__init__.py 17.7 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
9
from lm_eval.api.task import ConfigurableGroup, ConfigurableTask, GroupConfig, Task

10
11

GROUP_ONLY_KEYS = list(GroupConfig().to_dict().keys())
lintangsutawika's avatar
lintangsutawika committed
12

13

14
15
16
class TaskManager:
    """TaskManager indexes all tasks from the default `lm_eval/tasks/`
    and an optional directory if provided.
17

18
19
    """

20
    def __init__(self, verbosity="INFO", include_path: Optional[str] = None) -> None:
21
22
23
24
25
        self.verbosity = verbosity
        self.include_path = include_path
        self.logger = utils.eval_logger
        self.logger.setLevel(getattr(logging, f"{verbosity}"))

26
        self._task_index = self.initialize_tasks(include_path=include_path)
27
        self._all_tasks = sorted(list(self._task_index.keys()))
28

29
        self.task_group_map = collections.defaultdict(list)
30

31
32
    def initialize_tasks(self, include_path: Optional[str] = None):
        """Creates a dictionary of tasks index.
33
34
35
36
37
38
39
40
41
42
43
44

        :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)
45

46
47
48
49
        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
50

51
52
53
54
55
56
57
58
59
60
61
        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):
62
        return utils.pattern_match(task_list, self.all_tasks)
63

64
    def _name_is_registered(self, name) -> bool:
65
66
67
68
        if name in self.all_tasks:
            return True
        return False

69
    def _name_is_task(self, name) -> bool:
70
71
72
        if self._name_is_registered(name) and (
            self.task_index[name]["type"] == "task"
        ):
73
74
75
            return True
        return False

lintangsutawika's avatar
lintangsutawika committed
76
77
78
79
80
81
82
    def _name_is_tag(self, name) -> bool:
        if self._name_is_registered(name) and (
            self.task_index[name]["type"] == "tag"
        ):
            return True
        return False

83
    def _name_is_group(self, name) -> bool:
84
85
86
        if self._name_is_registered(name) and (
            self.task_index[name]["type"] == "group"
        ):
87
88
89
90
            return True
        return False

    def _name_is_python_task(self, name):
91
92
93
        if self._name_is_registered(name) and (
            self.task_index[name]["type"] == "python_task"
        ):
94
95
96
            return True
        return False

97
    def _config_is_task(self, config) -> bool:
98
99
100
101
        if ("task" in config) and isinstance(config["task"], str):
            return True
        return False

102
    def _config_is_group(self, config) -> bool:
103
104
105
106
        if ("task" in config) and isinstance(config["task"], list):
            return True
        return False

107
    def _config_is_python_task(self, config) -> bool:
108
109
110
111
112
        if "class" in config:
            return True
        return False

    def _get_yaml_path(self, name):
113
114
        if name not in self.task_index:
            raise ValueError
115
116
117
        return self.task_index[name]["yaml_path"]

    def _get_config(self, name):
118
119
        if name not in self.task_index:
            raise ValueError
120
121
122
123
124
125
126
        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):
127
128
        if self._name_is_task(name):
            raise ValueError
129
130
131
132
133
134
135
136
137
138
139
140
        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(
141
        self,
142
143
144
145
        name_or_config: Optional[Union[str, dict]] = None,
        parent_name: Optional[str] = None,
        update_config: Optional[dict] = None,
    ) -> Mapping:
146
        def load_task(config, task):
147
148
            print(f"loading {task}", self._name_is_registered(task))
            print(config)
149
            if "include" in config:
150
151
                config = {
                    **utils.load_yaml_config(
152
                        yaml_path=None,
153
154
                        yaml_config={"include": config.pop("include")},
                        mode="full",
155
156
157
                    ),
                    **config,
                }
158
159
160
161
162
163
164
165
166
167
168
169
            if self._config_is_python_task(config):
                task_object = config["class"]()
            else:
                task_object = ConfigurableTask(config=config)
            return {task: task_object}

        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}
            elif self._name_is_task(name_or_config):
                task_config = self._get_config(name_or_config)
170
                return load_task(task_config, task=name_or_config)
171
            else:
172
173
174
175
                subtask_list = self._get_tasklist(name_or_config)
                if subtask_list == -1:
                    group_config = self._get_config(name_or_config)
                    subtask_list = group_config["task"]
lintangsutawika's avatar
lintangsutawika committed
176
177
178
                    group_name = ConfigurableGroup(config=group_config)
                else:
                    group_name = name_or_config
179

180
181
        if isinstance(name_or_config, dict):
            if update_config is not None:
182
                name_or_config = {
183
184
                    **name_or_config,
                    **update_config,
185
186
                }

187
188
189
190
            if self._config_is_task(name_or_config):
                name = name_or_config["task"]
                # If the name is registered as a group
                # if self._name_is_task(name) is False:
191
                if self._name_is_group(name) or self._name_is_tag(name):
192
                    group_name = name
193
194
195
                    update_config = {
                        k: v for k, v in name_or_config.items() if k != "task"
                    }
196
197
                    subtask_list = self._get_tasklist(name)
                    if subtask_list == -1:
198
199
                        group_config = self._get_config(name)
                        subtask_list = group_config["task"]
200
201
202
                        group_name = ConfigurableGroup(config=group_config)
                    else:
                        group_name = name
203
204
205
206
207
208
209
                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:
                            name_or_config["group"] = parent_name
210
211
212
213
214
215
216
217
                            num_duplicate = len(
                                list(
                                    filter(
                                        lambda x: x.startswith(name),
                                        self.task_group_map[parent_name],
                                    )
                                )
                            )
218
219
220
221
                            if num_duplicate > 0:
                                name = f"{name}-{num_duplicate}"
                            self.task_group_map[parent_name].append(name)

222
223
224
225
                        task_config = {
                            **base_task_config,
                            **name_or_config,
                        }
226
227
                    else:
                        task_config = name_or_config
228
                    return load_task(task_config, task=name)
229
230
231
            else:
                group_name = name_or_config["group"]
                subtask_list = name_or_config["task"]
232
                if set(name_or_config.keys()) > {"task", "group"}:
233
                    update_config = {
234
235
                        k: v
                        for k, v in name_or_config.items()
236
                        if k not in GROUP_ONLY_KEYS
237
                    }
238
239
240
241
242
243
244
245
246
247
                else:
                    update_config = None

                group_config = {
                    k: v
                    for k, v in name_or_config.items()
                    if k in GROUP_ONLY_KEYS + ["task", "group"]
                }
                if bool(group_config):
                    group_name = ConfigurableGroup(config=group_config)
248

249
250
251
252
253
        fn = partial(
            self._load_individual_task_or_group,
            parent_name=group_name,
            update_config=update_config,
        )
254
255
256
        return {
            group_name: dict(collections.ChainMap(*map(fn, reversed(subtask_list))))
        }
257

258
    def load_task_or_group(self, task_list: Optional[Union[str, list]] = None) -> dict:
259
        """Loads a dictionary of task objects from a list
260

261
262
        :param task_list: Union[str, list] = None
            Single string or list of string of task names to be loaded
263

264
265
266
267
268
        :return
            Dictionary of task objects
        """
        if isinstance(task_list, str):
            task_list = [task_list]
269

270
        all_loaded_tasks = dict(
271
            collections.ChainMap(*map(self._load_individual_task_or_group, task_list))
272
273
274
275
276
277
278
        )
        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):
279
        """Creates a dictionary of tasks index with the following metadata,
lintangsutawika's avatar
lintangsutawika committed
280
        - `type`, that can be either `task`, `python_task`, `group` or `tags`.
281
282
            `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
283
284
            `group` are group configs. `tags` are labels that can be assigned
            to tasks to assist in sorting and calling tasks of certain themes.
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
        - `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
        """
300
301
        # TODO: remove group in next release
        print_info = True
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
        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",
318
319
320
321
322
                            "task": -1,  # This signals that
                            # we don't need to know
                            # the task list for indexing
                            # as it can be loaded
                            # when called.
323
324
                            "yaml_path": yaml_path,
                        }
lintangsutawika's avatar
lintangsutawika committed
325

326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
                        # # 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,
341
                        }
342

343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
                        # TODO: remove group in next release
                        for attr in ["tag", "group"]:
                            if attr in config:
                                if attr == "group" and print_info:
                                    self.logger.info(
                                        "`group` and `group_alias` will no longer be used in the next release of lm-eval. "
                                        "`tags` will be used to allow to call a collection of tasks just like `group`. "
                                        "`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

                                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)
368
369
370
371
                    else:
                        self.logger.debug(f"File {f} in {root} could not be loaded")

        return tasks_and_groups
lintangsutawika's avatar
lintangsutawika committed
372

373

374
375
376
377
378
379
380
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
381

382

lintangsutawika's avatar
lintangsutawika committed
383
def get_task_name_from_object(task_object):
384
385
    if hasattr(task_object, "config"):
        return task_object._config["task"]
lintangsutawika's avatar
lintangsutawika committed
386
387
388
389
390
391
392
393
394

    # 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__
    )

395
396

def get_task_dict(
397
    task_name_list: Union[str, List[Union[str, Dict, Task]]],
398
    task_manager: Optional[TaskManager] = None,
399
):
400
    """Creates a dictionary of task objects from either a name of task, config, or prepared Task object.
lintangsutawika's avatar
lintangsutawika committed
401

402
403
404
405
406
407
408
    :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`
409

410
411
412
413
    :return
        Dictionary of task objects
    """
    task_name_from_string_dict = {}
414
415
416
    task_name_from_config_dict = {}
    task_name_from_object_dict = {}

417
    if isinstance(task_name_list, str):
lintangsutawika's avatar
lintangsutawika committed
418
        task_name_list = [task_name_list]
419
420
421
422
423
424
425
426
427
    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
428

429
430
431
432
433
    string_task_name_list = [task for task in task_name_list if isinstance(task, str)]
    others_task_name_list = [task for task in task_name_list if ~isinstance(task, str)]
    if len(string_task_name_list) > 0:
        if task_manager is None:
            task_manager = TaskManager()
lintangsutawika's avatar
lintangsutawika committed
434

435
436
437
        task_name_from_string_dict = task_manager.load_task_or_group(
            string_task_name_list
        )
438

439
440
    for task_element in others_task_name_list:
        if isinstance(task_element, dict):
441
442
            task_name_from_config_dict = {
                **task_name_from_config_dict,
443
                **task_manager.load_config(config=task_element),
444
445
446
447
448
            }

        elif isinstance(task_element, Task):
            task_name_from_object_dict = {
                **task_name_from_object_dict,
lintangsutawika's avatar
lintangsutawika committed
449
                get_task_name_from_object(task_element): task_element,
450
            }
lintangsutawika's avatar
lintangsutawika committed
451

452
    if not set(task_name_from_string_dict.keys()).isdisjoint(
lintangsutawika's avatar
lintangsutawika committed
453
        set(task_name_from_object_dict.keys())
454
455
    ):
        raise ValueError
456

lintangsutawika's avatar
lintangsutawika committed
457
    return {
458
        **task_name_from_string_dict,
lintangsutawika's avatar
lintangsutawika committed
459
        **task_name_from_config_dict,
460
        **task_name_from_object_dict,
lintangsutawika's avatar
lintangsutawika committed
461
    }