Update caching and add tokenizer to create_states_mapping

outlines/caching.py

@@ -1,15 +1,40 @@
 import asyncio
 import functools
-import hashlib
 import os
 from typing import Callable, Optional
 
 import cloudpickle
-from diskcache import Cache
+from diskcache import Cache, Disk
+from diskcache.core import ENOVAL, UNKNOWN, args_to_key, full_name
 
 _caching_enabled = True
 
 
+class CloudpickleDisk(Disk):
+ def __init__(self, directory, compress_level=1, **kwargs):
+ self.compress_level = compress_level
+ super().__init__(directory, **kwargs)
+
+ def put(self, key):
+ data = cloudpickle.dumps(key)
+ return super().put(data)
+
+ def get(self, key, raw):
+ data = super().get(key, raw)
+ return cloudpickle.loads(data)
+
+ def store(self, value, read, key=UNKNOWN):
+ if not read:
+ value = cloudpickle.dumps(value)
+ return super().store(value, read, key=key)
+
+ def fetch(self, mode, filename, value, read):
+ data = super().fetch(mode, filename, value, read)
+ if not read:
+ data = cloudpickle.loads(data)
+ return data
+
+
 @functools.lru_cache(1)
 def get_cache():
  """Get the context object that contains previously-computed return values.
@@ -26,7 +51,12 @@ def get_cache():
 
  home_dir = os.path.expanduser("~")
  cache_dir = os.environ.get("OUTLINES_CACHE_DIR", f"{home_dir}/.cache/outlines")
- memory = Cache(cache_dir, eviction_policy="none", cull_limit=0)
+ memory = Cache(
+ cache_dir,
+ eviction_policy="none",
+ cull_limit=0,
+ disk=CloudpickleDisk,
+ )
 
  # ensure if version upgrade occurs, old cache is pruned
  if outlines_version != memory.get("__version__"):
@@ -36,63 +66,72 @@ def get_cache():
  return memory
 
 
-def hash_arguments(*args, **kwargs) -> str:
- """Create a hash out of the args and kwargs provided"""
- result = hashlib.md5()
- for item in list(args) + sorted(kwargs.items()):
- result.update(cloudpickle.dumps(item))
- return result.hexdigest()
-
-
-def cache(key_function: Optional[Callable] = None):
+def cache(expire: Optional[float] = None, typed=False, ignore=()):
  """Caching decorator for memoizing function calls.
+
  The cache key is created based on the values returned by the key_function callable
  if provided or based on the arguments of the decorated function directly otherwise
+
+ This is based on `diskcache`'s `memoize`.
+
  Parameters
  ----------
- key_function
- A callable function used to generate a unique key for each function call. It's
- called with the arguments of the decorated function as arguments
+ expire
+ Seconds until arguments expire.
+ typed
+ Cache different types separately.
+ ignore
+ Positional or keyword arguments to ignore.
+
  Returns
  -------
- A decorator function that can be applied to other functions.
+  A decorator function that can be applied to other functions.
  """
 
  def decorator(cached_function: Callable):
  memory = get_cache()
 
- def wrapper(*args, **kwargs):
- if not _caching_enabled:
- return cached_function(*args, **kwargs)
- if key_function:
- key_args = key_function(*args, **kwargs)
- cache_key = hash_arguments(*key_args)
- else:
- cache_key = hash_arguments(*args, **kwargs)
- if cache_key in memory:
- return memory[cache_key]
- result = cached_function(*args, **kwargs)
- memory[cache_key] = result
- return result
-
- async def async_wrapper(*args, **kwargs):
- if not _caching_enabled:
- return await cached_function(*args, **kwargs)
- if key_function:
- key_args = key_function(*args, **kwargs)
- cache_key = hash_arguments(*key_args)
- else:
- cache_key = hash_arguments(*args, **kwargs)
- if cache_key in memory:
- return memory[cache_key]
- result = await cached_function(*args, **kwargs)
- memory[cache_key] = result
- return result
+ base = (full_name(cached_function),)
 
  if asyncio.iscoroutinefunction(cached_function):
- return async_wrapper
+
+ async def wrapper(*args, **kwargs):
+ if not _caching_enabled:
+ return await cached_function(*args, **kwargs)
+
+ cache_key = wrapper.__cache_key__(*args, **kwargs)
+ result = wrapper.__memory__.get(cache_key, default=ENOVAL, retry=True)
+
+ if result is ENOVAL:
+ result = await cached_function(*args, **kwargs)
+ wrapper.__memory__.set(cache_key, result, expire, retry=True)
+
+ return result
+
  else:
- return wrapper
+
+ def wrapper(*args, **kwargs):
+ if not _caching_enabled:
+ return cached_function(*args, **kwargs)
+
+ cache_key = wrapper.__cache_key__(*args, **kwargs)
+ result = wrapper.__memory__.get(cache_key, default=ENOVAL, retry=True)
+
+ if result is ENOVAL:
+ result = cached_function(*args, **kwargs)
+ wrapper.__memory__.set(cache_key, result, expire, retry=True)
+
+ return result
+
+ def __cache_key__(*args, **kwargs):
+ """Make key for cache given function arguments."""
+ return args_to_key(base, args, kwargs, typed, ignore)
+
+ wrapper.__cache_key__ = __cache_key__ # type: ignore
+ wrapper.__memory__ = memory # type: ignore
+ wrapper.__wrapped__ = cached_function # type: ignore
+
+ return wrapper
 
  return decorator
 

outlines/fsm/guide.py

@@ -105,44 +105,44 @@ def copy(self):
  return self
 
 
+@cache()
+def create_states_mapping(
+ regex_string: str, tokenizer: "Tokenizer"
+) -> Tuple[dict, set, set]:
+ """Create the variables related to the mapping between states and tokens
+ The parameters of the function are used for caching purpose
+ """
+ regex_pattern = interegular.parse_pattern(regex_string)
+ byte_fsm = make_byte_level_fsm(regex_pattern.to_fsm().reduce(), keep_utf8=True)
+ regex_fsm, _ = make_deterministic_fsm(byte_fsm)
+ states_to_token_maps, empty_token_ids = create_fsm_index_tokenizer(
+ regex_fsm, tokenizer
+ )
+
+ # We make sure that it is possible to generate strings in the language
+ # of the regular expression with the tokens present in the model's
+ # vocabulary.
+ if not any(
+ regex_fsm.finals.intersection(v.values()) for v in states_to_token_maps.values()
+ ):
+ raise ValueError(
+ "The vocabulary does not allow us to build a sequence that matches the input regex"
+ )
+
+ return states_to_token_maps, empty_token_ids, regex_fsm.finals
+
+
 class RegexGuide(Guide):
  """Guide to generate text in the language of a regular expression."""
 
  initial_state = 0
 
  def __init__(self, regex_string: str, tokenizer):
- @cache()
- def create_states_mapping(regex_string: str) -> Tuple[dict, set, set]:
- """Create the variables related to the mapping between states and tokens
- The parameters of the function are used for caching purpose
- """
- regex_pattern = interegular.parse_pattern(regex_string)
- byte_fsm = make_byte_level_fsm(
- regex_pattern.to_fsm().reduce(), keep_utf8=True
- )
- regex_fsm, _ = make_deterministic_fsm(byte_fsm)
- states_to_token_maps, empty_token_ids = create_fsm_index_tokenizer(
- regex_fsm, tokenizer
- )
-
- # We make sure that it is possible to generate strings in the language
- # of the regular expression with the tokens present in the model's
- # vocabulary.
- if not any(
- regex_fsm.finals.intersection(v.values())
- for v in states_to_token_maps.values()
- ):
- raise ValueError(
- "The vocabulary does not allow us to build a sequence that matches the input regex"
- )
-
- return states_to_token_maps, empty_token_ids, regex_fsm.finals
-
  (
  self.states_to_token_maps,
  self.empty_token_ids,
  fsm_finals,
- ) = create_states_mapping(regex_string)
+ ) = create_states_mapping(regex_string, tokenizer)
  self.eos_token_id = tokenizer.eos_token_id
  self.final_states = fsm_finals | {-1}
 

outlines/models/transformers.py

@@ -1,5 +1,7 @@
 from typing import TYPE_CHECKING, List, Optional, Tuple, Union
 
+from datasets.fingerprint import Hasher
+
 from outlines.models.tokenizer import Tokenizer
 
 if TYPE_CHECKING:
@@ -109,10 +111,15 @@ def __eq__(self, other):
  return NotImplemented
 
  def __hash__(self):
- from datasets.fingerprint import Hasher
-
  return hash(Hasher.hash(self.tokenizer))
 
+ def __getstate__(self):
+ state = {"tokenizer": self.tokenizer}
+ return state
+
+ def __setstate__(self, state):
+ self.__init__(state["tokenizer"])
+
 
 class Transformers:
  """Represents a `transformers` model."""

pyproject.toml

@@ -35,7 +35,8 @@ dependencies = [
  "referencing",
  "jsonschema",
  "requests",
- "tqdm"
+ "tqdm",
+ "datasets",
 ]
 dynamic = ["version"]
 
@@ -50,7 +51,6 @@ test = [
  "diff-cover",
  "accelerate",
  "beartype<0.16.0",
- "datasets",
  "responses",
  "llama-cpp-python",
  "huggingface_hub",

tests/generate/test_integration_transformers.py

@@ -1,6 +1,7 @@
 import datetime
 import re
 from enum import Enum
+from importlib import reload
 from typing import List, Union
 
 import pytest
@@ -11,7 +12,28 @@
 import outlines.models as models
 from outlines.fsm.regex import reduced_vocabulary
 from outlines.models.transformers import Transformers, TransformerTokenizer
-from outlines.samplers import beam_search, multinomial
+from outlines.samplers import beam_search, greedy, multinomial
+
+
+@pytest.fixture
+def temp_cache_dir():
+ import os
+ import tempfile
+
+ import outlines.caching
+ import outlines.fsm.guide
+
+ with tempfile.TemporaryDirectory() as tempdir:
+ os.environ["OUTLINES_CACHE_DIR"] = tempdir
+ outlines.caching.get_cache.cache_clear()
+ reload(outlines)
+ reload(outlines.fsm.guide)
+ cache_status = outlines.caching._caching_enabled
+ try:
+ outlines.caching._caching_enabled = True
+ yield
+ finally:
+ outlines.caching._caching_enabled = cache_status
 
 
 def test_transformers_integration_text():
@@ -632,3 +654,47 @@ def test_transformers_use_existing_model_and_tokenizer():
  model = Transformers(hf_model, hf_tokenizer)
  sequence = generate.text(model)("Write a short sentence ", rng=rng)
  assert isinstance(sequence, str)
+
+
+def test_RegexGuide_caching(temp_cache_dir):
+ import outlines.caching
+ from outlines.fsm.guide import create_states_mapping
+
+ assert outlines.caching._caching_enabled
+
+ regex = r"((25[0-5]|2[0-4]\d|[01]?\d\d?)\.){3}(25[0-5]|2[0-4]\d|[01]?\d\d?)"
+ prompt = "What is the IP address of the Google DNS servers? "
+
+ cache = outlines.caching.get_cache()
+
+ # Returns (hits, misses)
+ _ = cache.stats(enable=True)
+ assert cache.statistics
+
+ assert create_states_mapping.__memory__ is cache
+
+ model = models.transformers(
+ "hf-internal-testing/tiny-random-XLMRobertaXLForCausalLM"
+ )
+ generator = generate.regex(model, regex, sampler=greedy())
+ assert cache.stats() == (0, 1)
+
+ model_2 = models.transformers("hf-internal-testing/tiny-random-GPTJForCausalLM")
+ generator_2 = generate.regex(model_2, regex, sampler=greedy())
+ assert cache.stats() == (0, 2)
+
+ # These two different models and tokenizers should not have the same state
+ # mapping results
+ assert generator.fsm.states_to_token_maps != generator_2.fsm.states_to_token_maps
+
+ generator_3 = generate.regex(model_2, regex, sampler=greedy())
+ assert cache.stats() == (1, 2)
+ assert generator_2.fsm.states_to_token_maps == generator_3.fsm.states_to_token_maps
+
+ # Just for fun...
+ structured = generator(prompt, max_tokens=30)
+ structured_2 = generator_2(prompt, max_tokens=30)
+
+ assert re.fullmatch(regex, structured)
+ assert re.fullmatch(regex, structured_2)
+ assert structured != structured_2

tests/models/test_transformers.py

@@ -107,6 +107,14 @@ def test_tokenizer_eq_hash():
  tokenizer_hf = AutoTokenizer.from_pretrained("gpt2")
 
  tokenizer = TransformerTokenizer(tokenizer_hf)
- tokenizer2 = TransformerTokenizer(tokenizer_hf)
- assert tokenizer == tokenizer2
- assert hash(tokenizer) == hash(tokenizer2)
+ tokenizer_2 = TransformerTokenizer(tokenizer_hf)
+
+ assert tokenizer == tokenizer_2
+ assert hash(tokenizer) == hash(tokenizer_2)
+
+ tokenizer_hf_2 = AutoTokenizer.from_pretrained("gpt2")
+ tokenizer_hf_2.add_tokens(["test_token"])
+
+ tokenizer_3 = TransformerTokenizer(tokenizer_hf_2)
+ assert tokenizer != tokenizer_3
+ assert hash(tokenizer) != hash(tokenizer_3)