llama.cpp/examples/low_level_api/low_level_api_chat_cpp.py

422 lines
14 KiB
Python
Raw Normal View History

2023-04-03 20:54:46 +00:00
"""
This is an example implementation of main.cpp from llama.cpp
Quirks:
* Its not exactly alike since this port is designed around programmatic I/O
* Input is always echoed if on, so it should be turned off when using "input()"
* The first antiprompt should be the userprompt like "\nUser:",
because its added when n_predict is reached (aka generation ended prematurely)
2023-04-04 09:48:48 +00:00
* n_predict can be set to -1 for unlimited length responses (or just a really high value)
* Instruction mode adds its own antiprompt.
You should also still be feeding the model with a "primer" prompt that
shows it the expected format.
2023-04-03 20:54:46 +00:00
"""
import sys
from time import time
from os import cpu_count
2023-04-03 20:54:46 +00:00
import llama_cpp
from common import GptParams, gpt_params_parse, gpt_random_prompt
ANSI_COLOR_RESET = "\x1b[0m"
ANSI_COLOR_YELLOW = "\x1b[33m"
ANSI_BOLD = "\x1b[1m"
ANSI_COLOR_GREEN = "\x1b[32m"
CONSOLE_COLOR_DEFAULT = ANSI_COLOR_RESET
CONSOLE_COLOR_PROMPT = ANSI_COLOR_YELLOW
CONSOLE_COLOR_USER_INPUT = ANSI_BOLD + ANSI_COLOR_GREEN
2023-04-03 20:54:46 +00:00
# Iterative search
# Actively searches and prevents a pattern from being returned
class IterSearch:
def __init__(self, pattern):
self.pattern = list(pattern)
self.buffer = []
def __call__(self, char):
self.buffer += [char]
if (self.pattern[:len(self.buffer)] == self.buffer):
if (len(self.buffer) >= len(self.pattern)):
self.buffer.clear()
return []
_tmp = self.buffer[:]
self.buffer.clear()
return _tmp
2023-04-03 20:54:46 +00:00
# A LLaMA interactive session
class LLaMAInteract:
def __init__(self, params: GptParams) -> None:
2023-04-03 20:54:46 +00:00
# input args
self.params = params
if (self.params.perplexity):
raise NotImplementedError("""************
please use the 'perplexity' tool for perplexity calculations
************""")
if (self.params.embedding):
raise NotImplementedError("""************
please use the 'embedding' tool for embedding calculations
************""")
if (self.params.n_ctx > 2048):
print(f"""warning: model does not support \
context sizes greater than 2048 tokens ({self.params.n_ctx} \
specified) expect poor results""", file=sys.stderr)
if (self.params.seed <= 0):
self.params.seed = int(time())
print(f"seed = {self.params.seed}", file=sys.stderr)
if (self.params.random_prompt):
self.params.prompt = gpt_random_prompt(self.params.seed)
2023-04-03 20:54:46 +00:00
# runtime args
self.input_consumed = 0
self.n_past = 0
self.first_antiprompt = []
self.remaining_tokens = self.params.n_predict
self.output_echo = self.params.input_echo
2023-04-03 20:54:46 +00:00
# model load
self.lparams = llama_cpp.llama_context_default_params()
self.lparams.n_ctx = self.params.n_ctx
self.lparams.n_parts = self.params.n_parts
self.lparams.seed = self.params.seed
self.lparams.memory_f16 = self.params.memory_f16
self.lparams.use_mlock = self.params.use_mlock
self.lparams.use_mmap = self.params.use_mmap
self.ctx = llama_cpp.llama_init_from_file(self.params.model.encode("utf8"), self.lparams)
if (not self.ctx):
raise RuntimeError(f"error: failed to load model '{self.params.model}'")
print(file=sys.stderr)
print(f"system_info: n_threads = {self.params.n_threads} / {cpu_count()} \
| {llama_cpp.llama_print_system_info().decode('utf8')}", file=sys.stderr)
2023-04-03 20:54:46 +00:00
# determine the required inference memory per token:
if (self.params.mem_test):
tmp = [0, 1, 2, 3]
llama_cpp.llama_eval(self.ctx, (llama_cpp.c_int * len(tmp))(*tmp), len(tmp), 0, self.n_threads)
llama_cpp.llama_print_timings(self.ctx)
self.exit()
return
# create internal context
self.n_ctx = llama_cpp.llama_n_ctx(self.ctx)
# Add a space in front of the first character to match OG llama tokenizer behavior
self.params.prompt = " " + self.params.prompt
# Load prompt file
if (self.params.file):
with open(self.params.file) as f:
self.params.prompt = f.read()
# tokenize the prompt
self.embd = []
self.embd_inp = self._tokenize(self.params.prompt)
if (len(self.embd_inp) > self.params.n_ctx - 4):
raise RuntimeError(f"error: prompt is too long ({len(self.embd_inp)} tokens, max {self.params.n_ctx - 4})")
# number of tokens to keep when resetting context
if (self.params.n_keep < 0 or self.params.n_keep > len(self.embd_inp) or self.params.instruct):
self.params.n_keep = len(self.embd_inp)
self.inp_prefix = self._tokenize(self.params.instruct_inp_prefix)
self.inp_suffix = self._tokenize(self.params.instruct_inp_suffix, False)
# in instruct mode, we inject a prefix and a suffix to each input by the user
if (self.params.instruct):
self.params.interactive_start = True
_ptn = self._tokenize(self.params.instruct_inp_prefix.strip(), False)
self.first_antiprompt.append(_ptn)
self.antiecho = IterSearch(_ptn)
# enable interactive mode if reverse prompt or interactive start is specified
if (len(self.params.antiprompt) != 0 or self.params.interactive_start):
self.params.interactive = True
2023-04-03 20:54:46 +00:00
# determine newline token
2023-04-04 09:48:48 +00:00
self.llama_token_newline = self._tokenize("\n", False)
if (self.params.verbose_prompt):
print(f"""
prompt: '{self.params.prompt}'
number of tokens in prompt = {len(self.embd_inp)}""", file=sys.stderr)
2023-04-03 20:54:46 +00:00
for i in range(len(self.embd_inp)):
print(f"{self.embd_inp[i]} -> '{llama_cpp.llama_token_to_str(self.ctx, self.embd_inp[i])}'", file=sys.stderr)
2023-04-04 09:48:48 +00:00
if (self.params.n_keep > 0):
print("static prompt based on n_keep: '")
for i in range(self.params.n_keep):
print(llama_cpp.llama_token_to_str(self.ctx, self.embd_inp[i]), file=sys.stderr)
print("'", file=sys.stderr)
print(file=sys.stderr)
2023-04-03 20:54:46 +00:00
if (self.params.interactive):
print("interactive mode on.", file=sys.stderr)
if (len(self.params.antiprompt) > 0):
for antiprompt in self.params.antiprompt:
print(f"Reverse prompt: '{antiprompt}'", file=sys.stderr)
if len(self.params.input_prefix) > 0:
print(f"Input prefix: '{self.params.input_prefix}'", file=sys.stderr)
print(f"""sampling: temp = {self.params.temp},\
top_k = {self.params.top_k},\
top_p = {self.params.top_p},\
repeat_last_n = {self.params.repeat_last_n},\
repeat_penalty = {self.params.repeat_penalty}
generate: n_ctx = {self.n_ctx}, \
n_batch = {self.params.n_batch}, \
n_predict = {self.params.n_predict}, \
n_keep = {self.params.n_keep}
""", file=sys.stderr)
2023-04-03 20:54:46 +00:00
# determine antiprompt tokens
for i in self.params.antiprompt:
2023-04-04 09:48:48 +00:00
self.first_antiprompt.append(self._tokenize(i, False))
self.last_n_tokens = [0]*self.n_ctx #TODO: deque doesnt support slices
if (params.interactive):
print("""== Running in interactive mode. ==
- Press Ctrl+C to interject at any time.
- Press Return to return control to LLaMa.
- If you want to submit another line, end your input in '\\'.
""", file=sys.stderr)
self.set_color(CONSOLE_COLOR_PROMPT)
2023-04-04 09:48:48 +00:00
# tokenize a prompt
def _tokenize(self, prompt, bos=True):
_arr = (llama_cpp.llama_token * (len(prompt) + 1))()
_n = llama_cpp.llama_tokenize(self.ctx, prompt.encode("utf8"), _arr, len(_arr), bos)
return _arr[:_n]
2023-04-03 20:54:46 +00:00
def use_antiprompt(self):
return len(self.first_antiprompt) > 0
def set_color(self, c):
if (self.params.use_color):
print(c, end="")
2023-04-04 09:48:48 +00:00
# generate tokens
2023-04-03 20:54:46 +00:00
def generate(self):
while self.remaining_tokens > 0 or self.params.interactive:
2023-04-03 20:54:46 +00:00
# predict
if len(self.embd) > 0:
# infinite text generation via context swapping
# if we run out of context:
# - take the n_keep first tokens from the original prompt (via n_past)
# - take half of the last (n_ctx - n_keep) tokens and recompute the logits in a batch
if (self.n_past + len(self.embd) > self.n_ctx):
n_left = self.n_past - self.params.n_keep
self.n_past = self.params.n_keep
2023-04-03 20:54:46 +00:00
# insert n_left/2 tokens at the start of embd from last_n_tokens
_insert = self.last_n_tokens[
self.n_ctx - int(n_left/2) - len(self.embd):-len(self.embd)
2023-04-03 20:54:46 +00:00
]
self.embd = _insert + self.embd
2023-04-03 20:54:46 +00:00
if (llama_cpp.llama_eval(
self.ctx, (llama_cpp.llama_token * len(self.embd))(*self.embd), len(self.embd), self.n_past, self.params.n_threads
2023-04-03 20:54:46 +00:00
) != 0):
raise Exception("Failed to llama_eval!")
self.n_past += len(self.embd)
self.embd = []
if len(self.embd_inp) <= self.input_consumed:
# out of user input, sample next token
if (self.params.ignore_eos):
logits = llama_cpp.llama_get_logits(self.ctx)
logits[llama_cpp.llama_token_eos()] = llama_cpp.c_float(0)
_arr = self.last_n_tokens[-min(self.params.repeat_last_n, self.n_past):]
2023-04-03 20:54:46 +00:00
id = llama_cpp.llama_sample_top_p_top_k(
self.ctx,
(llama_cpp.llama_token * len(_arr))(*_arr),
len(_arr),
self.params.top_k,
self.params.top_p,
self.params.temp,
self.params.repeat_penalty,
2023-04-03 20:54:46 +00:00
)
self.last_n_tokens.pop(0)
2023-04-04 09:48:48 +00:00
self.last_n_tokens.append(id)
2023-04-03 20:54:46 +00:00
# replace end of text token with newline token when in interactive mode
if (id == llama_cpp.llama_token_eos() and self.params.interactive and not self.params.instruct):
2023-04-03 20:54:46 +00:00
id = self.llama_token_newline[0]
if (self.use_antiprompt()):
# tokenize and inject first reverse prompt
self.embd_inp += self.first_antiprompt[0]
2023-04-03 20:54:46 +00:00
# add it to the context
2023-04-04 09:48:48 +00:00
self.embd.append(id)
2023-04-03 20:54:46 +00:00
# echo this to console
self.output_echo = True
# decrement remaining sampling budget
self.remaining_tokens -= 1
else:
# output to console if input echo is on
self.output_echo = self.params.input_echo
2023-04-03 20:54:46 +00:00
# some user input remains from prompt or interaction, forward it to processing
while len(self.embd_inp) > self.input_consumed:
2023-04-04 09:48:48 +00:00
self.embd.append(self.embd_inp[self.input_consumed])
2023-04-03 20:54:46 +00:00
self.last_n_tokens.pop(0)
2023-04-04 09:48:48 +00:00
self.last_n_tokens.append(self.embd_inp[self.input_consumed])
2023-04-03 20:54:46 +00:00
self.input_consumed += 1
if len(self.embd) >= self.params.n_batch:
2023-04-03 20:54:46 +00:00
break
# display tokens
if self.output_echo:
for id in self.embd:
if self.params.instruct:
for r in self.antiecho(id):
yield r
else:
yield id
2023-04-03 20:54:46 +00:00
# reset color to default if we there is no pending user input
if (self.params.input_echo and len(self.embd_inp) == self.input_consumed):
self.set_color(CONSOLE_COLOR_DEFAULT)
if (self.params.interactive and len(self.embd_inp) <= self.input_consumed):
2023-04-04 09:48:48 +00:00
# if antiprompt is present, stop
if (self.use_antiprompt()):
if True in [
i == self.last_n_tokens[-len(i):]
for i in self.first_antiprompt
]:
break
2023-04-04 09:48:48 +00:00
# if we are using instruction mode, and we have processed the initial prompt
if (self.params.interactive_start):
2023-04-04 09:48:48 +00:00
break
2023-04-03 20:54:46 +00:00
# end of text token
2023-04-03 20:54:46 +00:00
if len(self.embd) > 0 and self.embd[-1] == llama_cpp.llama_token_eos():
if (not self.params.instruct):
for i in " [end of text]\n":
yield i
2023-04-03 20:54:46 +00:00
break
# respect n_predict even if antiprompt is present
if (self.params.interactive and self.remaining_tokens <= 0 and self.params.n_predict != -1):
# If we arent in instruction mode, fix the current generation by appending the antiprompt.
# Makes it so if chat ends prematurely you dont append the AI's text etc.
if not self.params.instruct:
self.embd_inp += self.first_antiprompt[0]
self.n_remain = self.params.n_predict
2023-04-03 20:54:46 +00:00
break
self.params.interactive_start = False
def __enter__(self):
return self
def __exit__(self, type, value, tb):
self.exit()
def exit(self):
llama_cpp.llama_free(self.ctx)
self.set_color(CONSOLE_COLOR_DEFAULT)
2023-04-04 09:48:48 +00:00
# return past text
2023-04-03 20:54:46 +00:00
def past(self):
for id in self.last_n_tokens[-self.n_past:]:
yield llama_cpp.llama_token_to_str(self.ctx, id).decode("utf-8")
2023-04-04 09:48:48 +00:00
# write input
2023-04-03 20:54:46 +00:00
def input(self, prompt: str):
if (self.params.instruct and self.last_n_tokens[-len(self.inp_prefix):] != self.inp_prefix):
2023-04-04 09:48:48 +00:00
self.embd_inp += self.inp_prefix
self.embd_inp += self._tokenize(prompt)
if (self.params.instruct):
2023-04-04 09:48:48 +00:00
self.embd_inp += self.inp_suffix
2023-04-03 20:54:46 +00:00
2023-04-04 09:48:48 +00:00
# write output
2023-04-03 20:54:46 +00:00
def output(self):
self.remaining_tokens = self.params.n_predict
2023-04-03 20:54:46 +00:00
for id in self.generate():
yield llama_cpp.llama_token_to_str(self.ctx, id).decode("utf-8")
# read user input
def read_input(self):
out = ""
while (t := input()).endswith("\\"):
out += t[:-1] + "\n"
return out + t + "\n"
# interactive mode
def interact(self):
for i in self.output():
print(i,end="",flush=True)
self.params.input_echo = False
while self.params.interactive:
self.set_color(CONSOLE_COLOR_USER_INPUT)
if (self.params.instruct):
print('\n> ', end="")
self.input(self.read_input())
else:
print(self.params.input_prefix, end="")
self.input(f"{self.params.input_prefix}{self.read_input()}{self.params.output_postfix}")
print(self.params.output_postfix,end="")
self.set_color(CONSOLE_COLOR_DEFAULT)
try:
for i in self.output():
print(i,end="",flush=True)
except KeyboardInterrupt:
self.set_color(CONSOLE_COLOR_DEFAULT)
if not self.params.instruct:
print(self.params.fix_prefix,end="")
self.input(self.params.fix_prefix)
2023-04-03 20:54:46 +00:00
if __name__ == "__main__":
from datetime import datetime
USER_NAME="User"
AI_NAME="ChatLLaMa"
2023-04-04 09:48:48 +00:00
2023-04-03 20:54:46 +00:00
time_now = datetime.now()
prompt = f"""Text transcript of a never ending dialog, where {USER_NAME} interacts with an AI assistant named {AI_NAME}.
{AI_NAME} is helpful, kind, honest, friendly, good at writing and never fails to answer {USER_NAME}s requests immediately and with details and precision.
There are no annotations like (30 seconds passed...) or (to himself), just what {USER_NAME} and {AI_NAME} say aloud to each other.
The dialog lasts for years, the entirety of it is shared below. It's 10000 pages long.
The transcript only includes text, it does not include markup like HTML and Markdown.
{USER_NAME}: Hello, {AI_NAME}!
{AI_NAME}: Hello {USER_NAME}! How may I help you today?
{USER_NAME}: What time is it?
{AI_NAME}: It is {time_now.strftime("%H:%M")}.
{USER_NAME}: What year is it?
{AI_NAME}: We are in {time_now.strftime("%Y")}.
{USER_NAME}: What is a cat?
{AI_NAME}: A cat is a domestic species of small carnivorous mammal. It is the only domesticated species in the family Felidae.
{USER_NAME}: Name a color.
{AI_NAME}: Blue
{USER_NAME}:"""
args = gpt_params_parse()
params = GptParams(**vars(args))
with LLaMAInteract(params) as m:
m.interact()