File size: 7,964 Bytes
c95c7b0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
import logging
import pytest
import numpy as np
from typing import Optional, Tuple
from itertools import islice
from promptsource.templates import Template

import lm_eval.tasks as tasks
from lm_eval.api.task import Task
from lm_eval.api.request import Request
from lm_eval.api.utils import set_seed, DEFAULT_SEED


logger = logging.getLogger(__name__)


def _get_deterministic_template(
    task_name: str,
) -> Tuple[Optional[Template], bool]:
    """Some `promptsource` templates randomize the ordering of prompt attributes
    If `task_class` does not have a prompt template with non-random ordering we
    return None.

    Returns:
        (prompt template, is_deterministic)
    """
    # Only choose 1 promptsource template.
    prompt_template = None
    templates = tasks.get_templates(task_name)
    if templates.all_template_names:
        for template_name in templates.all_template_names:
            prompt_template = templates[template_name]
            # Hacky way to ensure we only grab a deterministic jinja template.
            if (
                "range(" not in prompt_template.jinja
                and "random" not in prompt_template.jinja
            ):
                return prompt_template, True
        # Return the last non-deterministic template.
        return prompt_template, False
    return None, False


def _filter_docs(task: Task):
    def _filter(doc: dict):
        return not task.invalid_doc_for_prompt(doc)

    return _filter


@pytest.mark.parametrize("task_name,task_class", tasks.TASK_REGISTRY.items())
def test_basic_interface(task_name: str, task_class: Task):
    logger.info("Evaluating task", task_name)
    prompt_template, is_deterministic = _get_deterministic_template(task_name)
    task = task_class(prompt_template=prompt_template)

    assert task.has_training_docs() in [True, False]
    assert task.has_validation_docs() in [True, False]
    assert task.has_test_docs() in [True, False]

    assert isinstance(task.aggregation(), dict)
    assert isinstance(task.higher_is_better(), dict)
    assert task.aggregation().keys() == task.higher_is_better().keys()
    for v in task.higher_is_better().values():
        assert v in [True, False]
    assert isinstance(task.VERSION, int)

    # Test deterministic docs (NOTE: Don't test train because it's slow).
    # Return if the prompts are non-deterministic here.
    if not is_deterministic:
        return

    limit = 100
    task2 = task_class(prompt_template=prompt_template)

    if task.has_validation_docs():
        arr = list(task.validation_docs().filter(_filter_docs(task)))[:limit]
        arr2 = list(task2.validation_docs().filter(_filter_docs(task2)))[:limit]
        assert arr == arr2
        requests = [
            task.construct_requests(doc, task.doc_to_text(doc), {"num_fewshot": 0})
            for doc in arr
        ]
        requests2 = [
            task.construct_requests(doc, task2.doc_to_text(doc), {"num_fewshot": 0})
            for doc in arr2
        ]
        assert requests == requests2

    if task.has_test_docs():
        arr = list(task.test_docs().filter(_filter_docs(task)))[:limit]
        arr2 = list(task2.test_docs().filter(_filter_docs(task2)))[:limit]
        assert arr == arr2
        requests = [
            task.construct_requests(doc, task.doc_to_text(doc), {"num_fewshot": 0})
            for doc in arr
        ]
        requests2 = [
            task2.construct_requests(doc, task2.doc_to_text(doc), {"num_fewshot": 0})
            for doc in arr2
        ]
        assert requests == requests2


@pytest.mark.parametrize("task_name,task_class", tasks.TASK_REGISTRY.items())
def test_documents_and_requests(task_name: str, task_class: Task):
    set_seed()
    logger.info("Evaluating task", task_name)
    prompt_template, _ = _get_deterministic_template(task_name)
    task = task_class(prompt_template=prompt_template)

    fns = []

    # Training docs are too expensive to run on CI.
    # if task.has_training_docs():
    #     fns.append(task.training_docs)
    if task.has_validation_docs():
        fns.append(task.validation_docs)
    for fn in fns:
        docs = fn().filter(_filter_docs(task))
        for doc in islice(docs, 5):
            text = task.doc_to_text(doc)
            target = task.doc_to_target(doc)

            assert isinstance(text, str)
            assert isinstance(target, list)

            requests = task.construct_requests(doc, text, {"num_fewshot": 0})

            # Construct_requests can return just one request
            if not isinstance(requests, (list, tuple)):
                requests = [requests]
            # TODO: Mock lm after refactoring evaluator.py to not be a mess
            for req in requests:
                assert isinstance(req, Request)


def test_arg_string_task_creation():
    import itertools

    TEST_EXAMPLE_SEPS = [
        # Test `=` symbol in value string
        "\n===TEST_SEPARATOR===\n",
        # Test whitespace only separators
        " ",
        " \t\t  ",
        "\n\n\n\n",
        # Test empty string separator
        "",
        # Test misc. symbols in separator
        "[[[[]]]]",
        "<<___>>",
        "(())",
    ]
    TEST_TEXT_TARGET_SEPS = [
        # Test whitespace separators
        "   ",
        " \t ",
        "\n\n\n",
    ]

    # Ensure parsing properly handles args.
    for example_sep, text_target_sep in itertools.product(
        TEST_EXAMPLE_SEPS, TEST_TEXT_TARGET_SEPS
    ):
        test_arg_string = f" save_examples=False,example_separator={example_sep},text_target_separator={text_target_sep}"
        task = tasks.get_task_list_from_args_string(
            "wnli",
            template_names=["confident"],
            task_args=test_arg_string,
        )[0]

        assert task.save_examples is False
        assert task.example_separator == example_sep
        assert task.text_target_separator == text_target_sep

    # Ensure fewshot context is formatted as expected.
    TEST_EXAMPLE_SEP = "\n===TEST_SEPARATOR===\n"
    TEST_TEXT_TARGET_SEP = "   "
    test_arg_string = f" save_examples=False,example_separator={TEST_EXAMPLE_SEP},text_target_separator={TEST_TEXT_TARGET_SEP}"
    task = tasks.get_task_list_from_args_string(
        "wnli",
        template_names=["confident"],
        task_args=test_arg_string,
    )[0]
    context = task.fewshot_context(
        task.validation_docs()[0],
        num_fewshot=2,
        rng=np.random.default_rng(DEFAULT_SEED),
    )[0]
    expected = f"""If it's true that
The man couldn't lift his son because he was so heavy.
how confident should I be that
The man was so heavy.
very confident or not confident?   not confident
===TEST_SEPARATOR===
If it's true that
As Ollie carried Tommy up the long winding steps, his legs ached.
how confident should I be that
Ollie's legs ached.
very confident or not confident?   very confident
===TEST_SEPARATOR===
If it's true that
The drain is clogged with hair. It has to be cleaned.
how confident should I be that
The hair has to be cleaned.
very confident or not confident?"""
    assert context == expected

    # Ensure tasks don't instantiate with invalid args.
    with pytest.raises(AssertionError):
        bad_save_examples_arg_string = "example_separator=\t,save_examples=yes"
        task = tasks.get_task_list_from_args_string(
            "wnli",
            template_names=["confident"],
            task_args=bad_save_examples_arg_string,
        )[0]

        bad_example_sep_arg_string = "example_separator=False,save_examples=False"
        task = tasks.get_task_list_from_args_string(
            "wnli",
            template_names=["confident"],
            task_args=bad_example_sep_arg_string,
        )[0]

        bad_text_sep_arg_string = "text_target_separator=___"
        task = tasks.get_task_list_from_args_string(
            "wnli",
            template_names=["confident"],
            task_args=bad_text_sep_arg_string,
        )[0]