[whatmodel.sh](../../README.md) / Pythia 1.4B
# Pythia 1.4B
EleutherAI · released 2023-02 · 1.41B params · apache-2.0
**score 13.8** ±1.4 · #713 of 732 · from 15 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 Pythia 1.4B among every model on that task or test, worst to best.
other tests (13)
test result vs the others in its run difficulty counts why
ARC Easy (maker) 60.6% ▃·▅▃▃▅·█●▃██ beats 11/19 easy · avg 87% 16% blunt · at its level · several runs
ARC Challenge (maker) 26.0% █··●·▁▁·▁▂▁▂ beats 10/19 easy · avg 74% 16% sharp · too hard for it · small run · s…
ImageNetVC visual commonsense, te… 42.2% ▂·▄·▂●▄▄█▄▂▆ beats 7/21 easy · avg 74% 11% blunt · at its level
WinoGrande (maker) 57.3% ▄····▄▆▂●▆▆█ beats 9/19 easy · avg 87% 10% typical
SciQ (maker) 86.5% ▂··▂▃▂▂▆●▃█▂ beats 11/19 very easy · avg 98% 9.5% blunt
PIQA (maker) 71.1% ▂·▄▄·▄▂●▂▄█▄ beats 10/19 easy · avg 87% 8.2% blunt · at its level
LAMBADA (OpenAI version), last-wo… 61.6% ▃··▅▃▃▅▅▅●██ beats 11/19 very easy · avg 87% 6.7% blunt · at its level
OpenBookQA (maker) 19.8% ●▁▁▁▂▂▃█▃▁▃▁ beats 0/107 very hard · avg 29% 6.4% too hard for it · small run
HellaSwag (maker) 40.2% ▁▁▁●▂▂▄▆█▇▄▂ beats 13/135 easy · avg 83% 5.7% small run
TriviaQA (maker) 4.8% ▁▁●▂▄▄▃▆▅█▅▂ beats 4/80 medium · avg 61% 5.5% small run
MedQA (USMLE, 4 options) 27.5% ▁▁··●▁▁··▄▃█ beats 2/17 medium · avg 71% 1.9% sharp · too hard for it
MedMCQA 31.5% ▂▂··▄●▂▂·▄██ beats 4/17 medium · avg 63% 1.8% too hard for it
PubMedQA 59.4% ▂·▂●·▄▂▄▄▂█▂ beats 3/17 medium · avg 75% 0.9% blunt · at its level
source results share of the score
Pythia: A Suite for Analyzing Large Lan… 6 54% ▇▇▇▇▇▇▇
trlX: A Framework for Large Scale Reinf… 5 30% ▇▇▇▇
ImageNetVC: Zero- and Few-Shot Visual C… 1 11% ▇
Open Medical-LLM Leaderboard results 3 5% ▇
prior: pull toward the middle 1% ▇
weighted average = 13.8
Licences: 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)) · trlX: A Framework for Large Scale Reinforcement Learning from Human Feedback ([CC BY 4.0](../../licenses/acl-2023-emnlp-main-530-CC-BY-4-0.txt)) · ImageNetVC: Zero- and Few-Shot Visual Commonsense Evaluation on 1000 ImageNet Categories ([CC BY 4.0](../../licenses/acl-2023-findings-emnlp-133-CC-BY-4-0.txt)) · Open Medical-LLM Leaderboard results ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
## 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 Pythia 1.4B (Q4_K_M, needs ~0.9 GB, from [huggingface.co/mradermacher](https://huggingface.co/mradermacher/pythia-1.4b-GGUF/tree/24c447e001d1db62d2af39a31084b71de854b24d))
```sh
llama cli -hf mradermacher/pythia-1.4b-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf mradermacher/pythia-1.4b-GGUF:Q4_K_M
```
Other versions: Q8_0 (~1.5 GB) `llama cli -hf mradermacher/pythia-1.4b-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 Pythia 1.4B (Q4_K_M, needs ~0.9 GB, from [huggingface.co/mradermacher](https://huggingface.co/mradermacher/pythia-1.4b-GGUF/tree/24c447e001d1db62d2af39a31084b71de854b24d))
```sh
lms get https://huggingface.co/mradermacher/pythia-1.4b-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 (~1.5 GB) `lms get https://huggingface.co/mradermacher/pythia-1.4b-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 Pythia 1.4B (Q4_K_M, needs ~0.9 GB, from [huggingface.co/mradermacher](https://huggingface.co/mradermacher/pythia-1.4b-GGUF/tree/24c447e001d1db62d2af39a31084b71de854b24d))
```sh
ollama run hf.co/mradermacher/pythia-1.4b-GGUF:Q4_K_M
```
Other versions: Q8_0 (~1.5 GB) `ollama run hf.co/mradermacher/pythia-1.4b-GGUF:Q8_0`
### Other ways to run it
1. Run it as published (fp16, needs ~2.93 GB, from [huggingface.co/EleutherAI](https://huggingface.co/EleutherAI/pythia-1.4b/tree/fedc38a16eea3bd36a96b906d78d11d2ce18ed79))
```sh
uv run --with torch --with 'transformers<5' python -c "from transformers import pipeline; print(pipeline('text-generation', model='EleutherAI/pythia-1.4b')('Once upon a time', max_new_tokens=40)[0]['generated_text'])"
```
Model repository and documentation: [huggingface.co/EleutherAI](https://huggingface.co/EleutherAI/pythia-1.4b)
[interactive version](https://whatmodel.sh/model/pythia-1-4b) · [JSON](https://whatmodel.sh/api?format=json&all=1)