vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge branch 'master' into speculative-tree

Browse source
This commit is contained in:
Georgi Gerganov 2023-10-17 19:31:40 +03:00 • committed by GitHub
commit bd9451ca2a
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
10 changed files with 385 additions and 90 deletions
Download patch file Download diff file Expand all files Collapse all files

2
examples/batched.swift/Sources/main.swift
View file

@ -217,7 +217,7 @@ llama_print_timings(context)
private func tokenize(text: String, add_bos: Bool) -> [llama_token] {
let n_tokens = text.count + (add_bos ? 1 : 0)
let tokens = UnsafeMutablePointer<llama_token>.allocate(capacity: n_tokens)
let tokenCount = llama_tokenize(model, text, Int32(text.count), tokens, Int32(n_tokens), add_bos)
let tokenCount = llama_tokenize(model, text, Int32(text.count), tokens, Int32(n_tokens), add_bos, /*special tokens*/ false)
var swiftTokens: [llama_token] = []
for i in 0 ..< tokenCount {
swiftTokens.append(tokens[Int(i)])

40
examples/main/main.cpp
View file

@ -237,7 +237,7 @@ int main(int argc, char ** argv) {
if (params.interactive_first || params.instruct || !params.prompt.empty() || session_tokens.empty()) {
LOG("tokenize the prompt\n");
embd_inp = ::llama_tokenize(ctx, params.prompt, add_bos);
embd_inp = ::llama_tokenize(ctx, params.prompt, add_bos, true);
} else {
LOG("use session tokens\n");
embd_inp = session_tokens;
@ -259,10 +259,10 @@ int main(int argc, char ** argv) {
if (ctx_guidance) {
LOG("cfg_negative_prompt: \"%s\"\n", log_tostr(sparams.cfg_negative_prompt));
guidance_inp = ::llama_tokenize(ctx_guidance, sparams.cfg_negative_prompt, add_bos);
guidance_inp = ::llama_tokenize(ctx_guidance, sparams.cfg_negative_prompt, add_bos, true);
LOG("guidance_inp tokenized: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx_guidance, guidance_inp).c_str());
std::vector<llama_token> original_inp = ::llama_tokenize(ctx, params.prompt, add_bos);
std::vector<llama_token> original_inp = ::llama_tokenize(ctx, params.prompt, add_bos, true);
LOG("original_inp tokenized: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx, original_inp).c_str());
original_prompt_len = original_inp.size();
@ -319,8 +319,8 @@ int main(int argc, char ** argv) {
}
// prefix & suffix for instruct mode
const auto inp_pfx = ::llama_tokenize(ctx, "\n\n### Instruction:\n\n", add_bos);
const auto inp_sfx = ::llama_tokenize(ctx, "\n\n### Response:\n\n", false);
const auto inp_pfx = ::llama_tokenize(ctx, "\n\n### Instruction:\n\n", add_bos, true);
const auto inp_sfx = ::llama_tokenize(ctx, "\n\n### Response:\n\n", false, true);
LOG("inp_pfx: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx, inp_pfx).c_str());
LOG("inp_sfx: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx, inp_sfx).c_str());
@ -382,6 +382,12 @@ int main(int argc, char ** argv) {
if (!params.antiprompt.empty()) {
for (const auto & antiprompt : params.antiprompt) {
LOG_TEE("Reverse prompt: '%s'\n", antiprompt.c_str());
if (params.verbose_prompt) {
auto tmp = ::llama_tokenize(ctx, antiprompt, false, true);
for (int i = 0; i < (int) tmp.size(); i++) {
LOG_TEE("%6d -> '%s'\n", tmp[i], llama_token_to_piece(ctx, tmp[i]).c_str());
}
}
}
}
@ -391,10 +397,22 @@ int main(int argc, char ** argv) {
if (!params.input_prefix.empty()) {
LOG_TEE("Input prefix: '%s'\n", params.input_prefix.c_str());
if (params.verbose_prompt) {
auto tmp = ::llama_tokenize(ctx, params.input_prefix, true, true);
for (int i = 0; i < (int) tmp.size(); i++) {
LOG_TEE("%6d -> '%s'\n", tmp[i], llama_token_to_piece(ctx, tmp[i]).c_str());
}
}
}
if (!params.input_suffix.empty()) {
LOG_TEE("Input suffix: '%s'\n", params.input_suffix.c_str());
if (params.verbose_prompt) {
auto tmp = ::llama_tokenize(ctx, params.input_suffix, false, true);
for (int i = 0; i < (int) tmp.size(); i++) {
LOG_TEE("%6d -> '%s'\n", tmp[i], llama_token_to_piece(ctx, tmp[i]).c_str());
}
}
}
}
LOG_TEE("sampling: repeat_last_n = %d, repeat_penalty = %f, presence_penalty = %f, frequency_penalty = %f, top_k = %d, tfs_z = %f, top_p = %f, typical_p = %f, temp = %f, mirostat = %d, mirostat_lr = %f, mirostat_ent = %f\n",
@ -687,7 +705,7 @@ int main(int argc, char ** argv) {
if (params.interactive) {
if (!params.antiprompt.empty()) {
// tokenize and inject first reverse prompt
const auto first_antiprompt = ::llama_tokenize(ctx, params.antiprompt.front(), false);
const auto first_antiprompt = ::llama_tokenize(ctx, params.antiprompt.front(), false, true);
embd_inp.insert(embd_inp.end(), first_antiprompt.begin(), first_antiprompt.end());
is_antiprompt = true;
}
@ -714,8 +732,7 @@ int main(int argc, char ** argv) {
std::string buffer;
if (!params.input_prefix.empty()) {
LOG("appending input prefix: '%s'\n", params.input_prefix.c_str());
buffer += params.input_prefix;
printf("%s", buffer.c_str());
printf("%s", params.input_prefix.c_str());
}
// color user input only
@ -737,7 +754,6 @@ int main(int argc, char ** argv) {
// append input suffix if any
if (!params.input_suffix.empty()) {
LOG("appending input suffix: '%s'\n", params.input_suffix.c_str());
buffer += params.input_suffix;
printf("%s", params.input_suffix.c_str());
}
@ -752,10 +768,14 @@ int main(int argc, char ** argv) {
embd_inp.insert(embd_inp.end(), inp_pfx.begin(), inp_pfx.end());
}
const auto line_inp = ::llama_tokenize(ctx, buffer, false);
const auto line_pfx = ::llama_tokenize(ctx, params.input_prefix, false, true);
const auto line_inp = ::llama_tokenize(ctx, buffer, false, false);
const auto line_sfx = ::llama_tokenize(ctx, params.input_suffix, false, true);
LOG("input tokens: %s\n", LOG_TOKENS_TOSTR_PRETTY(ctx, line_inp).c_str());
embd_inp.insert(embd_inp.end(), line_pfx.begin(), line_pfx.end());
embd_inp.insert(embd_inp.end(), line_inp.begin(), line_inp.end());
embd_inp.insert(embd_inp.end(), line_sfx.begin(), line_sfx.end());
// instruct mode: insert response suffix
if (params.instruct) {

97
examples/save-load-state/save-load-state.cpp
View file

@ -8,10 +8,7 @@
int main(int argc, char ** argv) {
gpt_params params;
llama_sampling_params & sparams = params.sampling_params;
params.seed = 42;
params.n_threads = 4;
sparams.repeat_last_n = 64;
params.prompt = "The quick brown fox";
if (!gpt_params_parse(argc, argv, params)) {
@ -25,56 +22,49 @@ int main(int argc, char ** argv) {
}
auto n_past = 0;
auto last_n_tokens_data = std::vector<llama_token>(sparams.repeat_last_n, 0);
std::string result0;
std::string result1;
// init
llama_model * model;
llama_context * ctx;
std::tie(model, ctx) = llama_init_from_gpt_params( params );
if (model == nullptr) {
return 1;
}
if (ctx == nullptr) {
llama_free_model(model);
std::tie(model, ctx) = llama_init_from_gpt_params(params);
if (model == nullptr || ctx == nullptr) {
fprintf(stderr, "%s : failed to init\n", __func__);
return 1;
}
// tokenize prompt
auto tokens = llama_tokenize(ctx, params.prompt, true);
auto n_prompt_tokens = tokens.size();
if (n_prompt_tokens < 1) {
fprintf(stderr, "%s : failed to tokenize prompt\n", __func__);
llama_free(ctx);
llama_free_model(model);
return 1;
}
// evaluate prompt
llama_decode(ctx, llama_batch_get_one(tokens.data(), n_prompt_tokens, n_past, 0));
llama_decode(ctx, llama_batch_get_one(tokens.data(), tokens.size(), n_past, 0));
n_past += tokens.size();
last_n_tokens_data.insert(last_n_tokens_data.end(), tokens.data(), tokens.data() + n_prompt_tokens);
n_past += n_prompt_tokens;
const size_t state_size = llama_get_state_size(ctx);
uint8_t * state_mem = new uint8_t[state_size];
// Save state (rng, logits, embedding and kv_cache) to file
// save state (rng, logits, embedding and kv_cache) to file
{
FILE *fp_write = fopen("dump_state.bin", "wb");
llama_copy_state_data(ctx, state_mem); // could also copy directly to memory mapped file
fwrite(state_mem, 1, state_size, fp_write);
fclose(fp_write);
std::vector<uint8_t> state_mem(llama_get_state_size(ctx));
{
FILE *fp_write = fopen("dump_state.bin", "wb");
llama_copy_state_data(ctx, state_mem.data()); // could also copy directly to memory mapped file
fwrite(state_mem.data(), 1, state_mem.size(), fp_write);
fclose(fp_write);
}
}
// save state (last tokens)
const auto last_n_tokens_data_saved = std::vector<llama_token>(last_n_tokens_data);
const auto n_past_saved = n_past;
// first run
printf("\n%s", params.prompt.c_str());
printf("\nfirst run: %s", params.prompt.c_str());
for (auto i = 0; i < params.n_predict; i++) {
auto * logits = llama_get_logits(ctx);
auto n_vocab = llama_n_vocab(model);
std::vector<llama_token_data> candidates;
candidates.reserve(n_vocab);
for (llama_token token_id = 0; token_id < n_vocab; token_id++) {
@ -83,9 +73,10 @@ int main(int argc, char ** argv) {
llama_token_data_array candidates_p = { candidates.data(), candidates.size(), false };
auto next_token = llama_sample_token(ctx, &candidates_p);
auto next_token_str = llama_token_to_piece(ctx, next_token);
last_n_tokens_data.push_back(next_token);
printf("%s", next_token_str.c_str());
result0 += next_token_str;
if (llama_decode(ctx, llama_batch_get_one(&next_token, 1, n_past, 0))) {
fprintf(stderr, "\n%s : failed to evaluate\n", __func__);
llama_free(ctx);
@ -103,32 +94,28 @@ int main(int argc, char ** argv) {
// make new context
auto * ctx2 = llama_new_context_with_model(model, llama_context_params_from_gpt_params(params));
// Load state (rng, logits, embedding and kv_cache) from file
{
FILE *fp_read = fopen("dump_state.bin", "rb");
if (state_size != llama_get_state_size(ctx2)) {
fprintf(stderr, "\n%s : failed to validate state size\n", __func__);
llama_free(ctx2);
llama_free_model(model);
return 1;
}
printf("\nsecond run: %s", params.prompt.c_str());
const size_t ret = fread(state_mem, 1, state_size, fp_read);
if (ret != state_size) {
// load state (rng, logits, embedding and kv_cache) from file
{
std::vector<uint8_t> state_mem(llama_get_state_size(ctx2));
FILE * fp_read = fopen("dump_state.bin", "rb");
const size_t ret = fread(state_mem.data(), 1, state_mem.size(), fp_read);
if (ret != state_mem.size()) {
fprintf(stderr, "\n%s : failed to read state\n", __func__);
llama_free(ctx2);
llama_free_model(model);
return 1;
}
llama_set_state_data(ctx2, state_mem); // could also read directly from memory mapped file
llama_set_state_data(ctx2, state_mem.data());
fclose(fp_read);
}
delete[] state_mem;
// restore state (last tokens)
last_n_tokens_data = last_n_tokens_data_saved;
n_past = n_past_saved;
// second run
@ -143,10 +130,11 @@ int main(int argc, char ** argv) {
llama_token_data_array candidates_p = { candidates.data(), candidates.size(), false };
auto next_token = llama_sample_token(ctx2, &candidates_p);
auto next_token_str = llama_token_to_piece(ctx2, next_token);
last_n_tokens_data.push_back(next_token);
printf("%s", next_token_str.c_str());
if (llama_decode(ctx, llama_batch_get_one(&next_token, 1, n_past, 0))) {
result1 += next_token_str;
if (llama_decode(ctx2, llama_batch_get_one(&next_token, 1, n_past, 0))) {
fprintf(stderr, "\n%s : failed to evaluate\n", __func__);
llama_free(ctx2);
llama_free_model(model);
@ -155,10 +143,17 @@ int main(int argc, char ** argv) {
n_past += 1;
}
printf("\n\n");
printf("\n");
llama_free(ctx2);
llama_free_model(model);
if (result0 != result1) {
fprintf(stderr, "\n%s : error : the 2 generations are different\n", __func__);
return 1;
}
fprintf(stderr, "\n%s : success\n", __func__);
return 0;
}