mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-09-10 10:43:48 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

Introduce prompt caching so prompts load instantly

Browse source 
This change also introduces an ephemeral status line in non-verbose mode
to display a load percentage status when slow operations are happening.
This commit is contained in:
Justine Tunney 2023-04-28 16:15:26 -07:00
parent bf6459e324
commit b31ba86ace
No known key found for this signature in database
GPG key ID: BE714B4575D6E328
7 changed files with 333 additions and 103 deletions
Download patch file Download diff file Expand all files Collapse all files

65
third_party/ggml/llama.cc vendored
View file

@ -1,5 +1,5 @@
/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:4;tab-width:8;coding:utf-8 -*-│
│vi: set net ft=c ts=4 sts=4 sw=4 fenc=utf-8 :vi│
/*-*-mode:c++;indent-tabs-mode:nil;c-basic-offset:4;tab-width:8;coding:utf-8-*-│
│vi: set net ft=c++ ts=4 sts=4 sw=4 fenc=utf-8 :vi│
╚──────────────────────────────────────────────────────────────────────────────╝
│ │
│ llama.cpp │
@ -25,6 +25,30 @@
│ SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. │
│ │
╚─────────────────────────────────────────────────────────────────────────────*/
#include "third_party/ggml/llama.h"
#include "libc/intrin/bits.h"
#include "third_party/ggml/ggml.h"
#include "third_party/ggml/llama_util.h"
#include "third_party/libcxx/algorithm"
#include "third_party/libcxx/array"
#include "third_party/libcxx/atomic"
#include "third_party/libcxx/cassert"
#include "third_party/libcxx/cinttypes"
#include "third_party/libcxx/climits"
#include "third_party/libcxx/cstdint"
#include "third_party/libcxx/cstdio"
#include "third_party/libcxx/cstring"
#include "third_party/libcxx/ctime"
#include "third_party/libcxx/fstream"
#include "third_party/libcxx/initializer_list"
#include "third_party/libcxx/map"
#include "third_party/libcxx/memory"
#include "third_party/libcxx/mutex"
#include "third_party/libcxx/queue"
#include "third_party/libcxx/random"
#include "third_party/libcxx/sstream"
#include "third_party/libcxx/thread"
#include "third_party/libcxx/unordered_map"
asm(".ident\t\"\\n\\n\
llama.cpp (MIT License)\\n\
@ -32,46 +56,9 @@ Copyright (c) 2023 Georgi Gerganov\"");
asm(".include \"libc/disclaimer.inc\"");
// clang-format off
// Defines fileno on msys:
#ifndef _GNU_SOURCE
#define _GNU_SOURCE
#include "third_party/libcxx/cstdint"
#include "third_party/libcxx/cstdio"
#endif
#include "third_party/ggml/llama_util.h"
#include "third_party/ggml/llama.h"
#include "third_party/ggml/ggml.h"
#include "third_party/libcxx/array"
#include "third_party/libcxx/ctime"
#include "third_party/libcxx/cinttypes"
#include "third_party/libcxx/fstream"
#include "third_party/libcxx/random"
#include "third_party/libcxx/map"
#include "third_party/libcxx/unordered_map"
#include "third_party/libcxx/queue"
#include "third_party/libcxx/cassert"
#include "third_party/libcxx/cstring"
#include "third_party/libcxx/climits"
#include "third_party/libcxx/memory"
#include "third_party/libcxx/algorithm"
#include "third_party/libcxx/initializer_list"
#include "third_party/libcxx/thread"
#include "third_party/libcxx/atomic"
#include "third_party/libcxx/mutex"
#include "third_party/libcxx/sstream"
#define LLAMA_USE_SCRATCH
#define LLAMA_MAX_SCRATCH_BUFFERS 16
#define READ32BE(s) \
((uint32_t)((const uint8_t *)(s))[0] << 030 | \
(uint32_t)((const uint8_t *)(s))[1] << 020 | \
(uint32_t)((const uint8_t *)(s))[2] << 010 | \
(uint32_t)((const uint8_t *)(s))[3] << 000)
// available llama models
enum e_model {
MODEL_UNKNOWN,