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
|
{
lib,
buildPythonPackage,
fetchFromGitHub,
# build-system
setuptools-scm,
# dependencies
accelerate,
datasets,
dill,
evaluate,
jsonlines,
more-itertools,
numexpr,
peft,
pybind11,
pytablewriter,
rouge-score,
sacrebleu,
scikit-learn,
sqlitedict,
torch,
tqdm-multiprocess,
transformers,
word2number,
zstandard,
# optional-dependencies
# api
aiohttp,
requests,
tenacity,
tiktoken,
tqdm,
# hf_transfer
hf-transfer,
# ifeval
immutabledict,
langdetect,
nltk,
# neuronx
optimum,
# mamba
causal-conv1d,
mamba-ssm,
# math
antlr4-python3-runtime,
sympy,
# sentencepiece
sentencepiece,
# vllm
vllm,
# wandb
numpy,
pandas,
wandb,
# tests
pytestCheckHook,
writableTmpDirAsHomeHook,
}:
buildPythonPackage rec {
pname = "lm-eval";
version = "0.4.9.1";
pyproject = true;
src = fetchFromGitHub {
owner = "EleutherAI";
repo = "lm-evaluation-harness";
tag = "v${version}";
hash = "sha256-N5NRRabjWxPchwOIkjqYTCKInCmVSY6T5cAmdxNbCkU=";
};
build-system = [
setuptools-scm
];
dependencies = [
accelerate
datasets
dill
evaluate
jsonlines
more-itertools
numexpr
peft
pybind11
pytablewriter
rouge-score
sacrebleu
scikit-learn
sqlitedict
torch
tqdm-multiprocess
transformers
word2number
zstandard
];
optional-dependencies = {
api = [
aiohttp
requests
tenacity
tiktoken
tqdm
];
hf_transfer = [ hf-transfer ];
ifeval = [
immutabledict
langdetect
nltk
];
neuronx = [ optimum ] ++ optimum.optional-dependencies.neuronx;
mamba = [
causal-conv1d
mamba-ssm
];
math = [
antlr4-python3-runtime
sympy
];
optimum = [ optimum ] ++ optimum.optional-dependencies.openvino;
sentencepiece = [ sentencepiece ];
vllm = [ vllm ];
wandb = [
numpy
pandas
wandb
];
# Still missing dependencies for the following:
# deepsparse, gptq, ibm_watsonx_ai, multilingual, promptsource, sparseml,
# zeno, gptqmodel, japanese_leaderboard; all = [...];
};
pythonRelaxDeps = [ "datasets" ];
pythonImportsCheck = [ "lm_eval" ];
nativeCheckInputs = [
pytestCheckHook
writableTmpDirAsHomeHook
]
++ optional-dependencies.api;
disabledTests = [
"test_deepsparse" # deepsparse is not available
# download models from the internet
"test_get_batched_requests_with_no_ssl"
"test_model_tokenized_call_usage"
];
disabledTestPaths = [
# attempts to download models
"tests/models/test_huggingface.py"
"tests/test_evaluator.py"
"tests/test_include_path.py"
"tests/test_prompt.py"
"tests/test_task_manager.py"
"tests/test_tasks.py"
"tests/test_unitxt_tasks.py"
# optimum-intel is not available
"tests/models/test_openvino.py"
# zeno-client is not packaged
"tests/scripts/test_zeno_visualize.py"
];
meta = {
changelog = "https://github.com/EleutherAI/lm-evaluation-harness/releases/tag/${src.tag}";
description = "Framework for few-shot evaluation of language models";
homepage = "https://github.com/EleutherAI/lm-evaluation-harness";
license = [ lib.licenses.mit ];
maintainers = [ lib.maintainers.booxter ];
};
}
|
