Jdudeo commited on
Commit
85bf632
·
verified ·
1 Parent(s): 3f17e82

Upload 3 files

Browse files
Files changed (3) hide show
  1. README.md +82 -0
  2. eval_run_metadata.json +102 -0
  3. model_scores_2026-08-27.json +0 -0
README.md ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc0-1.0
3
+ language:
4
+ - en
5
+ tags:
6
+ - benchmark-results
7
+ - language-model-evaluation
8
+ - small-language-models
9
+ ---
10
+
11
+ ![Veyra Banner](https://cdn-uploads.huggingface.co/production/uploads/6857f2cfae68b377f17aff8c/MfZTOclKIpOM1RbUlnfcr.jpeg)
12
+
13
+ # SLM Bench
14
+
15
+ SLM Bench is a collection of benchmark results that have been collected using [lm-evaluation-harness](https://github.com/eleutherai/lm-evaluation-harness)
16
+
17
+ ### Models Recorded
18
+
19
+ 2026-08-27:
20
+ - 154 model records across 49 organizations
21
+ - 125 records marked isValid: true, all other models haven't been evaluated because of gated access or lm-eval not being able to benchmark them.
22
+
23
+ ### What benchmarks are included:
24
+
25
+ - ARC-Easy: arc_easy_acc, arc_easy_acc_norm
26
+ - ARC-Challenge: arc_challenge_acc, arc_challenge_acc_norm
27
+ - HellaSwag: hellaswag_acc, hellaswag_acc_norm
28
+ - PIQA: piqa_acc, piqa_acc_norm
29
+ - SciQ: sciq_acc, sciq_acc_norm
30
+ - WinoGrande: winogrande_acc
31
+ - OpenBookQA: openbookqa_acc, openbookqa_acc_norm
32
+ - BoolQ: boolq_acc
33
+ - LAMBADA: lambada_acc
34
+ - LogiQA: logiqa_acc
35
+ - MathQA: mathqa_acc
36
+
37
+ ## Uses
38
+
39
+ This dataset has been made for people to be able to compare models under 150M against eachother and provide it in a fully open way without any restrictions. You can make leaderboards, visualisations, etc.
40
+
41
+ ## Dataset Structure
42
+
43
+ The JSON file contains two top-level objects:
44
+
45
+ <pre><code>{
46
+ "orgs": {
47
+ "exampleorg": {
48
+ "name": "Example Org",
49
+ "url": "https://huggingface.co/example-org"
50
+ }
51
+ },
52
+ "models": [
53
+ {
54
+ "name": "Example-10M-Base",
55
+ "org": "exampleorg",
56
+ "params": 10000000,
57
+ "links": {
58
+ "card": "https://huggingface.co/example-org/Example-10M-Base"
59
+ },
60
+ "isCustom": false,
61
+ "isLooped": false,
62
+ "loopCount": 1,
63
+ "firstCommit": "2026-08-27",
64
+ "arc_easy_acc": 0.0,
65
+ "arc_easy_acc_norm": 0.0,
66
+ "isValid": true
67
+ }
68
+ ]
69
+ }
70
+ </code></pre>
71
+
72
+ ### License
73
+
74
+ The results dataset is dedicated to the public domain under CC0 1.0. Individual models retain their own licenses and terms.
75
+
76
+ ### Attribution
77
+
78
+ No attribution is required but if you are using SLMBench for any public project we would appreciate a citation or link:
79
+
80
+ > https://huggingface.co/datasets/veyra-ai/SLMBench-Results
81
+
82
+ Attribution does not imply endorsement by Veyra AI or the model authors represented in this dataset.
eval_run_metadata.json ADDED
@@ -0,0 +1,102 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "created_utc": "2026-08-26T14:00:05.544555+00:00",
3
+ "input_file": "model_no_scores.json",
4
+ "input_sha256": "0e277777ae79d4b87512b061117ca5664794baf5d02af5ceac23a7d2e2acd18b",
5
+ "lm_eval_git_commit": "ccde35c889a90c06353bbf0dc01ed320907f10f2",
6
+ "datasets_version": "3.6.0",
7
+ "tasks": [
8
+ "arc_easy",
9
+ "arc_challenge",
10
+ "hellaswag",
11
+ "piqa",
12
+ "sciq",
13
+ "winogrande",
14
+ "openbookqa",
15
+ "boolq",
16
+ "lambada_openai",
17
+ "logiqa",
18
+ "mathqa"
19
+ ],
20
+ "metric_map": {
21
+ "arc_easy_acc": [
22
+ "arc_easy",
23
+ "acc"
24
+ ],
25
+ "arc_easy_acc_norm": [
26
+ "arc_easy",
27
+ "acc_norm"
28
+ ],
29
+ "arc_challenge_acc": [
30
+ "arc_challenge",
31
+ "acc"
32
+ ],
33
+ "arc_challenge_acc_norm": [
34
+ "arc_challenge",
35
+ "acc_norm"
36
+ ],
37
+ "hellaswag_acc": [
38
+ "hellaswag",
39
+ "acc"
40
+ ],
41
+ "hellaswag_acc_norm": [
42
+ "hellaswag",
43
+ "acc_norm"
44
+ ],
45
+ "piqa_acc": [
46
+ "piqa",
47
+ "acc"
48
+ ],
49
+ "piqa_acc_norm": [
50
+ "piqa",
51
+ "acc_norm"
52
+ ],
53
+ "sciq_acc": [
54
+ "sciq",
55
+ "acc"
56
+ ],
57
+ "sciq_acc_norm": [
58
+ "sciq",
59
+ "acc_norm"
60
+ ],
61
+ "winogrande_acc": [
62
+ "winogrande",
63
+ "acc"
64
+ ],
65
+ "openbookqa_acc": [
66
+ "openbookqa",
67
+ "acc"
68
+ ],
69
+ "openbookqa_acc_norm": [
70
+ "openbookqa",
71
+ "acc_norm"
72
+ ],
73
+ "boolq_acc": [
74
+ "boolq",
75
+ "acc"
76
+ ],
77
+ "lambada_acc": [
78
+ "lambada_openai",
79
+ "acc"
80
+ ],
81
+ "logiqa_acc": [
82
+ "logiqa",
83
+ "acc"
84
+ ],
85
+ "mathqa_acc": [
86
+ "mathqa",
87
+ "acc"
88
+ ]
89
+ },
90
+ "score_scale": 100.0,
91
+ "dtype": "float32",
92
+ "batch_size": "auto:4",
93
+ "device": "cuda:0",
94
+ "seeds": [
95
+ 0,
96
+ 1234,
97
+ 1234,
98
+ 1234
99
+ ],
100
+ "num_fewshot_override": null,
101
+ "full_dataset_limit": null
102
+ }
model_scores_2026-08-27.json ADDED
The diff for this file is too large to render. See raw diff