Einstein V4 Phi2 is an open-source language model by Weyaxi. Features: 2.8b LLM, VRAM: 5.6GB, Context: 2K, License: other, HF Score: 60.8, LLM Explorer Score: 0.15, Arc: 60, HellaSwag: 74.1, MMLU: 56.9, TruthfulQA: 45.8, WinoGrande: 73.9, GSM8K: 54.
Axolotl Base model:finetune:microsoft/... Base model:microsoft/phi-2 Biology Chatml Chemistry Conversational Custom code Dataset:allenai/ai2 arc Dataset:bigbio/med qa Dataset:camel-ai/biology Dataset:camel-ai/chemistry Dataset:camel-ai/math Dataset:camel-ai/physics Dataset:cot-alpaca-gpt4-from-o... Dataset:derek-thomas/scienceqa Dataset:glaiveai/glaive-code-a... Dataset:jondurbin/airoboros-3.... Dataset:knowrohit07/saraswati-... Dataset:ldjnr/capybara Dataset:lmsys/lmsys-chat-1m Dataset:mandyyyyii/scibench Dataset:meta-math/metamathqa-4... Dataset:metaeval/reclor Dataset:migtissera/synthia-v1.... Dataset:open-orca/slimorca Dataset:openbookqa Dataset:piqa Dataset:sablo/oasst2 curated Dataset:scibench Dataset:sciq Dataset:stem-ai-mtl/electrical... Dataset:tiger-lab/mathinstruct Dataset:tiger-lab/scienceeval Einstein En Endpoints compatible Finetuned Generated from trainer Gpt4 Instruct Math Model-index Phi Phi2 Physics Region:us Safetensors Science Sharded Synthetic data Tensorflow
Model Card on HF 🤗: https://huggingface.co/Weyaxi/Einstein-v4-phi2
Einstein V4 Phi2 Benchmarks
ARC: 59.98 vs 96.7 (so35)-38%
HellaSwag: 74.07 vs 95.3 (gpt4)-22.3%
MMLU: 56.89 vs 88.3 (so35)-35.6%
TruthfulQA: 45.8 vs 59 (gpt4)-22.4%
WinoGrande: 73.88 vs 87.5 (gpt4)-15.6%
GSM8K: 53.98 vs 96.4 (so35)-44%
LLME Score: 0.14775
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)
Einstein V4 Phi2 Parameters and Internals
| Model Type |
| |||||||||
| Additional Notes |
| |||||||||
| Training Details |
|
| LLM Name | Einstein V4 Phi2 |
| Repository 🤗 | https://huggingface.co/Weyaxi/Einstein-v4-phi2 |
| Base Model(s) | |
| Model Size | 2.8b |
| Required VRAM | 5.6 GB |
| Updated | 2026-04-05 |
| Maintainer | Weyaxi |
| Model Type | phi |
| Model Files | |
| Supported Languages | en |
| Model Architecture | PhiForCausalLM |
| License | other |
| Context Length | 2048 |
| Model Max Length | 2048 |
| Transformers Version | 4.38.2 |
| Tokenizer Class | CodeGenTokenizer |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 51200 |
| Torch Data Type | bfloat16 |
Best Alternatives to Einstein V4 Phi2
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| MFANN3bv0.24 | 128K / 11.1 GB | 5 | 0 |
| MFANN3b | 128K / 11.1 GB | 87 | 0 |
| MFANN3bv1.4 | 128K / 11.1 GB | 6 | 0 |
| MFANN Phigments Slerp V3.2 | 128K / 5.6 GB | 5 | 0 |
| MFANN3bv1.3 | 128K / 11.1 GB | 11 | 0 |
| MFANN3bv0.23 | 128K / 11.1 GB | 5 | 0 |
| MFANN3bv1.1 | 128K / 11.1 GB | 5 | 0 |
| MFANN3bv1.5 | 128K / 11.1 GB | 5 | 0 |
| MFANN3b Rebase | 128K / 11.1 GB | 6 | 0 |
| MFANN Liminerity Slerp 4a | 128K / 5.6 GB | 5 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than Weyaxi/Einstein-v4-phi2 .
Rank the Einstein V4 Phi2 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? 52721 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a