[whatmodel.sh](../../README.md) / StableLM-Base-Alpha 7B # StableLM-Base-Alpha 7B Stability AI · released 2023-04 · 7.9B params · cc-by-sa-4.0 **score 8.2** ±1.4 · #719 of 732 · from 16 benchmark results across 4 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks StableLM-Base-Alpha 7B among every model on that task or test, worst to best. 
task    vs the others         rank  counts  tests
coding  ●▁▁▃▅▇█▇▅▄▃▁   #468 of 469     18%      2
coding · #468 of 469 · 18% of its score Beats [Zyte-1B](zyte-1b.md) here, though they rank higher overall. 
test        result  vs the others in its run  difficulty      counts  why
HumanEval     2.4%  ●·▁▁▃▄▆▅█▃▁▁ beats 1/99   easy · avg 76%     11%  sharp · too hard for it · several runs
HumanEval+    2.4%  ●·▁▁▁▃▄▄▄█▃▁ beats 1/99   easy · avg 70%    6.9%  sharp · too hard for it
other tests (13) 
test                                result  vs the others in its run  difficulty           counts  why
SciQ (maker)                         80.1%  ▁·●·▁·▃▅▃█▄▁ beats 1/26   very easy · avg 97%     17%  blunt
ARC Challenge (maker)                27.0%  ▁·●··▁▁▇▂█▆▃ beats 1/26   easy · avg 82%          12%  sharp · too hard for it
HellaSwag (maker)                    41.2%  ●···▁▇▅█▇▃▁▁ beats 1/26   very easy · avg 91%     10%  at its level
ARC Easy (maker)                     44.9%  ▁●·▁▁▄▅▄█▅▁▃ beats 1/26   easy · avg 85%         9.0%  blunt · at its level
BoolQ (maker)                        60.1%  ▂●▂▅▆▂▅▆█▂▂▂ beats 1/26   easy · avg 88%         8.0%  typical
WinoGrande (maker)                   50.1%  ●····▁··▁▂█▄ beats 0/26   easy · avg 86%         7.2%  typical
PIQA (maker)                         66.8%  ▁·●··▁▁▃█▅█▂ beats 1/26   easy · avg 87%         6.8%  blunt · at its level
LAMBADA (OpenAI version), last-wo…   55.1%  ▁··▂●▁▁▂▆█▅▃ beats 3/26   very easy · avg 86%    5.4%  blunt · at its level
Social IQA (maker)                   39.5%  ▁···●▃█▁▁▁·▁ beats 1/26   hard · avg 56%         2.0%  blunt · too hard for it
MMLU                                 26.2%  ▄●·▄···▄▄·██ beats 1/8    easy · avg 74%         1.6%  too hard for it
OpenBookQA (maker)                   21.4%  ●·▁▁▃·▁▃▅▇▃▄ beats 4/26   hard · avg 49%         1.1%  too hard for it
Open PT-LLM Leaderboard average (…   25.4%  ●▁▁▁▁▁▁▁▂█▅▁ beats 1/276  medium · avg 72%       0.7%  too hard for it · noisy
TruthfulQA MC2 (maker)               40.0%  █·█·█▄···▄·● beats 8/9    medium · avg 55%       0.6%  blunt · at its level
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
Stability AI (submitted / maker-reporte…       11   79% ▇▇▇▇▇▇▇▇▇
EvalPlus / EvalPerf published evaluatio…        2   16% ▇▇
InstructEval                                    2    3% ▇
Open Portuguese LLM Leaderboard                 1    1% ▇
prior: pull toward the middle                        1% ▇
weighted average                                   = 8.2
Licences: Stability AI ([Apache-2.0](../../licenses/stablelm-original-eval-json-Apache-2.0.txt)) · EvalPlus / EvalPerf published evaluation runs ([Apache-2.0](../../licenses/evalplus-published-runs-Apache-2.0.txt)) · InstructEval ([MIT / Apache-2.0](https://github.com/declare-lab/instruct-eval/blob/49bbd2ce7e3a5bdc4024ba62a5c9c68c107ee966/LICENSE)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) ## How to run it ### Other ways to run it 1. Run it as published (fp32, needs ~31.74 GB, from [huggingface.co/stabilityai](https://huggingface.co/stabilityai/stablelm-base-alpha-7b/tree/1d43a3b40d7d708d8084d07dfae0c4e9fd95c04f)) ```sh uv run --with torch --with 'transformers<5' --with accelerate python -c "from transformers import pipeline; print(pipeline('text-generation', model='stabilityai/stablelm-base-alpha-7b', torch_dtype='auto', device_map='auto')('Once upon a time', max_new_tokens=40)[0]['generated_text'])" ``` Model repository and documentation: [huggingface.co/stabilityai](https://huggingface.co/stabilityai/stablelm-base-alpha-7b) [interactive version](https://whatmodel.sh/model/stablelm-base-alpha-7b) · [JSON](https://whatmodel.sh/api?format=json&all=1)