update version

lm_eval/api/filter.py

+from abc import ABC, abstractmethod
+from dataclasses import dataclass
+from typing import Callable, Iterable, List, Union
+
+from lm_eval.api.instance import Instance
+
+
+class Filter(ABC):
+    """
+    Filter classes operate on a per-task level.
+    They take all model outputs (`instance.resps` for all `task.instances`)
+    across all instances of a task, and perform operations.
+    In a single run, one can configure any number of separate filters or lists of filters.
+
+    """
+
+    def __init__(self, **kwargs) -> None:
+        """
+        Can define custom behavior here, if an individual instantiation of a Filter class should have state.
+        """
+
+    @abstractmethod
+    def apply(self, resps: Union[List, Iterable], docs: List[dict]) -> Iterable:
+        """
+        Defines the operation to perform on a list of the `inst.resps` properties of `Instance` objects.
+        Should return the list of (filtered) response lists *in the same order as they were input*, e.g.
+        if pass in [<inst.resps for instance 0>, <inst.resps for instance 1>] should return
+        [<filtered resps for instance 0>, <filtered resps for instance 1>]
+        """
+        return resps
+
+
+@dataclass
+class FilterEnsemble:
+    """
+    FilterEnsemble creates a pipeline applying multiple filters.
+    Its intended usage is to stack multiple post-processing steps in order.
+    `task.apply_filters` should use a list of FilterEnsemble classes that it stores, to apply each
+    pipeline separately.
+    """
+
+    name: str
+    filters: List[Callable[[], Filter]]
+
+    def apply(self, instances: List[Instance]) -> None:
+        resps, docs = zip(*((inst.resps, inst.doc) for inst in instances))
+        resps, docs = list(resps), list(docs)
+
+        for f in self.filters:
+            # apply filters in sequence
+            resps = f().apply(resps, docs)
+
+        # add the end results after filtering to filtered_requests of their respective source instances.
+        # has key `self.name`: each FilterEnsemble applied in a given run should use a different name.
+        for inst, resp in zip(instances, resps):
+            inst.filtered_resps[self.name] = resp

lm_eval/api/instance.py

+from dataclasses import dataclass, field
+from typing import Literal, Optional, Tuple
+
+
+OutputType = Literal[
+    "loglikelihood", "loglikelihood_rolling", "generate_until", "multiple_choice"
+]
+
+
+@dataclass
+class Instance:
+    request_type: OutputType
+    doc: dict
+    arguments: tuple
+    idx: int
+    metadata: Tuple[Optional[str], Optional[int], Optional[int]] = field(
+        default_factory=lambda: (None, None, None)
+    )
+    resps: list = field(default_factory=list)
+    filtered_resps: dict = field(default_factory=dict)
+
+    # initialized after init
+    task_name: Optional[str] = None
+    doc_id: Optional[int] = None
+    repeats: Optional[int] = None
+
+    def __post_init__(self) -> None:
+        # unpack metadata field
+        self.task_name, self.doc_id, self.repeats = self.metadata
+
+    @property
+    def args(self):
+        """
+        Returns (string,) where `string` is the string to calculate loglikelihood over
+        """
+        return (
+            self.arguments if isinstance(self.arguments, tuple) else (self.arguments,)
+        )

lm_eval/api/model.py

+import abc
+import hashlib
+import json
+import logging
+import os
+from typing import List, Optional, Tuple, Type, TypeVar
+
+import transformers
+from sqlitedict import SqliteDict
+from tqdm import tqdm
+
+from lm_eval import utils
+
+
+eval_logger = logging.getLogger("lm-eval")
+
+T = TypeVar("T", bound="LM")
+
+
+class LM(abc.ABC):
+    def __init__(self) -> None:
+        """Defines the interface that should be implemented by all LM subclasses.
+        LMs are assumed to take text (strings) as input and yield strings as output
+        (inputs/outputs should be tokenization-agnostic.)
+
+        """
+        # set rank and world size to a single process, by default.
+        self._rank = 0
+        self._world_size = 1
+        self.cache_hook = CacheHook(None)
+
+    @abc.abstractmethod
+    def loglikelihood(self, requests) -> List[Tuple[float, bool]]:
+        """Compute log-likelihood of generating a continuation from a context.
+        Downstream tasks should attempt to use loglikelihood instead of other
+        LM calls whenever possible.
+
+        :param requests: list[Instance]
+            A list of Instance objects, with property `args` which returns a tuple (context, continuation).
+            `context: str`
+                Context string. Implementations of LM must be able to handle an
+                empty context string.
+            `continuation: str`
+                The continuation over which log likelihood will be calculated. If
+                there is a word boundary, the space should be in the continuation.
+                For example, context="hello" continuation=" world" is correct.
+
+        :return: list[tuple[float, bool]]
+            A list of pairs (logprob, isgreedy)
+            `logprob: float`
+                The log probability of `continuation`.
+            `isgreedy`:
+                Whether `continuation` would be generated by greedy sampling from `context`.
+        """
+        pass
+
+    @abc.abstractmethod
+    def loglikelihood_rolling(self, requests) -> List[Tuple[float]]:
+        """Compute full log-likelihood of a string, with no truncation, for perplexity computation
+        - We will use the full max context length of the model.
+        - For inputs that exceed the max context length, we divide the tokenized string into chunks of up to
+        the max context length.
+        - IMPORTANT: Each document's loglikelihood/perplexity is computed *separately*, unlike other implementations
+          which may simply concatenate multiple documents together.
+        - IMPORTANT: We maximize the amount of context for each prediction. Specifically, for inputs that we break into
+          multiple chunks, the last input will still a full-sized context.
+          Example:
+            Input tokens: [ 0 1 2 3 4 5 6 7 8 9 ]
+            Prefix: BOS/EOS
+            Max context length: 4
+            Resulting input/prediction pairs:
+
+                INPUT:  BOS   0   1   2
+                PRED:     0   1   2   3
+
+                INPUT:    3   4   5   6
+                PRED:     4   5   6   7
+
+                INPUT:    5   6   7   8
+                PRED:             8   9
+
+          Observe that:
+            1. Each token is predicted exactly once
+            2. For the last pair, we provide the full context, but only score the last two tokens
+
+        :param requests: list[Instance]
+            A list of Instance objects with property `args` which returns a tuple (context,).
+            string: str
+                String for which we are computing overall loglikelihood
+        :return: list[tuple[float]]
+            A list of tuples (logprob,)
+            logprob: float
+                The log probability of `context` conditioned on the BOS/EOS token.
+                Can also be overridden for custom cases by `prefix_token_id`.
+        """
+        pass
+
+    # TODO: Add an optional max length
+    @abc.abstractmethod
+    def generate_until(self, requests) -> List[str]:
+        """Generate greedily until a stopping sequence
+
+        :param requests: list[Instance]
+            A list of Instance objects with property `args` which returns a tuple (context, until).
+            context: str
+                Context string
+            until: [str]
+                The string sequences to generate until. These string sequences
+                may each span across multiple tokens, or may be part of one token.
+        :return: list[str]
+            A list of strings continuation
+            continuation: str
+                The generated continuation.
+        """
+        pass
+
+    @classmethod
+    def create_from_arg_string(
+        cls: Type[T], arg_string: str, additional_config: Optional[dict] = None
+    ) -> T:
+        """
+        Creates an instance of the LM class using the given argument string and additional config.
+
+        Parameters:
+        - arg_string: A string containing arguments in the format key1=value1,key2=value2.
+        - additional_config: Optional dictionary containing additional configuration parameters.
+
+        Returns:
+        - Instance of the LM class.
+        """
+        additional_config = {} if additional_config is None else additional_config
+        args = utils.simple_parse_args_string(arg_string)
+        args2 = {k: v for k, v in additional_config.items() if v is not None}
+        return cls(**args, **args2)
+
+    @classmethod
+    def create_from_arg_obj(
+        cls: Type[T], arg_dict: dict, additional_config: Optional[dict] = None
+    ) -> T:
+        """
+        Creates an instance of the LM class using the given arg_obj
+
+        Parameters:
+        - arg_obj: A dict containing arguments in the format key1=value1,key2=value2.
+        - additional_config: Optional dictionary containing additional configuration parameters.
+
+        Returns:
+        - Instance of the LM class.
+        """
+
+        additional_config = {} if additional_config is None else additional_config
+        additional_config = {
+            k: v for k, v in additional_config.items() if v is not None
+        }
+
+        return cls(**arg_dict, **additional_config)
+
+    @property
+    def rank(self):
+        # used in the case of parallelism. Hardcoded to
+        # ensure no errors arise using API models which do
+        # not support multi-device parallelism nor expect it.
+        return self._rank
+
+    @property
+    def world_size(self):
+        # used in the case of parallelism. Hardcoded to
+        # ensure no errors arise using API models which do
+        # not support multi-device parallelism nor expect it.
+        return self._world_size
+
+    def set_cache_hook(self, cache_hook) -> None:
+        self.cache_hook = cache_hook
+
+
+### SQLite-based caching of LM responses
+def hash_args(attr, args):
+    dat = json.dumps([attr] + list(args))
+    return hashlib.sha256(dat.encode("utf-8")).hexdigest()
+
+
+class CacheHook:
+    def __init__(self, cachinglm) -> None:
+        if cachinglm is None:
+            self.dbdict = None
+            return
+
+        self.dbdict = cachinglm.dbdict
+
+    def add_partial(self, attr, req, res) -> None:
+        if self.dbdict is None:
+            return
+        hsh = hash_args(attr, req)
+        self.dbdict[hsh] = res
+
+
+class CachingLM:
+    def __init__(self, lm, cache_db) -> None:
+        """LM wrapper that returns cached results if they exist, and uses the underlying LM if not.
+
+        :param lm: LM
+            Underlying LM
+        :param cache_db: str
+            Path to cache db
+        """
+        self.lm = lm
+        self.cache_db = cache_db
+        if os.path.dirname(cache_db):
+            os.makedirs(os.path.dirname(cache_db), exist_ok=True)
+        self.dbdict = SqliteDict(cache_db, autocommit=True)
+
+        # add hook to lm
+        lm.set_cache_hook(self.get_cache_hook())
+
+    def __getattr__(self, attr):
+        lm_attr = getattr(self.lm, attr)
+        if not callable(lm_attr):
+            return lm_attr
+
+        def fn(requests):
+            res = []
+            remaining_reqs = []
+            warned = False
+            # figure out which ones are cached and which ones are new
+            eval_logger.info(
+                f"Loading '{attr}' responses from cache '{self.cache_db}' where possible..."
+            )
+            for req in tqdm(requests, desc="Checking cached requests"):
+                hsh = hash_args(attr, req.args)
+                if attr == "generate_until" and req.args[1].get("do_sample", False):
+                    # when we are doing non-greedy generation, don't use the cache
+                    # (else every "randomly sampled" generation would be identical for repeats > 1).
+                    if not warned:
+                        eval_logger.warning(
+                            f"Arguments to lm.generate_until() '{req.args[1]}' include non-deterministic sampling. Caching will not be performed for such requests."
+                        )
+                        warned = True
+                    res.append(None)
+                    remaining_reqs.append(req)
+                elif hsh in self.dbdict:
+                    ob = self.dbdict[hsh]
+
+                    assert ob is not None
+
+                    res.append(ob)
+                else:
+                    res.append(None)
+                    remaining_reqs.append(req)
+            eval_logger.info(
+                f"Cached requests: {len(requests) - len(remaining_reqs)}, Requests remaining: {len(remaining_reqs)}"
+            )
+            # actually run the LM on the requests that do not have cached results
+            rem_res = getattr(self.lm, attr)(remaining_reqs)
+
+            # stick the new ones back into the list and also cache any of the new ones
+            resptr = 0
+            for req, r in zip(remaining_reqs, rem_res):
+                while res[resptr] is not None:
+                    resptr += 1
+
+                res[resptr] = r
+
+                # caching
+                hsh = hash_args(attr, req.args)
+                self.dbdict[hsh] = r
+            self.dbdict.commit()
+
+            return res
+
+        return fn
+
+    def get_cache_hook(self):
+        return CacheHook(self)
+
+
+class TemplateLM(LM):
+    """
+    A class acting as intermediary between the LM base class
+    and boilerplate often included in other LM subclasses.
+    """
+
+    @property
+    @abc.abstractmethod
+    def eot_token_id(self):
+        pass
+
+    @property
+    def prefix_token_id(self):
+        # it is used as prefix for loglikelihood
+        return self.eot_token_id
+
+    @abc.abstractmethod
+    def tok_encode(self, string: str, **kwargs):
+        pass
+
+    @abc.abstractmethod
+    def _loglikelihood_tokens(self, requests, **kwargs):
+        pass
+
+    def _encode_pair(self, context, continuation):
+        n_spaces = len(context) - len(context.rstrip())
+        if n_spaces > 0:
+            continuation = context[-n_spaces:] + continuation
+            context = context[:-n_spaces]
+
+        model_class = getattr(self, "AUTO_MODEL_CLASS", None)
+
+        if model_class == transformers.AutoModelForSeq2SeqLM:
+            context_enc = self.tok_encode(context)
+            continuation_enc = self.tok_encode(continuation, add_special_tokens=False)
+        else:
+            whole_enc = self.tok_encode(context + continuation)
+            context_enc = self.tok_encode(context)
+
+            context_enc_len = len(context_enc)
+            continuation_enc = whole_enc[context_enc_len:]
+
+        return context_enc, continuation_enc
+
+    def loglikelihood(
+        self, requests, disable_tqdm: bool = False
+    ) -> List[Tuple[float, bool]]:
+        new_reqs = []
+        for context, continuation in [req.args for req in requests]:
+            if context == "":
+                # BOS or EOS as context
+                context_enc, continuation_enc = (
+                    [self.prefix_token_id],
+                    self.tok_encode(continuation),
+                )
+            else:
+                context_enc, continuation_enc = self._encode_pair(context, continuation)
+
+            new_reqs.append(((context, continuation), context_enc, continuation_enc))
+
+        return self._loglikelihood_tokens(new_reqs, disable_tqdm=disable_tqdm)
+
+    @abc.abstractmethod
+    def loglikelihood_rolling(
+        self, requests, disable_tqdm: bool = False
+    ) -> List[Tuple[float, bool]]:
+        pass
+
+    @abc.abstractmethod
+    def generate_until(self, requests, disable_tqdm: bool = False) -> List[str]:
+        pass

lm_eval/caching/cache.py

+import hashlib
+import os
+
+import dill
+
+from lm_eval.utils import eval_logger
+
+
+MODULE_DIR = os.path.dirname(os.path.realpath(__file__))
+
+OVERRIDE_PATH = os.getenv("LM_HARNESS_CACHE_PATH")
+
+
+PATH = OVERRIDE_PATH if OVERRIDE_PATH else f"{MODULE_DIR}/.cache"
+
+# This should be sufficient for uniqueness
+HASH_INPUT = "EleutherAI-lm-evaluation-harness"
+
+HASH_PREFIX = hashlib.sha256(HASH_INPUT.encode("utf-8")).hexdigest()
+
+FILE_SUFFIX = f".{HASH_PREFIX}.pickle"
+
+
+def load_from_cache(file_name):
+    try:
+        path = f"{PATH}/{file_name}{FILE_SUFFIX}"
+
+        with open(path, "rb") as file:
+            cached_task_dict = dill.loads(file.read())
+            return cached_task_dict
+
+    except Exception:
+        eval_logger.debug(f"{file_name} is not cached, generating...")
+        pass
+
+
+def save_to_cache(file_name, obj):
+    if not os.path.exists(PATH):
+        os.mkdir(PATH)
+
+    file_path = f"{PATH}/{file_name}{FILE_SUFFIX}"
+
+    eval_logger.debug(f"Saving {file_path} to cache...")
+    with open(file_path, "wb") as file:
+        file.write(dill.dumps(obj))
+
+
+# NOTE the "key" param is to allow for flexibility
+def delete_cache(key: str = ""):
+    files = os.listdir(PATH)
+
+    for file in files:
+        if file.startswith(key) and file.endswith(FILE_SUFFIX):
+            file_path = f"{PATH}/{file}"
+            os.unlink(file_path)