[whatmodel.sh](../../README.md) / BLOOM 1.7B
# BLOOM 1.7B
BigScience · released 2022-07 · 1.72B params · bigscience-bloom-rail-1.0
**score 10.7** ±1.0 · #717 of 732 · from 37 benchmark results across 7 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks BLOOM 1.7B among every model on that task or test, worst to best.
task vs the others rank counts tests
coding █··█·●·██··█ #4 of 6* 4.1% 1
math ▁●▃▅█▆▄█▄▂▁▁ #421 of 431* 2.0% 1
coding · #4 of 6 · 4.1% of its score
test result vs the others in its run difficulty counts why
HumanEval (maker) 4.0% ▄···▄·●·█··▄ beats 2/5 easy · avg 79% 4.1% sharp · too hard for it
math · #421 of 431 · 2.0% of its score
Beats [Flan-T5 Large](flan-t5-large.md), [OPT-66B](opt-66b.md), [Flan-T5 Base](flan-t5-base.md) here, though they rank higher overall. Trails [GPT-Neo 125M](gpt-neo-125m.md), [OPT-125M](opt-125m.md), [GPT-2 XL (1.5B)](gpt-2-xl.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
GSM8K 1.3% ▄▄●▄···▄·▄·▄ beats 3/7 easy · avg 65% 2.0% sharp · too hard for it
other tests (22)
test result vs the others in its run difficulty counts why
ARC Easy 56.2% ▃·▅▃▃▅·●█▃██ beats 7/19 easy · avg 87% 17% blunt · at its level · several runs
PIQA 68.6% ▂·▄▄·▄●▆▂▄█▄ beats 7/19 easy · avg 87% 11% blunt · at its level · several runs
ARC Challenge 23.8% ●··▁·▁▁·▁▂▁▂ beats 3/19 easy · avg 81% 10% sharp · too hard for it · several runs
WinoGrande 57.2% ▄····▄▆▂●▆▆█ beats 8/19 easy · avg 88% 10% several runs
LAMBADA (OpenAI version), last-wo… 46.2% ▃··▅▃▃●▅▅███ beats 6/19 very easy · avg 87% 6.9% blunt · at its level · several runs
SciQ 85.1% ▂··▂▃▂▂●▃▃█▂ beats 9/19 very easy · avg 98% 5.5% blunt
TriviaQA 12.7% ▁··▁█●▃▁▄▅▄█ beats 8/28 medium · avg 45% 5.0% too hard for it · several runs
Story Cloze Test 65.4% ▂▂▄●·▂·▆·▂█▂ beats 4/14 easy · avg 89% 4.7% blunt · at its level · several runs
COPA (maker) 54.0% ▃··●▅▃▃▅·▃▃▃ beats 1/12 very easy · avg 95% 4.7% at its level · several runs
HellaSwag 37.6% ▄·▄●·▄·▄··▄█ beats 2/7 easy · avg 81% 4.5% typical
CoQA 53.5% ▄····▄●·▄▄·█ beats 2/6 very easy · avg 92% 3.3% at its level
Winograd Schema Challenge (WSC273) 72.9% ▄▄·●·▄··█··█ beats 2/7 very easy · avg 93% 2.8% blunt · at its level
Natural Questions (closed book) 3.5% ▁·▁·▁●▁▂·▅▂█ beats 3/28 medium · avg 48% 2.2% too hard for it
BoolQ 61.8% ▄··▄·●▄··█·▄ beats 3/7 easy · avg 87% 1.7% typical
RACE-high 33.2% ▄···●··▄▄··█ beats 2/6 medium · avg 51% 1.3% blunt
DROP 6.9% ▄●···█·▄·▄·▄ beats 1/6 hard · avg 34% 1.2% blunt
OpenBookQA 21.4% ●········▂▂▂ beats 2/7 hard · avg 40% 0.8% too hard for it
TruthfulQA MC2 41.3% ▅▃●▆▁▂▅···▁▁ beats 11/31 medium · avg 57% 0.5% blunt · at its level · several runs
Open PT-LLM Leaderboard average (… 28.6% ●▁▁▁▁▁▁▁▂█▅▁ beats 29/276 medium · avg 72% 0.4% too hard for it · noisy
WebQuestions (closed book) 1.3% █···▄··●▄▄·▄ beats 3/6 very hard · avg 3.5% 0.1% blunt · too hard for it
HaluEval QA hallucination detecti… 49.5% ●▁▁·▁▁·▁▁▁▁▁ beats 10/27 hard · avg 66% 0.1% too hard for it · noisy
LogiQA (harness) 21.7% ●··········▁ beats 11/14 very hard · avg 25% 0.0% too hard for it · noisy
source results share of the score
EleutherAI lm-evaluation-harness result… 17 43% ▇▇▇▇▇
Pythia: A Suite for Analyzing Large Lan… 6 27% ▇▇▇
GPTQ: Accurate Post-Training Quantizati… 5 21% ▇▇
BLOOM: A 176B-Parameter Open-Access Mul… 1 4% ▇
Hallucinations Leaderboard results 4 3% ▇
Crosslingual Generalization through Mul… 3 1% ▇
Open Portuguese LLM Leaderboard 1 0% ▇
prior: pull toward the middle 0% ▇
weighted average = 10.7
Licences: EleutherAI lm-evaluation-harness results folder ([MIT](../../licenses/lm-eval-harness-results-2023-05-MIT.txt)) · Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling ([CC BY-SA 4.0](../../licenses/paper-2304.01373v2-CC-BY-SA-4.0.txt)) · GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers ([CC BY 4.0](../../licenses/paper-2210.17323v2-gptq-CC-BY-4-0.txt)) · BLOOM: A 176B-Parameter Open-Access Multilingual Language Model ([CC BY 4.0](../../licenses/paper-2211.05100v4-CC-BY-4.0.txt)) · Hallucinations Leaderboard results ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · Crosslingual Generalization through Multitask Finetuning ([CC BY 4.0](../../licenses/acl-2023-acl-long-891-CC-BY-4-0.txt)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt))
## How to run it
### llama.cpp (recommended)
1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md)
```sh
# macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
# Windows
irm https://llama.app/install.ps1 | iex
```
2. Run BLOOM 1.7B (Q4_K_M, needs ~1.5 GB, from [huggingface.co/mradermacher](https://huggingface.co/mradermacher/bloom-1b7-GGUF/tree/44b567abb1107979da1de4b66b736e47ef75ff49))
```sh
llama cli -hf mradermacher/bloom-1b7-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf mradermacher/bloom-1b7-GGUF:Q4_K_M
```
Other versions: Q8_0 (~2.4 GB) `llama cli -hf mradermacher/bloom-1b7-GGUF:Q8_0`
### LM Studio
1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download)
2. Run BLOOM 1.7B (Q4_K_M, needs ~1.5 GB, from [huggingface.co/mradermacher](https://huggingface.co/mradermacher/bloom-1b7-GGUF/tree/44b567abb1107979da1de4b66b736e47ef75ff49))
```sh
lms get https://huggingface.co/mradermacher/bloom-1b7-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q8_0 (~2.4 GB) `lms get https://huggingface.co/mradermacher/bloom-1b7-GGUF@Q8_0 -y`
### Ollama
1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download)
```sh
# Linux
curl -fsSL https://ollama.com/install.sh | sh
```
2. Run BLOOM 1.7B (Q4_K_M, needs ~1.5 GB, from [huggingface.co/mradermacher](https://huggingface.co/mradermacher/bloom-1b7-GGUF/tree/44b567abb1107979da1de4b66b736e47ef75ff49))
```sh
ollama run hf.co/mradermacher/bloom-1b7-GGUF:Q4_K_M
```
Other versions: Q8_0 (~2.4 GB) `ollama run hf.co/mradermacher/bloom-1b7-GGUF:Q8_0`
### Other ways to run it
1. Run it as published (fp16, needs ~3.44 GB, from [huggingface.co/bigscience](https://huggingface.co/bigscience/bloom-1b7/tree/cc72a88036c2fb937d65efeacc57a0c2ef5d6fe5))
```sh
uv run --with torch --with 'transformers<5' python -c "from transformers import pipeline; print(pipeline('text-generation', model='bigscience/bloom-1b7')('Once upon a time', max_new_tokens=40)[0]['generated_text'])"
```
Model repository and documentation: [huggingface.co/bigscience](https://huggingface.co/bigscience/bloom-1b7)
[interactive version](https://whatmodel.sh/model/bloom-1b7) · [JSON](https://whatmodel.sh/api?format=json&all=1)