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

llama : move vocab, grammar and sampling into separate files (#8508)

Browse source 
* llama : move sampling code into llama-sampling

ggml-ci

* llama : move grammar code into llama-grammar

ggml-ci

* cont

ggml-ci

* cont : pre-fetch rules

* cont

ggml-ci

* llama : deprecate llama_sample_grammar

* llama : move tokenizers into llama-vocab

ggml-ci

* make : update llama.cpp deps [no ci]

* llama : redirect external API to internal APIs

ggml-ci

* llama : suffix the internal APIs with "_impl"

ggml-ci

* llama : clean-up
This commit is contained in:
Georgi Gerganov 2024-07-23 13:10:17 +03:00 • committed by GitHub
parent 751fcfc6c3
commit 938943cdbf
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
18 changed files with 3521 additions and 2968 deletions
Download patch file Download diff file Expand all files Collapse all files

20
tests/test-llama-grammar.cpp
View file

@ -2,10 +2,12 @@
#undef NDEBUG
#endif
#include "llama.cpp" // TODO: not great
#define LLAMA_API_INTERNAL
#include "llama.h"
#include "grammar-parser.h"
#include <cassert>
#include <stdexcept>
int main()
{
@ -112,10 +114,10 @@ int main()
}
}
llama_grammar *grammar = NULL;
llama_grammar * grammar = NULL;
std::vector<const llama_grammar_element *> grammar_rules(parsed_grammar.c_rules());
grammar = llama_grammar_init(
grammar_rules.data(), grammar_rules.size(), parsed_grammar.symbol_ids.at("root"));
grammar = llama_grammar_init(grammar_rules.data(), grammar_rules.size(), parsed_grammar.symbol_ids.at("root"));
if (grammar == nullptr)
{
throw std::runtime_error("Failed to initialize llama_grammar");
@ -172,7 +174,7 @@ int main()
}};
auto index = 0;
for (auto stack : grammar->stacks)
for (auto stack : llama_grammar_get_stacks(grammar))
{
// compare stack to expected_stack
for (uint32_t i = 0; i < stack.size(); i++)
@ -374,13 +376,13 @@ int main()
},
};
std::vector<llama_grammar_candidate> rejects = llama_grammar_reject_candidates_for_stack(grammar->rules, grammar->stacks[0], next_candidates);
std::vector<llama_grammar_candidate> rejects = llama_grammar_reject_candidates_for_stack(llama_grammar_get_rules(grammar), llama_grammar_get_stacks(grammar)[0], next_candidates);
std::vector<std::vector<llama_grammar_candidate>> all_rejects;
for (std::size_t count = 0; count < grammar->stacks.size(); ++count)
for (std::size_t count = 0; count < llama_grammar_get_stacks(grammar).size(); ++count)
{
rejects = llama_grammar_reject_candidates_for_stack(grammar->rules, grammar->stacks[count], next_candidates);
rejects = llama_grammar_reject_candidates_for_stack(llama_grammar_get_rules(grammar), llama_grammar_get_stacks(grammar)[count], next_candidates);
all_rejects.push_back(rejects);
}
@ -401,6 +403,6 @@ int main()
delete[] candidate.code_points;
candidate.code_points = nullptr;
}
delete grammar;
llama_grammar_free(grammar);
return 0;
}