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

now starting to refactor the code

Browse source
This commit is contained in:
mike dupont 2023-11-24 11:49:09 -05:00
parent 9722cfd0bb
commit bc3b93b942
6 changed files with 895 additions and 722 deletions
Download patch file Download diff file Expand all files Collapse all files

166
ggml-internal.hpp
View file

@ -90,3 +90,169 @@ struct ggml_allocr {
ggml_tallocr_t talloc;
ggml_gallocr_t galloc;
};
#define GGML_NUMA_MAX_NODES 8
#define GGML_NUMA_MAX_CPUS 512
struct ggml_numa_node {
uint32_t cpus[GGML_NUMA_MAX_CPUS]; // hardware threads on this node
uint32_t n_cpus;
};
struct ggml_numa_nodes {
struct ggml_numa_node nodes[GGML_NUMA_MAX_NODES];
uint32_t n_nodes;
uint32_t total_cpus; // hardware threads on system
};
struct ggml_state {
struct ggml_context_container contexts[GGML_MAX_CONTEXTS];
struct ggml_numa_nodes numa;
ggml_state():contexts(), numa()
{
}
};
struct gguf_str {
uint64_t n; // GGUFv2
char * data;
};
struct ggml_map_custom1_op_params {
ggml_custom1_op_t fun;
int n_tasks;
void * userdata;
};
struct ggml_map_custom2_op_params {
ggml_custom2_op_t fun;
int n_tasks;
void * userdata;
};
struct ggml_map_custom3_op_params {
ggml_custom3_op_t fun;
int n_tasks;
void * userdata;
};
struct hash_map {
struct ggml_hash_set set;
struct ggml_tensor ** vals;
};
#if defined(_WIN32)
typedef volatile LONG atomic_int;
typedef atomic_int atomic_bool;
#else
#include<atomic>
using namespace std;
#endif
struct ggml_compute_state_shared {
const struct ggml_cgraph * cgraph;
const struct ggml_cplan * cplan;
int64_t perf_node_start_cycles;
int64_t perf_node_start_time_us;
const int n_threads;
// synchronization primitives
atomic_int n_active; // num active threads
atomic_int node_n; // active graph node
bool (*abort_callback)(void * data); // abort ggml_graph_compute when true
void * abort_callback_data;
};
typedef pthread_t ggml_thread_t;
struct ggml_compute_state {
ggml_thread_t thrd;
int ith;
struct ggml_compute_state_shared * shared;
};
union gguf_value {
uint8_t uint8;
int8_t int8;
uint16_t uint16;
int16_t int16;
uint32_t uint32;
int32_t int32;
float float32;
uint64_t uint64;
int64_t int64;
double float64;
bool bool_;
struct gguf_str str;
struct gguf_array_T {
enum gguf_type type;
uint64_t n; // GGUFv2
void * data;
} arr;
};
struct ggml_lbfgs_iteration_data {
float alpha;
float ys;
float * s;
float * y;
};
struct gguf_kv {
struct gguf_str key;
enum gguf_type type;
union gguf_value value;
};
struct gguf_header {
char magic[4];
uint32_t version;
uint64_t n_tensors; // GGUFv2
uint64_t n_kv; // GGUFv2
};
struct gguf_tensor_info {
struct gguf_str name;
uint32_t n_dims;
uint64_t ne[GGML_MAX_DIMS];
enum ggml_type type;
uint64_t offset; // offset from start of `data`, must be a multiple of `ALIGNMENT`
// for writing API
const void * data;
size_t size;
};
struct gguf_context {
struct gguf_header header;
struct gguf_kv * kv;
struct gguf_tensor_info * infos;
size_t alignment;
size_t offset; // offset of `data` from beginning of file
size_t size; // size of `data` in bytes
//uint8_t * padding;
void * data;
};
struct gguf_buf {
void * data;
size_t size;
size_t offset;
};
#include "ggml-backend-impl.h"

143
ggml.cpp
View file

@ -1625,33 +1625,12 @@ static void ggml_setup_op_has_task_pass(void) {
// NUMA support
//
#define GGML_NUMA_MAX_NODES 8
#define GGML_NUMA_MAX_CPUS 512
struct ggml_numa_node {
uint32_t cpus[GGML_NUMA_MAX_CPUS]; // hardware threads on this node
uint32_t n_cpus;
};
struct ggml_numa_nodes {
struct ggml_numa_node nodes[GGML_NUMA_MAX_NODES];
uint32_t n_nodes;
uint32_t total_cpus; // hardware threads on system
};
//
// ggml state
//
struct ggml_state {
struct ggml_context_container contexts[GGML_MAX_CONTEXTS];
struct ggml_numa_nodes numa;
ggml_state():contexts(), numa()
{
}
};
// global state
static struct ggml_state g_state;
@ -1986,10 +1965,6 @@ static inline int ggml_up(int n, int m) {
////////////////////////////////////////////////////////////////////////////////
static size_t GGUF_TYPE_SIZE[GGUF_TYPE_COUNT]={};
struct gguf_str {
uint64_t n; // GGUFv2
char * data;
};
static const char * GGUF_TYPE_NAME[GGUF_TYPE_COUNT] = {};
@ -6084,11 +6059,6 @@ struct ggml_tensor * ggml_map_custom3_inplace_f32(
}
// ggml_map_custom1
struct ggml_map_custom1_op_params {
ggml_custom1_op_t fun;
int n_tasks;
void * userdata;
};
static struct ggml_tensor * ggml_map_custom1_impl(
struct ggml_context * ctx,
@ -6141,11 +6111,6 @@ struct ggml_tensor * ggml_map_custom1_inplace(
// ggml_map_custom2
struct ggml_map_custom2_op_params {
ggml_custom2_op_t fun;
int n_tasks;
void * userdata;
};
static struct ggml_tensor * ggml_map_custom2_impl(
struct ggml_context * ctx,
@ -6202,11 +6167,6 @@ struct ggml_tensor * ggml_map_custom2_inplace(
// ggml_map_custom3
struct ggml_map_custom3_op_params {
ggml_custom3_op_t fun;
int n_tasks;
void * userdata;
};
static struct ggml_tensor * ggml_map_custom3_impl(
struct ggml_context * ctx,
@ -14475,10 +14435,6 @@ static void ggml_hash_set_free(struct ggml_hash_set hash_set) {
free(hash_set.keys);
}
struct hash_map {
struct ggml_hash_set set;
struct ggml_tensor ** vals;
};
static struct hash_map * ggml_new_hash_map(size_t size) {
struct hash_map * result = (hash_map *)malloc(sizeof(struct hash_map));
@ -15734,7 +15690,7 @@ typedef int ggml_lock_t;
#define GGML_LOCK_INITIALIZER 0
typedef pthread_t ggml_thread_t;
#define ggml_thread_create pthread_create
#define ggml_thread_join pthread_join
@ -15824,28 +15780,7 @@ static void set_numa_thread_affinity(int thread_n, int n_threads) { UNUSED(threa
static void clear_numa_thread_affinity(void) {}
#endif
struct ggml_compute_state_shared {
const struct ggml_cgraph * cgraph;
const struct ggml_cplan * cplan;
int64_t perf_node_start_cycles;
int64_t perf_node_start_time_us;
const int n_threads;
// synchronization primitives
atomic_int n_active; // num active threads
atomic_int node_n; // active graph node
bool (*abort_callback)(void * data); // abort ggml_graph_compute when true
void * abort_callback_data;
};
struct ggml_compute_state {
ggml_thread_t thrd;
int ith;
struct ggml_compute_state_shared * shared;
};
static void ggml_graph_compute_perf_stats_node(struct ggml_tensor * node, const struct ggml_compute_state_shared * st) {
int64_t cycles_cur = ggml_perf_cycles() - st->perf_node_start_cycles;
@ -17456,12 +17391,6 @@ static enum ggml_opt_result ggml_opt_adam(
// https://github.com/chokkan/liblbfgs
//
struct ggml_lbfgs_iteration_data {
float alpha;
float ys;
float * s;
float * y;
};
static enum ggml_opt_result linesearch_backtracking(
const struct ggml_opt_params * params,
@ -18328,71 +18257,6 @@ static_assert(GGUF_TYPE_COUNT == 13, "GGUF_TYPE_COUNT != 13");
//};
static_assert(GGUF_TYPE_COUNT == 13, "GGUF_TYPE_COUNT != 13");
union gguf_value {
uint8_t uint8;
int8_t int8;
uint16_t uint16;
int16_t int16;
uint32_t uint32;
int32_t int32;
float float32;
uint64_t uint64;
int64_t int64;
double float64;
bool bool_;
struct gguf_str str;
struct {
enum gguf_type type;
uint64_t n; // GGUFv2
void * data;
} arr;
};
struct gguf_kv {
struct gguf_str key;
enum gguf_type type;
union gguf_value value;
};
struct gguf_header {
char magic[4];
uint32_t version;
uint64_t n_tensors; // GGUFv2
uint64_t n_kv; // GGUFv2
};
struct gguf_tensor_info {
struct gguf_str name;
uint32_t n_dims;
uint64_t ne[GGML_MAX_DIMS];
enum ggml_type type;
uint64_t offset; // offset from start of `data`, must be a multiple of `ALIGNMENT`
// for writing API
const void * data;
size_t size;
};
struct gguf_context {
struct gguf_header header;
struct gguf_kv * kv;
struct gguf_tensor_info * infos;
size_t alignment;
size_t offset; // offset of `data` from beginning of file
size_t size; // size of `data` in bytes
//uint8_t * padding;
void * data;
};
static bool gguf_fread_el(FILE * file, void * dst, size_t size, size_t * offset) {
const size_t n = fread(dst, 1, size, file);
@ -19185,11 +19049,6 @@ void gguf_set_tensor_data(struct gguf_context * ctx, const char * name, const vo
// fwrite(val, sizeof(char), size, file);
//}
struct gguf_buf {
void * data;
size_t size;
size_t offset;
};
static struct gguf_buf gguf_buf_init(size_t size) {
struct gguf_buf buf = {

392
llama-internal.hpp
View file

@ -1,5 +1,5 @@
#include <set>
#include <queue>
enum llm_arch {
LLM_ARCH_LLAMA,
LLM_ARCH_FALCON,
@ -90,7 +90,7 @@ enum llama_fver {
struct LLM_KV {
LLM_KV(llm_arch arch) : arch(arch) {}
llm_arch arch;
std::string operator()(llm_kv kv) const; // moved to llama.cpp file
@ -196,7 +196,7 @@ struct llama_buffer {
// useful in cases where CUDA can try to allocate PINNED memory
bool fallback = false;
void resize(size_t n) ;
void resize(size_t n) ;
~llama_buffer();
@ -293,9 +293,9 @@ struct llama_vocab {
struct llama_mmap {
void * addr;
size_t size;
llama_mmap(const llama_mmap &) = delete;
llama_mmap(struct llama_file * file, size_t prefetch = (size_t) -1 /* -1 = max value */, bool numa = false);
~llama_mmap();
@ -371,8 +371,8 @@ struct llama_mlock {
#undef MLOCK_SUGGESTION
static void raw_unlock(void * addr, size_t size);
#elif defined(_WIN32)
static constexpr bool SUPPORTED = true;
static size_t lock_granularity();
static constexpr bool SUPPORTED = true;
static size_t lock_granularity();
bool raw_lock(void * ptr, size_t len) const ;
static void raw_unlock(void * ptr, size_t len);
#else
@ -516,3 +516,381 @@ struct LLM_TN {
std::string operator()(llm_tensor tensor, const std::string & suffix, int bid) const ;
};
struct llama_file {
// use FILE * so we don't have to re-open the file to mmap
FILE * fp;
size_t size;
llama_file(const char * fname, const char * mode) ;
size_t tell() const;
void seek(size_t offset, int whence) const;
void read_raw(void * ptr, size_t len) const;
uint32_t read_u32() const;
void write_raw(const void * ptr, size_t len) const ;
void write_u32(std::uint32_t val) const;
~llama_file();
};
struct llama_state {
llama_state();
// We save the log callback globally
ggml_log_callback log_callback;
void * log_callback_user_data = nullptr;
};
struct llama_model_loader {
int n_kv = 0;
int n_tensors = 0;
int n_created = 0;
int64_t n_elements = 0;
size_t n_bytes = 0;
bool use_mmap = false;
llama_file file;
llama_ftype ftype;
llama_fver fver;
std::unique_ptr<llama_mmap> mapping;
struct gguf_context * ctx_gguf = NULL;
struct ggml_context * ctx_meta = NULL;
llama_model_loader(const std::string & fname, bool use_mmap) ;
~llama_model_loader();
std::string get_arch_name() const;
enum llm_arch get_arch() const ;
const char * get_tensor_name(int i) const;
struct ggml_tensor * get_tensor_meta(int i) const;
void calc_sizes(size_t & ctx_size_p, size_t & mmapped_size_p) const;
struct ggml_tensor * create_tensor_for(struct ggml_context * ctx, struct ggml_tensor * meta, ggml_backend_type backend) ;
struct ggml_tensor * create_tensor(struct ggml_context * ctx, const std::string & name, const std::vector<int64_t> & ne, ggml_backend_type backend) ;
void done_getting_tensors() const;
size_t file_offset(const char * name) const;
void load_data_for(struct ggml_tensor * cur) const ;
void load_all_data(struct ggml_context * ctx, llama_progress_callback progress_callback, void * progress_callback_user_data, llama_mlock * lmlock) ;
};
struct llama_data_context {
virtual void write(const void * src, size_t size) = 0;
virtual size_t get_size_written() = 0;
virtual ~llama_data_context() = default;
};
struct llama_data_buffer_context : llama_data_context {
uint8_t * ptr;
size_t size_written = 0;
llama_data_buffer_context(uint8_t * p) ;
void write(const void * src, size_t size) override ;
size_t get_size_written() override ;
};
struct llama_data_file_context : llama_data_context {
llama_file * file;
size_t size_written = 0;
llama_data_file_context(llama_file * f);
size_t get_size_written() override ;
void write(const void * src, size_t size);
};
struct llama_beam {
std::vector<llama_token> tokens;
float p; // Cumulative beam probability (renormalized relative to all beams)
bool eob; // Initialize end-of-beam to false. Callback sets this to true.
// Sort beams by probability. In case of ties, prefer beams at eob.
bool operator<(const llama_beam & rhs) const ;
void shift_tokens(const size_t n) ;
llama_beam_view view() const;
};
// A struct for calculating logit-related info.
struct llama_logit_info {
const float * const logits;
const int n_vocab;
const float max_l;
const float normalizer;
struct sum_exp {
float max_l;
float operator()(float sum, float l) const { return sum + std::exp(l - max_l); }
};
llama_logit_info(llama_context * ctx);
llama_token_data get_token_data(const llama_token token_id) const ;
std::vector<llama_token_data> top_k(size_t k) ;
float probability_from_logit(float logit) const ;
};
struct llama_beam_search_data {
llama_context * ctx;
size_t n_beams;
int n_past;
int n_predict;
std::vector<llama_beam> beams;
std::vector<llama_beam> next_beams;
size_t common_prefix_length;
std::vector<llama_beam_view> beam_views;
llama_beam_search_data(llama_context * ctx, size_t n_beams, int n_past, int n_predict);
void collapse_beams(const size_t beam_idx) ;
void fill_next_beams_by_top_probabilities(llama_beam & beam) ;
size_t find_common_prefix_length() ;
llama_beams_state get_beams_state(const bool last_call) ;
void loop(const llama_beam_search_callback_fn_t callback, void * const callback_data);
static void renormalize_beam_probabilities(std::vector<llama_beam> & beams) ;
size_t top_beam_index();
void update_beams_from_beam_views();
};
using llm_build_cb = std::function<void(struct ggml_tensor * cur, const char * name, int nl)>;
enum llm_rope_type {
LLM_ROPE,
LLM_ROPE_NEOX,
LLM_ROPE_GLM,
};
enum llm_ffn_op_type {
LLM_FFN_SILU,
LLM_FFN_GELU,
LLM_FFN_RELU,
LLM_FFN_RELU_SQR,
};
enum llm_ffn_gate_type {
LLM_FFN_SEQ,
LLM_FFN_PAR, // ffn_gate is parallel to ffn_up
};
enum llm_norm_type {
LLM_NORM,
LLM_NORM_RMS,
};
struct llm_build_context {
const llama_model & model;
const llama_hparams & hparams;
const llama_cparams & cparams;
const llama_batch & batch;
const llama_kv_cache & kv_self;
const int64_t n_embd;
const int64_t n_layer;
const int64_t n_ctx; // user-specified context size (can be different from n_ctx_train)
const int64_t n_head;
const int64_t n_head_kv;
const int64_t n_embd_head;
const int64_t n_embd_gqa;
const float freq_base;
const float freq_scale;
const float ext_factor;
const float attn_factor;
const float beta_fast;
const float beta_slow;
const float norm_eps;
const float norm_rms_eps;
const int32_t n_tokens;
const int32_t n_kv; // size of KV cache to consider (n_kv <= n_ctx)
const int32_t kv_head; // index of where we store new KV data in the cache
const int32_t n_orig_ctx;
const bool do_rope_shift;
const llm_build_cb & cb;
llama_buffer & buf_compute;
struct ggml_context * ctx0 = nullptr;
// TODO: consider making the entire interface noexcept
llm_build_context(
llama_context & lctx,
const llama_batch & batch,
const llm_build_cb & cb,
bool worst_case);
void init() ;
void free() ;
struct ggml_cgraph * build_llama() ;
struct ggml_cgraph * build_baichuan() ;
struct ggml_cgraph * build_falcon() ;
struct ggml_cgraph * build_starcoder() ;
struct ggml_cgraph * build_persimmon() ;
struct ggml_cgraph * build_refact() ;
struct ggml_cgraph * build_bloom() ;
struct ggml_cgraph * build_mpt() ;
struct ggml_cgraph * build_stablelm();
};
enum llm_offload_func_e {
OFFLOAD_FUNC_NOP,
OFFLOAD_FUNC,
OFFLOAD_FUNC_KQ,
OFFLOAD_FUNC_V,
OFFLOAD_FUNC_NR,
OFFLOAD_FUNC_EMB,
OFFLOAD_FUNC_OUT,
};
struct llm_offload_trie {
struct node {
~node() ;
node * children[256] = { nullptr };
llm_offload_func_e func = OFFLOAD_FUNC_NOP;
};
node * root = nullptr;
llm_offload_trie();
llm_offload_trie(const std::unordered_map<const char *, llm_offload_func_e> & map) ;
~llm_offload_trie();
void add(const char * name, llm_offload_func_e func);
llm_offload_func_e find(const char * name) const;
};
struct llm_symbol {
using index = int;
index prev;
index next;
const char * text;
size_t n;
};
struct llm_bigram_spm {
struct comparator {
bool operator()(llm_bigram_spm & l, llm_bigram_spm & r);
};
using queue_storage = std::vector<llm_bigram_spm>;
using queue = std::priority_queue<llm_bigram_spm, queue_storage, comparator>;
llm_symbol::index left;
llm_symbol::index right;
float score;
size_t size;
};
struct llm_tokenizer_spm {
llm_tokenizer_spm(const llama_vocab & vocab);
void tokenize(const std::string & text, std::vector<llama_vocab::id> & output);
private:
void resegment(llm_symbol & symbol, std::vector<llama_vocab::id> & output) ;
void try_add_bigram(int left, int right) ;
const llama_vocab & vocab;
std::vector<llm_symbol> symbols;
llm_bigram_spm::queue work_queue;
std::map<std::string, std::pair<int, int>> rev_merge;
};
// BPE tokenizer
// adapted from https://github.com/cmp-nct/ggllm.cpp [MIT License]
// tried to simplify unicode stuff, so most likely does not work 100% correctly!
// TODO: there are a lot of common parts between spm and bpe tokenizers, should be refactored and reused
struct llm_bigram_bpe {
struct comparator {
bool operator()(const llm_bigram_bpe & l, const llm_bigram_bpe & r) const ;
};
using queue_storage = std::vector<llm_bigram_bpe>;
using queue = std::priority_queue<llm_bigram_bpe, queue_storage, comparator>;
llm_symbol::index left;
llm_symbol::index right;
std::string text;
int rank;
size_t size;
};
struct llm_tokenizer_bpe {
llm_tokenizer_bpe(const llama_vocab & vocab);
void tokenize(const std::string & text, std::vector<llama_vocab::id> & output);
private:
void add_new_bigram(int left, int right) ;
std::vector<std::string> bpe_gpt2_preprocess(const std::string & text) ;
const llama_vocab & vocab;
std::vector<llm_symbol> symbols;
std::vector<llm_symbol> symbols_final;
llm_bigram_bpe::queue work_queue;
};
typedef enum FRAGMENT_BUFFER_VARIANT_TYPE{
FRAGMENT_BUFFER_VARIANT_TYPE_TOKEN,
FRAGMENT_BUFFER_VARIANT_TYPE_RAW_TEXT
} FRAGMENT_BUFFER_VARIANT_TYPE;
struct fragment_buffer_variant{
fragment_buffer_variant(llama_vocab::id _token);
fragment_buffer_variant(const std::string & _raw_text, int64_t _offset, int64_t _length);
const FRAGMENT_BUFFER_VARIANT_TYPE type;
const llama_vocab::id token;
const std::string _dummy;
const std::string & raw_text;
const uint64_t offset;
const uint64_t length;
};
struct llama_partial_utf8 {
uint32_t value; // bit value so far (unshifted)
int n_remain; // num bytes remaining; -1 indicates invalid sequence
};
struct llama_grammar {
const std::vector<std::vector<llama_grammar_element>> rules;
std::vector<std::vector<const llama_grammar_element *>> stacks;
// buffer for partially generated UTF-8 sequence from accepted tokens
llama_partial_utf8 partial_utf8;
};
struct llama_grammar_candidate {
size_t index;
const uint32_t * code_points;
llama_partial_utf8 partial_utf8;
};
struct quantize_state_internal {
const llama_model & model;
const llama_model_quantize_params * params;
int n_attention_wv = 0;
int n_feed_forward_w2 = 0;
int i_attention_wv = 0;
int i_feed_forward_w2 = 0;
int n_k_quantized = 0;
int n_fallback = 0;
quantize_state_internal(const llama_model & model, const llama_model_quantize_params * params)
: model(model)
, params(params)
{}
};

473
llama.cpp
View file

@ -639,12 +639,8 @@ llama_buffer::~llama_buffer() {
}
struct llama_file {
// use FILE * so we don't have to re-open the file to mmap
FILE * fp;
size_t size;
llama_file(const char * fname, const char * mode) {
llama_file::llama_file(const char * fname, const char * mode) {
fp = std::fopen(fname, mode);
if (fp == NULL) {
throw std::runtime_error(format("failed to open %s: %s", fname, strerror(errno)));
@ -654,7 +650,7 @@ struct llama_file {
seek(0, SEEK_SET);
}
size_t tell() const {
size_t llama_file::tell() const {
#ifdef _WIN32
__int64 ret = _ftelli64(fp);
#else
@ -664,7 +660,8 @@ struct llama_file {
return (size_t) ret;
}
void seek(size_t offset, int whence) const {
void llama_file::seek(size_t offset, int whence) const {
#ifdef _WIN32
int ret = _fseeki64(fp, (__int64) offset, whence);
#else
@ -673,7 +670,7 @@ struct llama_file {
GGML_ASSERT(ret == 0); // same
}
void read_raw(void * ptr, size_t len) const {
void llama_file::read_raw(void * ptr, size_t len) const {
if (len == 0) {
return;
}
@ -687,13 +684,13 @@ struct llama_file {
}
}
uint32_t read_u32() const {
uint32_t llama_file::read_u32() const {
uint32_t ret;
read_raw(&ret, sizeof(ret));
return ret;
}
void write_raw(const void * ptr, size_t len) const {
void llama_file::write_raw(const void * ptr, size_t len) const {
if (len == 0) {
return;
}
@ -704,16 +701,16 @@ struct llama_file {
}
}
void write_u32(std::uint32_t val) const {
void llama_file::write_u32(std::uint32_t val) const {
write_raw(&val, sizeof(val));
}
~llama_file() {
llama_file::~llama_file() {
if (fp) {
std::fclose(fp);
}
}
};
//
@ -985,12 +982,6 @@ static std::string llama_token_to_piece(const struct llama_context * ctx, llama_
// globals
//
struct llama_state {
// We save the log callback globally
ggml_log_callback log_callback = llama_log_callback_default;
void * log_callback_user_data = nullptr;
};
static llama_state g_state;
@ -1276,26 +1267,8 @@ static std::string llama_format_tensor_shape(const struct ggml_tensor * t) {
return buf;
}
struct llama_model_loader {
int n_kv = 0;
int n_tensors = 0;
int n_created = 0;
int64_t n_elements = 0;
size_t n_bytes = 0;
bool use_mmap = false;
llama_file file;
llama_ftype ftype;
llama_fver fver;
std::unique_ptr<llama_mmap> mapping;
struct gguf_context * ctx_gguf = NULL;
struct ggml_context * ctx_meta = NULL;
llama_model_loader(const std::string & fname, bool use_mmap) : file(fname.c_str(), "rb") {
llama_model_loader::llama_model_loader(const std::string & fname, bool use_mmap) : file(fname.c_str(), "rb") {
struct gguf_init_params params(
/*.no_alloc =*/ true,
/*.ctx = */ &ctx_meta
@ -1409,7 +1382,7 @@ struct llama_model_loader {
this->use_mmap = use_mmap;
}
~llama_model_loader() {
llama_model_loader::~llama_model_loader() {
if (ctx_gguf) {
gguf_free(ctx_gguf);
}
@ -1418,7 +1391,7 @@ struct llama_model_loader {
}
}
std::string get_arch_name() const {
std::string llama_model_loader::get_arch_name() const {
const auto kv = LLM_KV(LLM_ARCH_UNKNOWN);
std::string arch_name;
@ -1427,21 +1400,21 @@ struct llama_model_loader {
return arch_name;
}
enum llm_arch get_arch() const {
enum llm_arch llama_model_loader::get_arch() const {
const std::string arch_name = get_arch_name();
return llm_arch_from_string(arch_name);
}
const char * get_tensor_name(int i) const {
const char * llama_model_loader::get_tensor_name(int i) const {
return gguf_get_tensor_name(ctx_gguf, i);
}
struct ggml_tensor * get_tensor_meta(int i) const {
struct ggml_tensor * llama_model_loader::get_tensor_meta(int i) const {
return ggml_get_tensor(ctx_meta, get_tensor_name(i));
}
void calc_sizes(size_t & ctx_size_p, size_t & mmapped_size_p) const {
void llama_model_loader::calc_sizes(size_t & ctx_size_p, size_t & mmapped_size_p) const {
ctx_size_p = 0;
mmapped_size_p = 0;
@ -1452,7 +1425,7 @@ struct llama_model_loader {
}
}
struct ggml_tensor * create_tensor_for(struct ggml_context * ctx, struct ggml_tensor * meta, ggml_backend_type backend) {
struct ggml_tensor * llama_model_loader::create_tensor_for(struct ggml_context * ctx, struct ggml_tensor * meta, ggml_backend_type backend) {
if (backend != GGML_BACKEND_CPU) {
ggml_set_no_alloc(ctx, true);
}
@ -1470,7 +1443,7 @@ struct llama_model_loader {
return tensor;
}
struct ggml_tensor * create_tensor(struct ggml_context * ctx, const std::string & name, const std::vector<int64_t> & ne, ggml_backend_type backend) {
struct ggml_tensor * llama_model_loader::create_tensor(struct ggml_context * ctx, const std::string & name, const std::vector<int64_t> & ne, ggml_backend_type backend) {
struct ggml_tensor * cur = ggml_get_tensor(ctx_meta, name.c_str());
if (cur == NULL) {
@ -1503,13 +1476,13 @@ struct llama_model_loader {
return create_tensor_for(ctx, cur, backend);
}
void done_getting_tensors() const {
void llama_model_loader::done_getting_tensors() const {
if (n_created != n_tensors) {
throw std::runtime_error(format("%s: wrong number of tensors; expected %d, got %d", __func__, n_tensors, n_created));
}
}
size_t file_offset(const char * name) const {
size_t llama_model_loader::file_offset(const char * name) const {
const int idx = gguf_find_tensor(ctx_gguf, name);
if (idx < 0) {
@ -1519,7 +1492,7 @@ struct llama_model_loader {
return gguf_get_data_offset(ctx_gguf) + gguf_get_tensor_offset(ctx_gguf, idx);
}
void load_data_for(struct ggml_tensor * cur) const {
void llama_model_loader::load_data_for(struct ggml_tensor * cur) const {
const size_t offs = file_offset(ggml_get_name(cur));
if (use_mmap) {
@ -1530,7 +1503,7 @@ struct llama_model_loader {
}
}
void load_all_data(struct ggml_context * ctx, llama_progress_callback progress_callback, void * progress_callback_user_data, llama_mlock * lmlock) {
void llama_model_loader::load_all_data(struct ggml_context * ctx, llama_progress_callback progress_callback, void * progress_callback_user_data, llama_mlock * lmlock) {
size_t size_data = 0;
size_t size_lock = 0;
size_t size_pref = 0; // prefetch
@ -1606,7 +1579,7 @@ struct llama_model_loader {
done_size += ggml_nbytes(cur);
}
}
};
//};
//
// load LLaMA models
@ -2940,30 +2913,6 @@ static bool llama_model_load(const std::string & fname, llama_model & model, con
// llm_build
//
using llm_build_cb = std::function<void(struct ggml_tensor * cur, const char * name, int nl)>;
enum llm_rope_type {
LLM_ROPE,
LLM_ROPE_NEOX,
LLM_ROPE_GLM,
};
enum llm_ffn_op_type {
LLM_FFN_SILU,
LLM_FFN_GELU,
LLM_FFN_RELU,
LLM_FFN_RELU_SQR,
};
enum llm_ffn_gate_type {
LLM_FFN_SEQ,
LLM_FFN_PAR, // ffn_gate is parallel to ffn_up
};
enum llm_norm_type {
LLM_NORM,
LLM_NORM_RMS,
};
static struct ggml_tensor * llm_build_inp_embd(
struct ggml_context * ctx,
@ -3278,45 +3227,10 @@ static struct ggml_tensor * llm_build_kqv(
return cur;
}
struct llm_build_context {
const llama_model & model;
const llama_hparams & hparams;
const llama_cparams & cparams;
const llama_batch & batch;
const llama_kv_cache & kv_self;
const int64_t n_embd;
const int64_t n_layer;
const int64_t n_ctx; // user-specified context size (can be different from n_ctx_train)
const int64_t n_head;
const int64_t n_head_kv;
const int64_t n_embd_head;
const int64_t n_embd_gqa;
const float freq_base;
const float freq_scale;
const float ext_factor;
const float attn_factor;
const float beta_fast;
const float beta_slow;
const float norm_eps;
const float norm_rms_eps;
const int32_t n_tokens;
const int32_t n_kv; // size of KV cache to consider (n_kv <= n_ctx)
const int32_t kv_head; // index of where we store new KV data in the cache
const int32_t n_orig_ctx;
const bool do_rope_shift;
const llm_build_cb & cb;
llama_buffer & buf_compute;
struct ggml_context * ctx0 = nullptr;
// struct llm_build_context {
// TODO: consider making the entire interface noexcept
llm_build_context(
llm_build_context::llm_build_context(
llama_context & lctx,
const llama_batch & batch,
const llm_build_cb & cb,
@ -3353,7 +3267,7 @@ struct llm_build_context {
// all initializations should be done in init()
}
void init() {
void llm_build_context::init() {
struct ggml_init_params params(
//.mem_size =
buf_compute.size,
@ -3366,14 +3280,14 @@ struct llm_build_context {
ctx0 = ggml_init(params);
}
void free() {
void llm_build_context::free() {
if (ctx0) {
ggml_free(ctx0);
ctx0 = nullptr;
}
}
struct ggml_cgraph * build_llama() {
struct ggml_cgraph * llm_build_context::build_llama() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
GGML_ASSERT(n_embd_head == hparams.n_rot);
@ -3485,7 +3399,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_baichuan() {
struct ggml_cgraph * llm_build_context::build_baichuan() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
struct ggml_tensor * cur;
@ -3605,7 +3519,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_falcon() {
struct ggml_cgraph * llm_build_context::build_falcon() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
struct ggml_tensor * cur;
@ -3727,7 +3641,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_starcoder() {
struct ggml_cgraph * llm_build_context::build_starcoder() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
struct ggml_tensor * cur;
@ -3826,7 +3740,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_persimmon() {
struct ggml_cgraph * llm_build_context::build_persimmon() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
const int64_t n_rot = n_embd_head / 2;
@ -4036,7 +3950,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_refact() {
struct ggml_cgraph * llm_build_context::build_refact() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
struct ggml_tensor * cur;
@ -4127,7 +4041,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_bloom() {
struct ggml_cgraph * llm_build_context::build_bloom() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
struct ggml_tensor * cur;
@ -4221,7 +4135,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_mpt() {
struct ggml_cgraph * llm_build_context::build_mpt() {
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
struct ggml_tensor * cur;
@ -4320,7 +4234,7 @@ struct llm_build_context {
return gf;
}
struct ggml_cgraph * build_stablelm() {
struct ggml_cgraph * llm_build_context::build_stablelm() {
struct ggml_cgraph * gf = ggml_new_graph(ctx0);
struct ggml_tensor * cur;
@ -4432,27 +4346,18 @@ struct llm_build_context {
return gf;
}
};
//
// tensor offloading helpers
//
// TODO: will be removed with backend v2
enum llm_offload_func_e {
OFFLOAD_FUNC_NOP,
OFFLOAD_FUNC,
OFFLOAD_FUNC_KQ,
OFFLOAD_FUNC_V,
OFFLOAD_FUNC_NR,
OFFLOAD_FUNC_EMB,
OFFLOAD_FUNC_OUT,
};
// TODO: will be removed with backend v2
struct llm_offload_trie {
struct node {
~node() {
//struct llm_offload_trie {
// struct node {
llm_offload_trie::node::~node() {
for (int i = 0; i < 256; ++i) {
if (children[i]) {
delete children[i];
@ -4460,28 +4365,28 @@ struct llm_offload_trie {
}
}
node * children[256] = { nullptr };
llm_offload_func_e func = OFFLOAD_FUNC_NOP;
};
// node * children[256] = { nullptr };
// llm_offload_func_e func = OFFLOAD_FUNC_NOP;
// };
llm_offload_trie() {
llm_offload_trie::llm_offload_trie() {
root = new node;
}
llm_offload_trie(const std::unordered_map<const char *, llm_offload_func_e> & map) {
root = new node;
llm_offload_trie::llm_offload_trie(const std::unordered_map<const char *, llm_offload_func_e> & map) {
root = new node;
for (const auto & kv : map) {
add(kv.first, kv.second);
}
}
for (const auto & kv : map) {
add(kv.first, kv.second);
}
}
llm_offload_trie::~llm_offload_trie() {
delete root;
}
~llm_offload_trie() {
delete root;
}
void add(const char * name, llm_offload_func_e func) {
node * cur = root;
void llm_offload_trie::add(const char * name, llm_offload_func_e func) {
node * cur = root;
for (int i = 0; ; ++i) {
const uint8_t c = name[i];
@ -4500,7 +4405,7 @@ struct llm_offload_trie {
cur->func = func;
}
llm_offload_func_e find(const char * name) const {
llm_offload_func_e llm_offload_trie::find(const char * name) const {
const node * cur = root;
for (int i = 0; ; ++i) {
@ -4520,8 +4425,8 @@ struct llm_offload_trie {
return cur->func;
}
node * root = nullptr;
};
// node * root = nullptr;
//};
// TODO: will be removed with backend v2
static const std::unordered_map<const char *, llm_offload_func_e> k_offload_map = {
@ -5255,13 +5160,6 @@ static void llama_unescape_whitespace(std::string & word) {
replace_all(word, "\xe2\x96\x81", " ");
}
struct llm_symbol {
using index = int;
index prev;
index next;
const char * text;
size_t n;
};
static_assert(std::is_trivially_copyable<llm_symbol>::value, "llm_symbol is not trivially copyable");
@ -5269,24 +5167,16 @@ static_assert(std::is_trivially_copyable<llm_symbol>::value, "llm_symbol is not
// original implementation:
// https://github.com/ggerganov/llama.cpp/commit/074bea2eb1f1349a0118239c4152914aecaa1be4
struct llm_bigram_spm {
struct comparator {
bool operator()(llm_bigram_spm & l, llm_bigram_spm & r) {
return (l.score < r.score) || (l.score == r.score && l.left > r.left);
}
};
using queue_storage = std::vector<llm_bigram_spm>;
using queue = std::priority_queue<llm_bigram_spm, queue_storage, comparator>;
llm_symbol::index left;
llm_symbol::index right;
float score;
size_t size;
};
struct llm_tokenizer_spm {
llm_tokenizer_spm(const llama_vocab & vocab): vocab(vocab) {}
bool llm_bigram_spm::comparator::operator()(llm_bigram_spm & l, llm_bigram_spm & r) {
return (l.score < r.score) || (l.score == r.score && l.left > r.left);
}
void tokenize(const std::string & text, std::vector<llama_vocab::id> & output) {
// struct llm_tokenizer_spm {
llm_tokenizer_spm::llm_tokenizer_spm(const llama_vocab & vocab): vocab(vocab) {}
void llm_tokenizer_spm::tokenize(const std::string & text, std::vector<llama_vocab::id> & output) {
// split string into utf8 chars
int index = 0;
size_t offs = 0;
@ -5344,8 +5234,8 @@ struct llm_tokenizer_spm {
}
}
private:
void resegment(llm_symbol & symbol, std::vector<llama_vocab::id> & output) {
//private:
void llm_tokenizer_spm::resegment(llm_symbol & symbol, std::vector<llama_vocab::id> & output) {
auto text = std::string(symbol.text, symbol.n);
auto token = vocab.token_to_id.find(text);
@ -5370,7 +5260,7 @@ private:
resegment(symbols[p->second.second], output);
}
void try_add_bigram(int left, int right) {
void llm_tokenizer_spm::try_add_bigram(int left, int right) {
if (left == -1 || right == -1) {
return;
}
@ -5400,13 +5290,6 @@ private:
rev_merge[text] = std::make_pair(left, right);
}
const llama_vocab & vocab;
std::vector<llm_symbol> symbols;
llm_bigram_spm::queue work_queue;
std::map<std::string, std::pair<int, int>> rev_merge;
};
// BPE tokenizer
// adapted from https://github.com/cmp-nct/ggllm.cpp [MIT License]
@ -5414,26 +5297,15 @@ private:
// TODO: there are a lot of common parts between spm and bpe tokenizers, should be refactored and reused
struct llm_bigram_bpe {
struct comparator {
bool operator()(const llm_bigram_bpe & l, const llm_bigram_bpe & r) const {
return l.rank > r.rank || (l.rank == r.rank && l.left > r.left);
}
};
using queue_storage = std::vector<llm_bigram_bpe>;
using queue = std::priority_queue<llm_bigram_bpe, queue_storage, comparator>;
llm_symbol::index left;
llm_symbol::index right;
std::string text;
int rank;
size_t size;
};
bool llm_bigram_bpe::comparator::operator()(const llm_bigram_bpe & l, const llm_bigram_bpe & r) const {
return l.rank > r.rank || (l.rank == r.rank && l.left > r.left);
}
struct llm_tokenizer_bpe {
llm_tokenizer_bpe(const llama_vocab & vocab): vocab(vocab) {}
//struct llm_tokenizer_bpe {
llm_tokenizer_bpe::llm_tokenizer_bpe(const llama_vocab & vocab): vocab(vocab) {}
void tokenize(const std::string & text, std::vector<llama_vocab::id> & output) {
void llm_tokenizer_bpe::tokenize(const std::string & text, std::vector<llama_vocab::id> & output) {
int final_prev_index = -1;
auto word_collection = bpe_gpt2_preprocess(text);
@ -5534,8 +5406,8 @@ struct llm_tokenizer_bpe {
}
}
private:
void add_new_bigram(int left, int right) {
//private:
void llm_tokenizer_bpe::add_new_bigram(int left, int right) {
if (left == -1 || right == -1) {
return;
}
@ -5562,7 +5434,7 @@ private:
work_queue.push(bigram);
}
std::vector<std::string> bpe_gpt2_preprocess(const std::string & text) {
std::vector<std::string> llm_tokenizer_bpe::bpe_gpt2_preprocess(const std::string & text) {
std::vector<std::string> bpe_words;
std::vector<std::string> bpe_encoded_words;
@ -5701,28 +5573,17 @@ private:
return bpe_encoded_words;
}
const llama_vocab & vocab;
std::vector<llm_symbol> symbols;
std::vector<llm_symbol> symbols_final;
llm_bigram_bpe::queue work_queue;
};
typedef enum FRAGMENT_BUFFER_VARIANT_TYPE{
FRAGMENT_BUFFER_VARIANT_TYPE_TOKEN,
FRAGMENT_BUFFER_VARIANT_TYPE_RAW_TEXT
} FRAGMENT_BUFFER_VARIANT_TYPE;
struct fragment_buffer_variant{
fragment_buffer_variant(llama_vocab::id _token)
//struct fragment_buffer_variant{
fragment_buffer_variant::fragment_buffer_variant(llama_vocab::id _token)
:
type(FRAGMENT_BUFFER_VARIANT_TYPE_TOKEN),
token(_token),
raw_text(_dummy),
offset(0),
length(0){}
fragment_buffer_variant(const std::string & _raw_text, int64_t _offset, int64_t _length)
fragment_buffer_variant::fragment_buffer_variant(const std::string & _raw_text, int64_t _offset, int64_t _length)
:
type(FRAGMENT_BUFFER_VARIANT_TYPE_RAW_TEXT),
token((llama_vocab::id)-1),
@ -5734,13 +5595,6 @@ struct fragment_buffer_variant{
GGML_ASSERT( offset + length <= raw_text.length() );
}
const FRAGMENT_BUFFER_VARIANT_TYPE type;
const llama_vocab::id token;
const std::string _dummy;
const std::string & raw_text;
const uint64_t offset;
const uint64_t length;
};
// #define PRETOKENIZERDEBUG
@ -5946,24 +5800,6 @@ static std::vector<llama_vocab::id> llama_tokenize_internal(const llama_vocab &
// grammar - internal
//
struct llama_partial_utf8 {
uint32_t value; // bit value so far (unshifted)
int n_remain; // num bytes remaining; -1 indicates invalid sequence
};
struct llama_grammar {
const std::vector<std::vector<llama_grammar_element>> rules;
std::vector<std::vector<const llama_grammar_element *>> stacks;
// buffer for partially generated UTF-8 sequence from accepted tokens
llama_partial_utf8 partial_utf8;
};
struct llama_grammar_candidate {
size_t index;
const uint32_t * code_points;
llama_partial_utf8 partial_utf8;
};
// Decodes a UTF-8 string which may end in an incomplete sequence. Adds a terminating 0 for use as
// pointer. If an invalid sequence is encountered, returns `llama_partial_utf8.n_remain == -1`.
@ -6895,22 +6731,19 @@ void llama_grammar_accept_token(struct llama_context * ctx, struct llama_grammar
// Beam search
//
struct llama_beam {
std::vector<llama_token> tokens;
float p; // Cumulative beam probability (renormalized relative to all beams)
bool eob; // Initialize end-of-beam to false. Callback sets this to true.
// Sort beams by probability. In case of ties, prefer beams at eob.
bool operator<(const llama_beam & rhs) const {
// llama_beam {
bool llama_beam::operator<(const llama_beam & rhs) const {
return std::make_pair(p, eob) < std::make_pair(rhs.p, rhs.eob);
}
// Shift off first n tokens and discard them.
void shift_tokens(const size_t n) {
void llama_beam::shift_tokens(const size_t n) {
if (n) {
std::copy(tokens.begin() + n, tokens.end(), tokens.begin());
tokens.resize(tokens.size() - n);
}
}
llama_beam_view view() const {
llama_beam_view llama_beam::view() const {
llama_beam_view bv = {
.tokens =tokens.data(),
.n_tokens= tokens.size(),
@ -6919,25 +6752,25 @@ struct llama_beam {
};
return bv;
}
};
// A struct for calculating logit-related info.
struct llama_logit_info {
const float * const logits;
const int n_vocab;
const float max_l;
const float normalizer;
struct sum_exp {
float max_l;
float operator()(float sum, float l) const { return sum + std::exp(l - max_l); }
};
llama_logit_info(llama_context * ctx)
//struct llama_logit_info {
// const float * const logits;
// const int n_vocab;
// const float max_l;
// const float normalizer;
// struct sum_exp {
// float max_l;
// float operator()(float sum, float l) const { return sum + std::exp(l - max_l); }
// };
llama_logit_info::llama_logit_info(llama_context * ctx)
: logits(llama_get_logits(ctx))
, n_vocab(llama_n_vocab(llama_get_model(ctx)))
, max_l(*std::max_element(logits, logits + n_vocab))
, normalizer(1.0f / std::accumulate(logits, logits + n_vocab, 0.0f, sum_exp{max_l}))
{ }
llama_token_data get_token_data(const llama_token token_id) const {
llama_token_data llama_logit_info::get_token_data(const llama_token token_id) const {
constexpr auto p = std::numeric_limits<float>::quiet_NaN(); // never used
llama_token_data dd(
token_id,
@ -6947,7 +6780,7 @@ struct llama_logit_info {
return dd;
}
// Return top k token_data by logit.
std::vector<llama_token_data> top_k(size_t k) {
std::vector<llama_token_data> llama_logit_info::top_k(size_t k) {
std::vector<llama_token_data> min_heap; // min-heap by logit
const llama_token k_min = std::min(static_cast<llama_token>(k), n_vocab);
min_heap.reserve(k_min);
@ -6966,26 +6799,15 @@ struct llama_logit_info {
}
return min_heap;
}
float probability_from_logit(float logit) const {
float llama_logit_info::probability_from_logit(float logit) const {
return normalizer * std::exp(logit - max_l);
}
};
struct llama_beam_search_data {
llama_context * ctx;
size_t n_beams;
int n_past;
int n_predict;
std::vector<llama_beam> beams;
std::vector<llama_beam> next_beams;
// Re-calculated on each loop iteration
size_t common_prefix_length;
//struct llama_beam_search_data {
// Used to communicate to/from callback on beams state.
std::vector<llama_beam_view> beam_views;
llama_beam_search_data(llama_context * ctx, size_t n_beams, int n_past, int n_predict)
llama_beam_search_data::llama_beam_search_data(llama_context * ctx, size_t n_beams, int n_past, int n_predict)
: ctx(ctx)
, n_beams(n_beams)
, n_past(n_past)
@ -6996,7 +6818,7 @@ struct llama_beam_search_data {
}
// Collapse beams to a single beam given by index.
void collapse_beams(const size_t beam_idx) {
void llama_beam_search_data::collapse_beams(const size_t beam_idx) {
if (0u < beam_idx) {
std::swap(beams[0], beams[beam_idx]);
}
@ -7008,7 +6830,7 @@ struct llama_beam_search_data {
// * Gather elements until the vector is full, then call std::make_heap() on it.
// * If the heap is full and a new element is found that should be included, pop the
// least element to the back(), replace it with the new, then push it into the heap.
void fill_next_beams_by_top_probabilities(llama_beam & beam) {
void llama_beam_search_data::fill_next_beams_by_top_probabilities(llama_beam & beam) {
// Min-heaps use a greater-than comparator.
const auto comp = [](const llama_beam & a, const llama_beam & b) { return a.p > b.p; };
if (beam.eob) {
@ -7063,7 +6885,7 @@ struct llama_beam_search_data {
// Find common_prefix_length based on beams.
// Requires beams is not empty.
size_t find_common_prefix_length() {
size_t llama_beam_search_data::find_common_prefix_length() {
size_t common_prefix_length = beams[0].tokens.size();
for (size_t i = 1 ; i < beams.size() ; ++i) {
common_prefix_length = std::min(common_prefix_length, beams[i].tokens.size());
@ -7079,7 +6901,7 @@ struct llama_beam_search_data {
// Construct beams_state to send back to caller via the callback function.
// Side effect: set common_prefix_length = find_common_prefix_length();
llama_beams_state get_beams_state(const bool last_call) {
llama_beams_state llama_beam_search_data::get_beams_state(const bool last_call) {
for (size_t i = 0 ; i < beams.size() ; ++i) {
beam_views[i] = beams[i].view();
}
@ -7098,7 +6920,7 @@ struct llama_beam_search_data {
// * any of the beams have not yet reached end-of-beam (eob), AND
// * the highest probability beam(s) (plural in case of ties) are not at end-of-sentence
// (since all other beam probabilities can only decrease)
void loop(const llama_beam_search_callback_fn_t callback, void * const callback_data) {
void llama_beam_search_data::loop(const llama_beam_search_callback_fn_t callback, void * const callback_data) {
beams.push_back({{}, 1.0f, false}); // Start with one empty beam w/ probability = 1.0 and !eob.
const auto not_eob = [](const llama_beam & beam) { return !beam.eob; };
for (int i = 0 ; i < n_predict && std::any_of(beams.begin(),beams.end(),not_eob) &&
@ -7125,25 +6947,25 @@ struct llama_beam_search_data {
// As beams grow, the cumulative probabilities decrease.
// Renormalize them to avoid floating point underflow.
static void renormalize_beam_probabilities(std::vector<llama_beam> & beams) {
void llama_beam_search_data::renormalize_beam_probabilities(std::vector<llama_beam> & beams) {
const auto sum_p = [](float sum, llama_beam & beam) { return sum + beam.p; };
const float inv_sum = 1.0f / std::accumulate(beams.begin(), beams.end(), 0.0f, sum_p);
std::for_each(beams.begin(), beams.end(), [=](llama_beam & beam) { beam.p *= inv_sum; });
}
// Assumes beams is non-empty. Uses llama_beam::operator<() for ordering.
size_t top_beam_index() {
size_t llama_beam_search_data::top_beam_index() {
return std::max_element(beams.begin(), beams.end()) - beams.begin();
}
// Copy (p,eob) for each beam which may have been changed by the callback.
void update_beams_from_beam_views() {
void llama_beam_search_data::update_beams_from_beam_views() {
for (size_t i = 0 ; i < beams.size() ; ++i) {
beams[i].p = beam_views[i].p;
beams[i].eob = beam_views[i].eob;
}
}
};
void llama_beam_search(llama_context * ctx,
llama_beam_search_callback_fn_t callback, void * callback_data,
@ -7169,23 +6991,6 @@ struct no_init {
no_init() { /* do nothing */ }
};
struct quantize_state_internal {
const llama_model & model;
const llama_model_quantize_params * params;
int n_attention_wv = 0;
int n_feed_forward_w2 = 0;
int i_attention_wv = 0;
int i_feed_forward_w2 = 0;
int n_k_quantized = 0;
int n_fallback = 0;
quantize_state_internal(const llama_model & model, const llama_model_quantize_params * params)
: model(model)
, params(params)
{}
};
static void llama_convert_tensor_internal(
struct ggml_tensor * tensor, std::vector<no_init<float>> & output, std::vector<std::thread> & workers,
@ -8442,45 +8247,32 @@ size_t llama_get_state_size(const struct llama_context * ctx) {
return s_total;
}
// llama_context_data
struct llama_data_context {
virtual void write(const void * src, size_t size) = 0;
virtual size_t get_size_written() = 0;
virtual ~llama_data_context() = default;
};
struct llama_data_buffer_context : llama_data_context {
uint8_t * ptr;
size_t size_written = 0;
llama_data_buffer_context(uint8_t * p) : ptr(p) {}
llama_data_buffer_context::llama_data_buffer_context(uint8_t * p) : ptr(p) {}
void write(const void * src, size_t size) override {
memcpy(ptr, src, size);
ptr += size;
size_written += size;
}
void llama_data_buffer_context::write(const void * src, size_t size) {
memcpy(ptr, src, size);
ptr += size;
size_written += size;
}
size_t get_size_written() override {
return size_written;
}
};
size_t llama_data_buffer_context::get_size_written() {
return size_written;
}
struct llama_data_file_context : llama_data_context {
llama_file * file;
size_t size_written = 0;
llama_data_file_context(llama_file * f) : file(f) {}
llama_data_file_context::llama_data_file_context(llama_file * f) : file(f) {}
void write(const void * src, size_t size) override {
file->write_raw(src, size);
size_written += size;
}
void llama_data_file_context::write(const void * src, size_t size) {
file->write_raw(src, size);
size_written += size;
}
size_t get_size_written() override {
return size_written;
}
};
size_t llama_data_file_context::get_size_written() {
return size_written;
}
/** copy state data into either a buffer or file depending on the passed in context
*
@ -9287,3 +9079,6 @@ llama_context::~llama_context() {
ggml_allocr_free(alloc);
}
}
llama_state::llama_state(){
log_callback= llama_log_callback_default;
}

20
llama.h
View file

@ -114,7 +114,7 @@
LLAMA_ROPE_SCALING_MAX_VALUE = LLAMA_ROPE_SCALING_YARN,
};
typedef struct llama_token_data : refl::attr::usage::type{
typedef struct llama_token_data {
llama_token_data( llama_token id, float logit, float p):
id( id),logit(logit),p(p){ }
llama_token id; // token id
@ -122,7 +122,7 @@
float p; // probability of the token
} llama_token_data;
typedef struct llama_token_data_array : refl::attr::usage::type{
typedef struct llama_token_data_array {
llama_token_data_array(llama_token_data * data,
size_t size,
bool sorted):
@ -146,7 +146,7 @@
// - seq_id : the sequence to which the respective token belongs
// - logits : if zero, the logits for the respective token will not be output
//
typedef struct llama_batch : refl::attr::usage::type{
typedef struct llama_batch {
llama_batch(int32_t n_tokens,
llama_token * token,
@ -205,7 +205,7 @@
bool use_mlock; // force system to keep model in RAM
};
struct llama_context_params : refl::attr::usage::type{
struct llama_context_params{
uint32_t seed; // RNG seed, -1 for random
uint32_t n_ctx; // text context, 0 = from model
uint32_t n_batch; // prompt processing maximum batch size
@ -230,7 +230,7 @@
};
// model quantization parameters
typedef struct llama_model_quantize_params : refl::attr::usage::type{
typedef struct llama_model_quantize_params {
int nthread; // number of threads to use for quantizing, if <=0 will use std::thread::hardware_concurrency()
enum llama_ftype ftype; // quantize to this llama_ftype
bool allow_requantize; // allow quantizing non-f32/f16 tensors
@ -268,7 +268,7 @@
LLAMA_GRETYPE_CHAR_ALT = 6,
};
typedef struct llama_grammar_element : refl::attr::usage::type {
typedef struct llama_grammar_element {
llama_grammar_element( enum llama_gretype type,
uint32_t value // Unicode code point or rule ID
):type(type), value(value){}
@ -278,7 +278,7 @@
} llama_grammar_element;
// performance timing information
struct llama_timings : refl::attr::usage::type{
struct llama_timings {
double t_start_ms;
double t_end_ms;
double t_load_ms;
@ -755,7 +755,7 @@
// Beam search
//
struct llama_beam_view : refl::attr::usage::type{
struct llama_beam_view {
const llama_token * tokens;
size_t n_tokens;
@ -767,7 +767,7 @@
// Whenever 0 < common_prefix_length, this number of tokens should be copied from any of the beams
// (e.g. beams[0]) as they will be removed (shifted) from all beams in all subsequent callbacks.
// These pointers are valid only during the synchronous callback, so should not be saved.
struct llama_beams_state : refl::attr::usage::type{
struct llama_beams_state {
struct llama_beam_view * beam_views;
size_t n_beams; // Number of elements in beam_views[].
@ -831,3 +831,5 @@ const std::vector<std::pair<std::string, struct ggml_tensor *>> & llama_internal
#endif // LLAMA_H

423
print.hpp
View file

@ -1,7 +1,4 @@
//template<typename T> void print_fields(const T& obj);
#include <iostream>
//#include <refl.hpp>
#include "llama.h"
#include "ggml-internal.hpp"
#include "llama-internal.hpp"
@ -56,9 +53,9 @@ REFL_FIELD(prompt_file )
REFL_FIELD(path_prompt_cache )
REFL_FIELD(input_prefix )
REFL_FIELD(input_suffix )
//REFL_FIELD( antiprompt)
REFL_FIELD( antiprompt)
REFL_FIELD(logdir )
//REFL_FIELD( lora_adapter)
REFL_FIELD( lora_adapter)
REFL_FIELD(lora_base )
REFL_FIELD( ppl_stride )
REFL_FIELD( ppl_output_type )
@ -95,9 +92,6 @@ REFL_END
REFL_TYPE(llama_sampling_params)
REFL_END
REFL_TYPE(llama_buffer)
REFL_END
REFL_TYPE(llm_arch)
REFL_END
@ -106,8 +100,8 @@ REFL_FIELD( params)
REFL_FIELD( mirostat_mu)
REFL_FIELD( grammar)
REFL_FIELD( parsed_grammar)
//REFL_FIELD( prev) vector of ints
//REFL_FIELD( cur)
REFL_FIELD( prev)
REFL_FIELD( cur)
REFL_END
REFL_TYPE(llama_token_data )
@ -183,87 +177,82 @@ REFL_TYPE(ggml_context_container)
REFL_FIELD(context)
REFL_END
// REFL_TYPE(ggml_numa_node)
// REFL_FIELD(cpus)
// REFL_FIELD(n_cpus)
// REFL_END
REFL_TYPE(ggml_numa_node)
REFL_FIELD(cpus)
REFL_FIELD(n_cpus)
REFL_END
// REFL_TYPE(ggml_numa_nodes)
// REFL_FIELD(nodes)
// REFL_FIELD(n_nodes)
// REFL_END
REFL_TYPE(ggml_numa_nodes)
REFL_FIELD(nodes)
REFL_FIELD(n_nodes)
REFL_END
// REFL_TYPE(ggml_state)
// REFL_FIELD(contexts)
// REFL_FIELD(numa)
// REFL_END
REFL_TYPE(ggml_state)
REFL_FIELD(contexts)
REFL_FIELD(numa)
REFL_END
// REFL_TYPE(gguf_str)
// REFL_FIELD(n)
// REFL_FIELD(data)
// REFL_END
REFL_TYPE(gguf_str)
REFL_FIELD(n)
REFL_FIELD(data)
REFL_END
// REFL_TYPE(ggml_map_custom1_op_params)
// REFL_FIELD(fun)
// REFL_FIELD(n_tasks)
// REFL_END
REFL_TYPE(ggml_map_custom1_op_params)
REFL_FIELD(fun)
REFL_FIELD(n_tasks)
REFL_END
// REFL_TYPE(ggml_map_custom2_op_params)
// REFL_FIELD(fun)
// REFL_FIELD(n_tasks)
// REFL_END
// REFL_TYPE(ggml_map_custom3_op_params)
// REFL_FIELD(fun)
// REFL_FIELD(n_tasks)
// REFL_END
// REFL_TYPE(hash_map)
// REFL_FIELD(set)
// REFL_FIELD(vals)
// REFL_END
// REFL_TYPE(ggml_compute_state_shared)
// REFL_FIELD(cgraph)
// REFL_FIELD(cplan)
// REFL_END
// REFL_TYPE(ggml_compute_state)
// REFL_FIELD(thrd)
// REFL_FIELD(ith)
// REFL_END
// REFL_TYPE(ggml_lbfgs_iteration_data)
// REFL_FIELD(alpha)
// REFL_FIELD(ys)
// REFL_END
//REFL_TYPE()
// REFL_FIELD(type)
//REFL_END
// REFL_TYPE(gguf_kv)
// REFL_FIELD(key)
// REFL_FIELD(type)
// REFL_END
// REFL_TYPE(gguf_header)
// REFL_FIELD(magic)
// REFL_FIELD(version)
// REFL_END
// REFL_TYPE(gguf_tensor_info)
// REFL_FIELD(name)
// REFL_FIELD(n_dims)
// REFL_END
REFL_TYPE(gguf_context)
// REFL_FIELD(header)
// REFL_FIELD(kv)
REFL_TYPE(ggml_map_custom2_op_params)
REFL_FIELD(fun)
REFL_FIELD(n_tasks)
REFL_END
// REFL_TYPE(gguf_buf)
// REFL_FIELD(data)
// REFL_FIELD(size)
// REFL_END
REFL_TYPE(ggml_map_custom3_op_params)
REFL_FIELD(fun)
REFL_FIELD(n_tasks)
REFL_END
//REFL_TYPE(llama_token_data)
//REFL_END
REFL_TYPE(hash_map)
REFL_FIELD(set)
REFL_FIELD(vals)
REFL_END
REFL_TYPE(ggml_compute_state_shared)
REFL_FIELD(cgraph)
REFL_FIELD(cplan)
REFL_END
REFL_TYPE(ggml_compute_state)
REFL_FIELD(thrd)
REFL_FIELD(ith)
REFL_END
REFL_TYPE(ggml_lbfgs_iteration_data)
REFL_FIELD(alpha)
REFL_FIELD(ys)
REFL_END
REFL_TYPE(gguf_kv)
REFL_FIELD(key)
REFL_FIELD(type)
REFL_END
REFL_TYPE(gguf_header)
REFL_FIELD(magic)
REFL_FIELD(version)
REFL_END
REFL_TYPE(gguf_tensor_info)
REFL_FIELD(name)
REFL_FIELD(n_dims)
REFL_END
REFL_TYPE(gguf_context)
REFL_FIELD(header)
REFL_FIELD(kv)
REFL_END
REFL_TYPE(gguf_buf)
REFL_FIELD(data)
REFL_FIELD(size)
REFL_END
REFL_TYPE(llama_model_params)
@ -290,55 +279,55 @@ REFL_TYPE(llama_beams_state)
REFL_FIELD(beam_views)
REFL_END
//REFL_TYPE(ggml_backend)
//REFL_END
REFL_TYPE(ggml_backend)
REFL_END
REFL_TYPE(ggml_backend_buffer)
REFL_END
//REFL_TYPE(ggml_allocr)
//REFL_END
REFL_TYPE(ggml_allocr)
REFL_END
//REFL_TYPE(ggml_tallocr)
//REFL_END
REFL_TYPE(ggml_tallocr)
REFL_END
//REFL_TYPE(ggml_gallocr)
//REFL_END
REFL_TYPE(ggml_gallocr)
REFL_END
//REFL_TYPE(llama_buffer)
//REFL_FIELD(data)
//REFL_FIELD(size)
//REFL_END
REFL_TYPE(llama_buffer)
REFL_FIELD(data)
REFL_FIELD(size)
REFL_END
// REFL_TYPE(llama_file)
// REFL_FIELD(fp)
// REFL_FIELD(size)
// REFL_END
REFL_TYPE(llama_file)
REFL_FIELD(fp)
REFL_FIELD(size)
REFL_END
// REFL_TYPE(llama_mmap)
// REFL_FIELD(addr)
// REFL_FIELD(size)
// REFL_END
REFL_TYPE(llama_mmap)
REFL_FIELD(addr)
REFL_FIELD(size)
REFL_END
// REFL_TYPE(llama_mlock)
// REFL_FIELD(addr)
// REFL_FIELD(size)
// REFL_END
REFL_TYPE(llama_mlock)
REFL_FIELD(addr)
REFL_FIELD(size)
REFL_END
//REFL_TYPE(llama_state)
// REFL_FIELD(log_callback)
// REFL_FIELD(log_callback_user_data)
// REFL_END
REFL_TYPE(llama_state)
REFL_FIELD(log_callback)
REFL_FIELD(log_callback_user_data)
REFL_END
// REFL_TYPE(llama_hparams)
// REFL_FIELD(vocab_only)
// REFL_FIELD(n_vocab)
// REFL_END
REFL_TYPE(llama_hparams)
REFL_FIELD(vocab_only)
REFL_FIELD(n_vocab)
REFL_END
REFL_TYPE(llama_cparams)
@ -346,24 +335,21 @@ REFL_TYPE(llama_cparams)
REFL_FIELD(n_batch)
REFL_END
//REFL_TYPE(llama_layer)
// REFL_FIELD(attn_norm)
// REFL_FIELD(attn_norm_b)
//REFL_END
REFL_TYPE(llama_layer)
REFL_FIELD(attn_norm)
REFL_FIELD(attn_norm_b)
REFL_END
// REFL_TYPE(llama_kv_cell)
// REFL_FIELD(pos)
// REFL_FIELD(delta)
// REFL_END
REFL_TYPE(llama_kv_cell)
REFL_FIELD(pos)
REFL_FIELD(delta)
REFL_END
REFL_TYPE(llama_kv_cache)
REFL_FIELD(has_shift)
REFL_FIELD(head)
REFL_END
// REFL_TYPE(llama_vocab)
// REFL_END
REFL_TYPE(e_model)
REFL_END
@ -389,29 +375,22 @@ REFL_FIELD( output_norm)
REFL_FIELD( output_norm_b)
REFL_FIELD( output)
//REFL_FIELD( layers)
REFL_FIELD( layers)
REFL_FIELD( n_gpu_layers)
//REFL_FIELD( gguf_kv) unordered map
REFL_FIELD( gguf_kv) //unordered map
REFL_FIELD( ctx)
REFL_FIELD( buf)
//REFL_FIELD( mapping) std::unique_ptr
//REFL_FIELD( mlock_buf)
//REFL_FIELD( mlock_mmap)
//REFL_FIELD( tensors_by_name)
REFL_FIELD( mapping) //std::unique_ptr
REFL_FIELD( mlock_buf)
REFL_FIELD( mlock_mmap)
REFL_FIELD( tensors_by_name)
REFL_FIELD( t_load_us)
REFL_FIELD( t_start_us)
REFL_END
REFL_TYPE(llama_hparams)
REFL_END
//REFL_TYPE(std::vector<int> >)
//REFL_END
REFL_TYPE(llama_vocab)
REFL_END
@ -422,7 +401,7 @@ REFL_TYPE(llama_context)
REFL_FIELD( cparams)
//REFL_FIELD(model)
REFL_FIELD(kv_self)
//REFL_FIELD(rng) random numbers
REFL_FIELD(rng) //random numbers
REFL_FIELD(has_evaluated_once )
REFL_FIELD( t_start_us)
REFL_FIELD( t_load_us)
@ -432,13 +411,13 @@ REFL_FIELD( t_p_eval_us )
REFL_FIELD( n_sample )
REFL_FIELD( n_p_eval )
REFL_FIELD( n_eval )
//REFL_FIELD( logits)
REFL_FIELD( logits)
REFL_FIELD( logits_all )
//REFL_FIELD( embedding)
//REFL_FIELD( work_buffer)
REFL_FIELD( embedding)
REFL_FIELD( work_buffer)
REFL_FIELD( buf_compute)
REFL_FIELD( buf_alloc)
//REFL_FIELD( alloc )
REFL_FIELD( alloc )
#ifdef GGML_USE_METAL
REFL_FIELD( ctx_metal )
@ -450,108 +429,102 @@ REFL_FIELD( ctx_mpi )
#endif
REFL_END
// REFL_TYPE(llama_model_loader)
// REFL_FIELD(n_kv)
// REFL_FIELD(n_tensors)
// REFL_END
REFL_TYPE(llama_model_loader)
REFL_FIELD(n_kv)
REFL_FIELD(n_tensors)
REFL_END
// REFL_TYPE(llm_build_context)
// REFL_FIELD(model)
// REFL_FIELD(hparams)
// REFL_END
REFL_TYPE(llm_build_context)
// REFL_FIELD(model) cannot create pointer to reference member ‘llm_build_context::model’
// REFL_FIELD(hparams) cannot create pointer to reference member ‘llm_build_context::hparams’
REFL_END
// REFL_TYPE(llm_offload_trie)
// REFL_END
REFL_TYPE(llm_offload_trie)
REFL_END
// REFL_TYPE(llm_symbol)
// REFL_FIELD(prev)
// REFL_END
REFL_TYPE(llm_symbol)
REFL_FIELD(prev)
REFL_END
// REFL_TYPE(llm_bigram_spm)
// REFL_END
REFL_TYPE(llm_bigram_spm)
REFL_END
// REFL_TYPE(llm_tokenizer_spm)
// REFL_END
REFL_TYPE(llm_tokenizer_spm)
REFL_END
// REFL_TYPE(llm_bigram_bpe)
// REFL_END
REFL_TYPE(llm_bigram_bpe)
REFL_END
// REFL_TYPE(llm_tokenizer_bpe)
// REFL_END
// REFL_TYPE(fragment_buffer_variant)
// REFL_END
// REFL_TYPE(llama_partial_utf8)
// REFL_FIELD(value)
// REFL_FIELD(n_remain)
// REFL_END
REFL_TYPE(llama_grammar)
// REFL_FIELD(rules)
// REFL_FIELD(stacks)
REFL_TYPE(llm_tokenizer_bpe)
REFL_END
//REFL_TYPE(llama_grammar_candidate)
// REFL_FIELD(index)
// REFL_FIELD(code_points)
//REFL_END
REFL_TYPE(fragment_buffer_variant)
REFL_END
// REFL_TYPE(llama_beam)
// REFL_FIELD(tokens)
// REFL_FIELD(p)
// REFL_END
REFL_TYPE(llama_partial_utf8)
REFL_FIELD(value)
REFL_FIELD(n_remain)
REFL_END
// REFL_TYPE(llama_logit_info)
// REFL_FIELD(logits)
// REFL_FIELD(n_vocab)
// REFL_END
// REFL_TYPE(llama_beam_search_data)
// REFL_FIELD(ctx)
// REFL_FIELD(n_beams)
// REFL_END
// REFL_TYPE(quantize_state_internal)
// REFL_FIELD(model)
// REFL_FIELD(params)
// REFL_END
// REFL_TYPE(llama_data_context)
// REFL_END
REFL_TYPE(llama_grammar)
REFL_FIELD(rules)
REFL_FIELD(stacks)
REFL_END
// REFL_TYPE(llama_data_buffer_context)
// REFL_FIELD(ptr)
// REFL_END
// REFL_TYPE(llama_data_file_context)
// REFL_FIELD(file)
// REFL_END
REFL_TYPE(llama_grammar_candidate)
REFL_FIELD(index)
REFL_FIELD(code_points)
REFL_END
// // A simple struct with some fields and a function
// // A custom attribute to mark some fields as hidden
struct hidden : refl::attr::usage::field {};
// // Another struct with some fields and a function, using the custom attribute
// struct Person {
// std::string name;
// int age;
// [[hidden]] std::string password;
// void say_hello() const {
// std::cout << "Hello, I'm " << name << " and I'm " << age << " years old.\n";
// }
// };
REFL_TYPE(llama_beam)
REFL_FIELD(tokens)
REFL_FIELD(p)
REFL_END
REFL_TYPE(llama_logit_info)
REFL_FIELD(logits)
REFL_FIELD(n_vocab)
REFL_END
REFL_TYPE(llama_beam_search_data)
REFL_FIELD(ctx)
REFL_FIELD(n_beams)
REFL_END
REFL_TYPE(quantize_state_internal)
// REFL_FIELD(model)
REFL_FIELD(params)
REFL_FIELD( n_attention_wv )
REFL_FIELD( n_feed_forward_w2 )
REFL_FIELD( i_attention_wv )
REFL_FIELD( i_feed_forward_w2 )
REFL_FIELD( n_k_quantized )
REFL_FIELD( n_fallback )
REFL_END
REFL_TYPE(llama_data_context)
REFL_END
REFL_TYPE(llama_data_buffer_context)
REFL_FIELD(ptr)
REFL_END
REFL_TYPE(llama_data_file_context)
REFL_FIELD(file)
REFL_END
// // A generic function to print out the fields of any object
template<typename T>
void print_fields(const T& t) {
void print_fields(const T& ) {
//return;
// // Get the type descriptor of the object
constexpr auto type = refl::reflect<T>();