__init__.py 20.7 KB
Newer Older
1
import collections
2
3
import logging
import os
4
from functools import partial
5
from typing import Dict, List, Mapping, Optional, Union
&'s avatar
& committed
6

7
from lm_eval import utils
8
from lm_eval.api.task import ConfigurableGroup, ConfigurableTask, GroupConfig, Task
haileyschoelkopf's avatar
haileyschoelkopf committed
9
from lm_eval.evaluator_utils import get_subtask_list
10

11
12

GROUP_ONLY_KEYS = list(GroupConfig().to_dict().keys())
lintangsutawika's avatar
lintangsutawika committed
13

14

15
16
17
class TaskManager:
    """TaskManager indexes all tasks from the default `lm_eval/tasks/`
    and an optional directory if provided.
18

19
20
    """

21
    def __init__(self, verbosity="INFO", include_path: Optional[str] = None) -> None:
22
23
24
25
26
        self.verbosity = verbosity
        self.include_path = include_path
        self.logger = utils.eval_logger
        self.logger.setLevel(getattr(logging, f"{verbosity}"))

27
        self._task_index = self.initialize_tasks(include_path=include_path)
28
        self._all_tasks = sorted(list(self._task_index.keys()))
29

30
        self.task_group_map = collections.defaultdict(list)
31

32
33
    def initialize_tasks(self, include_path: Optional[str] = None):
        """Creates a dictionary of tasks index.
34
35
36
37
38
39
40
41
42
43
44
45

        :param include_path: str = None
            An additional path to be searched for tasks

        :return
            Dictionary of task names as key and task metadata
        """
        all_paths = [os.path.dirname(os.path.abspath(__file__)) + "/"]
        if include_path is not None:
            if isinstance(include_path, str):
                include_path = [include_path]
            all_paths.extend(include_path)
46

47
48
49
50
        task_index = {}
        for task_dir in all_paths:
            tasks = self._get_task_and_group(task_dir)
            task_index = {**tasks, **task_index}
lintangsutawika's avatar
format  
lintangsutawika committed
51

52
53
54
55
56
57
58
59
60
61
62
        return task_index

    @property
    def all_tasks(self):
        return self._all_tasks

    @property
    def task_index(self):
        return self._task_index

    def match_tasks(self, task_list):
63
        return utils.pattern_match(task_list, self.all_tasks)
64

65
    def _name_is_registered(self, name) -> bool:
66
67
68
69
        if name in self.all_tasks:
            return True
        return False

70
    def _name_is_task(self, name) -> bool:
lintangsutawika's avatar
lintangsutawika committed
71
        if self._name_is_registered(name) and (self.task_index[name]["type"] == "task"):
72
73
74
            return True
        return False

lintangsutawika's avatar
lintangsutawika committed
75
    def _name_is_tag(self, name) -> bool:
lintangsutawika's avatar
lintangsutawika committed
76
        if self._name_is_registered(name) and (self.task_index[name]["type"] == "tag"):
lintangsutawika's avatar
lintangsutawika committed
77
78
79
            return True
        return False

80
    def _name_is_group(self, name) -> bool:
81
82
83
        if self._name_is_registered(name) and (
            self.task_index[name]["type"] == "group"
        ):
84
85
86
87
            return True
        return False

    def _name_is_python_task(self, name):
88
89
90
        if self._name_is_registered(name) and (
            self.task_index[name]["type"] == "python_task"
        ):
91
92
93
            return True
        return False

94
    def _config_is_task(self, config) -> bool:
95
96
97
98
        if ("task" in config) and isinstance(config["task"], str):
            return True
        return False

99
    def _config_is_group(self, config) -> bool:
100
101
102
103
        if ("task" in config) and isinstance(config["task"], list):
            return True
        return False

104
    def _config_is_python_task(self, config) -> bool:
105
106
107
108
109
        if "class" in config:
            return True
        return False

    def _get_yaml_path(self, name):
110
111
        if name not in self.task_index:
            raise ValueError
112
113
114
        return self.task_index[name]["yaml_path"]

    def _get_config(self, name):
115
116
        if name not in self.task_index:
            raise ValueError
117
118
119
120
121
122
123
        yaml_path = self._get_yaml_path(name)
        if yaml_path == -1:
            return {}
        else:
            return utils.load_yaml_config(yaml_path, mode="full")

    def _get_tasklist(self, name):
124
125
        if self._name_is_task(name):
            raise ValueError
126
127
128
129
130
131
132
133
134
135
136
137
        return self.task_index[name]["task"]

    def _process_alias(self, config, group=None):
        # If the group is not the same as the original
        # group which the group alias was intended for,
        # Set the group_alias to None instead.
        if ("group_alias" in config) and ("group" in config) and group is not None:
            if config["group"] != group:
                config["group_alias"] = None
        return config

    def _load_individual_task_or_group(
138
        self,
139
140
141
142
        name_or_config: Optional[Union[str, dict]] = None,
        parent_name: Optional[str] = None,
        update_config: Optional[dict] = None,
    ) -> Mapping:
143
        def _load_task(config, task):
144
            if "include" in config:
145
146
                config = {
                    **utils.load_yaml_config(
147
                        yaml_path=None,
148
149
                        yaml_config={"include": config.pop("include")},
                        mode="full",
150
151
152
                    ),
                    **config,
                }
153
            if self._config_is_python_task(config):
154
155
156
157
158
                task_object = (
                    config["class"](config=config)
                    if isinstance(config["class"], ConfigurableTask)
                    else config["class"]()
                )
159
                # very scuffed: set task name here. TODO: fixme?
160
                task_object.config.task = config["task"]
161
162
            else:
                task_object = ConfigurableTask(config=config)
163

164
165
            return {task: task_object}

166
167
168
169
        def _get_group_and_subtask_from_config(config):
            group_name = ConfigurableGroup(config=config)
            subtask_list = []
            for task in group_name.config["task"]:
lintangsutawika's avatar
lintangsutawika committed
170
                if isinstance(task, str) and self._name_is_tag(task):
171
172
173
174
175
                    subtask_list.extend(self._get_tasklist(task))
                else:
                    subtask_list.append(task)
            return group_name, subtask_list

lintangsutawika's avatar
lintangsutawika committed
176
        def _process_group_config(config, update_config=None):
lintangsutawika's avatar
lintangsutawika committed
177
178
179
180
181
182
183
            if update_config is not None:
                config = {**config, **update_config}
            _update_config = {
                k: v for k, v in config.items() if k not in GROUP_ONLY_KEYS
            }
            if not bool(_update_config):
                _update_config = None
lintangsutawika's avatar
lintangsutawika committed
184
185

            group_config = {k: v for k, v in config.items() if k in GROUP_ONLY_KEYS}
lintangsutawika's avatar
lintangsutawika committed
186
            return group_config, _update_config
lintangsutawika's avatar
lintangsutawika committed
187

188
189
190
191
        if isinstance(name_or_config, str):
            if update_config is not None:
                # Process name_or_config as a dict instead
                name_or_config = {"task": name_or_config, **update_config}
192
193
194
            elif self._name_is_task(name_or_config) or self._name_is_python_task(
                name_or_config
            ):
195
                task_config = self._get_config(name_or_config)
196
                return _load_task(task_config, task=name_or_config)
197
            else:
198
199
200
                subtask_list = self._get_tasklist(name_or_config)
                if subtask_list == -1:
                    group_config = self._get_config(name_or_config)
lintangsutawika's avatar
lintangsutawika committed
201
                    group_config, update_config = _process_group_config(group_config)
lintangsutawika's avatar
lintangsutawika committed
202
203
204
                    group_name, subtask_list = _get_group_and_subtask_from_config(
                        group_config
                    )
lintangsutawika's avatar
lintangsutawika committed
205
                else:
206
207
208
                    if self._name_is_tag(name_or_config):
                        fn = partial(
                            self._load_individual_task_or_group,
lintangsutawika's avatar
lintangsutawika committed
209
210
211
212
213
214
                            update_config=name_or_config
                            if isinstance(name_or_config, dict)
                            else None,
                        )
                        return dict(
                            collections.ChainMap(*map(fn, reversed(subtask_list)))
215
216
217
218
219
                        )
                    else:
                        group_name = ConfigurableGroup(
                            config={"group": name_or_config, "task": subtask_list}
                        )
220

221
222
        if isinstance(name_or_config, dict):
            if self._config_is_task(name_or_config):
lintangsutawika's avatar
lintangsutawika committed
223
224
225
                name = name_or_config.pop("task")
                if update_config is not None:
                    name_or_config = {**name_or_config, **update_config}
226
                # If the name is registered as a group
lintangsutawika's avatar
lintangsutawika committed
227
228
229
230
231
232
233
234
235
236
                if self._name_is_group(name):
                    group_config = self._get_config(name)

                    group_config, update_config = _process_group_config(
                        group_config, name_or_config
                    )
                    group_name, subtask_list = _get_group_and_subtask_from_config(
                        group_config
                    )
                elif self._name_is_tag(name):
237
                    subtask_list = self._get_tasklist(name)
lintangsutawika's avatar
lintangsutawika committed
238
239
240
241
242
                    fn = partial(
                        self._load_individual_task_or_group,
                        update_config=name_or_config,
                    )
                    return dict(collections.ChainMap(*map(fn, reversed(subtask_list))))
243
244
245
246
247
248
                else:
                    if self._name_is_registered(name):
                        base_task_config = self._get_config(name)

                        # Check if this is a duplicate.
                        if parent_name is not None:
249
250
251
252
253
254
255
256
                            num_duplicate = len(
                                list(
                                    filter(
                                        lambda x: x.startswith(name),
                                        self.task_group_map[parent_name],
                                    )
                                )
                            )
257
258
259
260
                            if num_duplicate > 0:
                                name = f"{name}-{num_duplicate}"
                            self.task_group_map[parent_name].append(name)

261
262
263
264
                        task_config = {
                            **base_task_config,
                            **name_or_config,
                        }
265
266
                    else:
                        task_config = name_or_config
267
                    return _load_task(task_config, task=name)
268
            else:
lintangsutawika's avatar
lintangsutawika committed
269
                group_config, update_config = _process_group_config(name_or_config)
lintangsutawika's avatar
lintangsutawika committed
270
271
272
                group_name, subtask_list = _get_group_and_subtask_from_config(
                    group_config
                )
273

274
275
276
277
278
        fn = partial(
            self._load_individual_task_or_group,
            parent_name=group_name,
            update_config=update_config,
        )
279
280
281
        return {
            group_name: dict(collections.ChainMap(*map(fn, reversed(subtask_list))))
        }
282

283
    def load_task_or_group(self, task_list: Optional[Union[str, list]] = None) -> dict:
284
        """Loads a dictionary of task objects from a list
285

286
287
        :param task_list: Union[str, list] = None
            Single string or list of string of task names to be loaded
288

289
290
291
292
293
        :return
            Dictionary of task objects
        """
        if isinstance(task_list, str):
            task_list = [task_list]
294

295
        all_loaded_tasks = dict(
296
            collections.ChainMap(*map(self._load_individual_task_or_group, task_list))
297
298
299
300
301
302
303
        )
        return all_loaded_tasks

    def load_config(self, config: Dict):
        return self._load_individual_task_or_group(config)

    def _get_task_and_group(self, task_dir: str):
304
        """Creates a dictionary of tasks index with the following metadata,
lintangsutawika's avatar
lintangsutawika committed
305
        - `type`, that can be either `task`, `python_task`, `group` or `tags`.
306
307
            `task` refer to regular task configs, `python_task` are special
            yaml files that only consists of `task` and `class` parameters.
lintangsutawika's avatar
lintangsutawika committed
308
309
            `group` are group configs. `tags` are labels that can be assigned
            to tasks to assist in sorting and calling tasks of certain themes.
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
        - `yaml_path`, path to the yaml file. If the entry is a `group` that
            was configured through a task config, the yaml_path will be -1
            and all subtasks will be listed in `task` (see below)
        - `task`, reserved for entries with `type` as `group`. This will list
            all subtasks. When a group config is created (as opposed to task
            config having `group` parameter set), this will be set to -1 to
            avoid recursive indexing. The whole list of subtasks will be loaded
            at evaluation.

        :param task_dir: str
            A directory to check for tasks

        :return
            Dictionary of task names as key and task metadata
        """
325
326
        # TODO: remove group in next release
        print_info = True
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
        tasks_and_groups = collections.defaultdict()
        for root, _, file_list in os.walk(task_dir):
            for f in file_list:
                if f.endswith(".yaml"):
                    yaml_path = os.path.join(root, f)
                    config = utils.load_yaml_config(yaml_path, mode="simple")
                    if self._config_is_python_task(config):
                        # This is a python class config
                        tasks_and_groups[config["task"]] = {
                            "type": "python_task",
                            "yaml_path": yaml_path,
                        }
                    elif self._config_is_group(config):
                        # This is a group config
                        tasks_and_groups[config["group"]] = {
                            "type": "group",
343
344
345
346
347
                            "task": -1,  # This signals that
                            # we don't need to know
                            # the task list for indexing
                            # as it can be loaded
                            # when called.
348
349
                            "yaml_path": yaml_path,
                        }
lintangsutawika's avatar
lintangsutawika committed
350

351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
                        # # Registered the level 1 tasks from a group config
                        # for config in config["task"]:
                        #     if isinstance(config, dict) and self._config_is_task(config):
                        #         task = config["task"]
                        #         tasks_and_groups[task] = {
                        #             "type": "task",
                        #             "yaml_path": yaml_path,
                        #             }

                    elif self._config_is_task(config):
                        # This is a task config
                        task = config["task"]
                        tasks_and_groups[task] = {
                            "type": "task",
                            "yaml_path": yaml_path,
366
                        }
367

368
369
370
371
372
                        # TODO: remove group in next release
                        for attr in ["tag", "group"]:
                            if attr in config:
                                if attr == "group" and print_info:
                                    self.logger.info(
haileyschoelkopf's avatar
haileyschoelkopf committed
373
                                        "`group` and `group_alias` keys in tasks' configs will no longer be used in the next release of lm-eval. "
374
                                        "`tag` will be used to allow to call a collection of tasks just like `group`. "
375
376
377
378
                                        "`group` will be removed in order to not cause confusion with the new ConfigurableGroup "
                                        "which will be the offical way to create groups with addition of group-wide configuations."
                                    )
                                    print_info = False
379
                                    # attr = "tag"
380
381
382
383
384
385
386
387
388
389
390
391
392
393

                                attr_list = config[attr]
                                if isinstance(attr_list, str):
                                    attr_list = [attr_list]

                                for tag in attr_list:
                                    if tag not in tasks_and_groups:
                                        tasks_and_groups[tag] = {
                                            "type": "tag",
                                            "task": [task],
                                            "yaml_path": -1,
                                        }
                                    else:
                                        tasks_and_groups[tag]["task"].append(task)
394
395
396
397
                    else:
                        self.logger.debug(f"File {f} in {root} could not be loaded")

        return tasks_and_groups
lintangsutawika's avatar
lintangsutawika committed
398

399

400
401
402
403
404
405
406
def get_task_name_from_config(task_config: Dict[str, str]) -> str:
    if "task" in task_config:
        return task_config["task"]
    if "dataset_name" in task_config:
        return "{dataset_path}_{dataset_name}".format(**task_config)
    else:
        return "{dataset_path}".format(**task_config)
lintangsutawika's avatar
lintangsutawika committed
407

408

lintangsutawika's avatar
lintangsutawika committed
409
def get_task_name_from_object(task_object):
410
411
    if hasattr(task_object, "config"):
        return task_object._config["task"]
lintangsutawika's avatar
lintangsutawika committed
412
413
414
415
416
417
418
419
420

    # TODO: scrap this
    # this gives a mechanism for non-registered tasks to have a custom name anyways when reporting
    return (
        task_object.EVAL_HARNESS_NAME
        if hasattr(task_object, "EVAL_HARNESS_NAME")
        else type(task_object).__name__
    )

421

haileyschoelkopf's avatar
haileyschoelkopf committed
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
def _check_duplicates(task_dict: dict) -> List[str]:
    """helper function solely used in validating get_task_dict output.
    Takes the output of lm_eval.evaluator_utils.get_subtask_list and
    returns a list of all leaf subtasks contained within, and errors if any such leaf subtasks are
    "oversubscribed" to several disjoint groups.
    """
    subtask_names = []
    for key, value in task_dict.items():
        subtask_names.extend(value)

    duplicate_tasks = {
        task_name for task_name in subtask_names if subtask_names.count(task_name) > 1
    }

    # locate the potentially problematic groups that seem to 'compete' for constituent subtasks
    competing_groups = [
        group
        for group in task_dict.keys()
        if len(set(task_dict[group]).intersection(duplicate_tasks)) > 0
    ]

    if len(duplicate_tasks) > 0:
        raise ValueError(
            f"Found 1 or more tasks while trying to call get_task_dict() that were members of more than 1 called group: {list(duplicate_tasks)}. Offending groups: {competing_groups}. Please call groups which overlap their constituent tasks in separate evaluation runs."
        )


449
def get_task_dict(
450
    task_name_list: Union[str, List[Union[str, Dict, Task]]],
451
    task_manager: Optional[TaskManager] = None,
452
):
453
    """Creates a dictionary of task objects from either a name of task, config, or prepared Task object.
lintangsutawika's avatar
lintangsutawika committed
454

455
456
457
458
459
460
461
    :param task_name_list: List[Union[str, Dict, Task]]
        Name of model or LM object, see lm_eval.models.get_model
    :param task_manager: TaskManager = None
        A TaskManager object that stores indexed tasks. If not set,
        task_manager will load one. This should be set by the user
        if there are additional paths that want to be included
        via `include_path`
462

463
464
465
    :return
        Dictionary of task objects
    """
haileyschoelkopf's avatar
haileyschoelkopf committed
466

467
    task_name_from_string_dict = {}
468
469
470
    task_name_from_config_dict = {}
    task_name_from_object_dict = {}

471
    if isinstance(task_name_list, str):
lintangsutawika's avatar
lintangsutawika committed
472
        task_name_list = [task_name_list]
473
474
475
476
477
478
479
480
481
    elif isinstance(task_name_list, list):
        if not all([isinstance(task, (str, dict, Task)) for task in task_name_list]):
            raise TypeError(
                "Expected all list items to be of types 'str', 'dict', or 'Task', but at least one entry did not match."
            )
    else:
        raise TypeError(
            f"Expected a 'str' or 'list' but received {type(task_name_list)}."
        )
lintangsutawika's avatar
lintangsutawika committed
482

483
    string_task_name_list = [task for task in task_name_list if isinstance(task, str)]
484
485
486
    others_task_name_list = [
        task for task in task_name_list if not isinstance(task, str)
    ]
487
488
489
    if len(string_task_name_list) > 0:
        if task_manager is None:
            task_manager = TaskManager()
lintangsutawika's avatar
lintangsutawika committed
490

491
492
493
        task_name_from_string_dict = task_manager.load_task_or_group(
            string_task_name_list
        )
494

495
496
    for task_element in others_task_name_list:
        if isinstance(task_element, dict):
497
498
            task_name_from_config_dict = {
                **task_name_from_config_dict,
499
                **task_manager.load_config(config=task_element),
500
501
502
503
504
            }

        elif isinstance(task_element, Task):
            task_name_from_object_dict = {
                **task_name_from_object_dict,
lintangsutawika's avatar
lintangsutawika committed
505
                get_task_name_from_object(task_element): task_element,
506
            }
lintangsutawika's avatar
lintangsutawika committed
507

508
    if not set(task_name_from_string_dict.keys()).isdisjoint(
lintangsutawika's avatar
lintangsutawika committed
509
        set(task_name_from_object_dict.keys())
510
511
    ):
        raise ValueError
512

haileyschoelkopf's avatar
haileyschoelkopf committed
513
    final_task_dict = {
514
        **task_name_from_string_dict,
lintangsutawika's avatar
lintangsutawika committed
515
        **task_name_from_config_dict,
516
        **task_name_from_object_dict,
lintangsutawika's avatar
lintangsutawika committed
517
    }
haileyschoelkopf's avatar
haileyschoelkopf committed
518
519
520
521
522
523
524
525

    # behavior can get odd if one tries to invoke several groups that "compete" for the same task.
    # (notably, because one could request several num_fewshot values at once in GroupConfig overrides for the subtask
    # and we'd be unsure which to use and report.)
    # we explicitly check and error in this case.
    _check_duplicates(get_subtask_list(final_task_dict))

    return final_task_dict