int32_t llama_get_kv_cache_token_count(const struct llama_context * ctx) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

int result = 0;

for (uint32_t i = 0; i < ctx->kv_self.size; i++) {
Expand All @@ -10008,7 +10008,7 @@ int llama_get_kv_cache_token_count(const struct llama_context * ctx) {
return result;
}

int llama_get_kv_cache_used_cells(const struct llama_context * ctx) {
int32_t llama_get_kv_cache_used_cells(const struct llama_context * ctx) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

return ctx->kv_self.used;
}

Expand Down Expand Up @@ -10488,7 +10488,7 @@ int llama_eval(
struct llama_context * ctx,
llama_token * tokens,
int32_t n_tokens,
int n_past) {
int32_t n_past) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t (and n_tokens should also be size_t across the whole codebase)

llama_kv_cache_seq_rm(ctx->kv_self, -1, n_past, -1);

const int ret = llama_decode_internal(*ctx, llama_batch_get_one(tokens, n_tokens, n_past, 0));
Expand All @@ -10503,7 +10503,7 @@ int llama_eval_embd(
struct llama_context * ctx,
float * embd,
int32_t n_tokens,
int n_past) {
int32_t n_past) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

size_t

llama_kv_cache_seq_rm(ctx->kv_self, -1, n_past, -1);

llama_batch batch = { n_tokens, nullptr, embd, nullptr, nullptr, nullptr, nullptr, n_past, 1, 0, };
Expand Down Expand Up @@ -10574,7 +10574,7 @@ void llama_batch_free(struct llama_batch batch) {
if (batch.logits) free(batch.logits);
}

int llama_decode(
int32_t llama_decode(
struct llama_context * ctx,
struct llama_batch batch) {
const int ret = llama_decode_internal(*ctx, batch);
Expand Down Expand Up @@ -10622,11 +10622,11 @@ llama_token llama_token_nl(const struct llama_model * model) {
return model->vocab.linefeed_id;
}

int llama_add_bos_token(const struct llama_model * model) {
int32_t llama_add_bos_token(const struct llama_model * model) {
return model->vocab.special_add_bos;
}

int llama_add_eos_token(const struct llama_model * model) {
int32_t llama_add_eos_token(const struct llama_model * model) {
return model->vocab.special_add_eos;
}

Expand All @@ -10646,12 +10646,12 @@ llama_token llama_token_eot(const struct llama_model * model) {
return model->vocab.special_eot_id;
}

int llama_tokenize(
int32_t llama_tokenize(
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ssize_t

const struct llama_model * model,
const char * text,
int text_len,
int32_t text_len,
llama_token * tokens,
int n_max_tokens,
int32_t n_max_tokens,
bool add_bos,
bool special) {
auto res = llama_tokenize_internal(model->vocab, std::string(text, text_len), add_bos, special);
Expand Down Expand Up @@ -10679,7 +10679,7 @@ static std::string llama_decode_text(const std::string & text) {
}

// does not write null-terminator to buf
int llama_token_to_piece(const struct llama_model * model, llama_token token, char * buf, int length) {
int32_t llama_token_to_piece(const struct llama_model * model, llama_token token, char * buf, int32_t length) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ssize_t

if (0 <= token && token < llama_n_vocab(model)) {
switch (llama_vocab_get_type(model->vocab)) {
case LLAMA_VOCAB_TYPE_SPM: {
Expand Down
Loading