Arxiv:2101.00027 Arxiv:2201.07311 Arxiv:2304.01373 Autotrain compatible Dataset:eleutherai/pile En Endpoints compatible Gpt neox Pythia Pytorch Region:us Safetensors Sharded Tensorflow
Pythia 12B Benchmarks
MMLU Pro: 1.21
MUSR: 3.79
BBH: 4.99
IFEval: 24.71 vs 88 (so35)-71.9%
ARC: 39.59 vs 96.7 (so35)-59.1%
HellaSwag: 68.82 vs 95.3 (gpt4)-27.8%
MMLU: 26.76 vs 88.3 (so35)-69.7%
TruthfulQA: 31.85 vs 59 (gpt4)-46%
WinoGrande: 64.17 vs 87.5 (gpt4)-26.7%
GSM8K: 1.74 vs 96.4 (so35)-98.2%
MATH Lvl 5: 0.91
LLME Score: 0.18912
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Pythia 12B Parameters and Internals
| LLM Name | Pythia 12B |
| Repository 🤗 | https://huggingface.co/EleutherAI/pythia-12b |
| Model Size | 12b |
| Required VRAM | 23.8 GB |
| Updated | 2024-10-17 |
| Maintainer | EleutherAI |
| Model Type | gpt_neox |
| Model Files | |
| Supported Languages | en |
| Model Architecture | GPTNeoXForCausalLM |
| License | apache-2.0 |
| Context Length | 2048 |
| Model Max Length | 2048 |
| Transformers Version | 4.24.0 |
| Tokenizer Class | GPTNeoXTokenizer |
| Vocabulary Size | 50688 |
| Torch Data Type | float16 |
Best Alternatives to Pythia 12B
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Dolly V2 12B | 2K / 23.8 GB | 7146 | 1948 |
| ...sst Sft 4 Pythia 12B Epoch 3.5 | 2K / 23.8 GB | 477165 | 360 |
| Oasst Sft 1 Pythia 12B | 2K / 23.8 GB | 3789 | 279 |
| Pythia 12B Sft V8 7K Steps | 2K / 23.8 GB | 1877 | 21 |
| Pythia 12B Deduped | 2K / 23.8 GB | 6054 | 51 |
| Pythia 12B Pre V8.12.5K Steps | 2K / 23.8 GB | 1262 | 6 |
| ...ythia 12B Sft V8 Rlhf 2K Steps | 2K / 23.8 GB | 1257 | 0 |
| H2ogpt Gm Oasst1 En 1024 12B | 2K / 23.8 GB | 1240 | 5 |
| Pythia 12B Sft V8.2.5K Steps | 2K / 23.8 GB | 1196 | 0 |
| H2ogpt Oasst1 512 12B | 2K / 23.9 GB | 1282 | 27 |
Note: green Score (e.g. "73.2") means that the model is better than EleutherAI/pythia-12b .
Rank the Pythia 12B Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 36966 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v2024072803