Autotrain compatible Custom code Endpoints compatible Instruct Phi-msft Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/ArtifactAI/phi-2-arxiv-math-instruct
Phi 2 Arxiv Math Instruct Benchmarks
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)
Phi 2 Arxiv Math Instruct Parameters and Internals
| LLM Name | Phi 2 Arxiv Math Instruct |
| Repository 🤗 | https://huggingface.co/ArtifactAI/phi-2-arxiv-math-instruct |
| Model Size | 2.8b |
| Required VRAM | 5.6 GB |
| Updated | 2024-08-14 |
| Maintainer | ArtifactAI |
| Model Type | phi-msft |
| Instruction-Based | Yes |
| Model Files | |
| Model Architecture | PhiForCausalLM |
| Model Max Length | 2048 |
| Transformers Version | 4.37.0.dev0 |
| Tokenizer Class | CodeGenTokenizer |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 51200 |
| Torch Data Type | float16 |
| Activation Function | gelu_new |
Best Alternatives to Phi 2 Arxiv Math Instruct
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Phi 2 Instruct Apo | 2K / 5.6 GB | 6 | 0 |
| Phi 2 Instruct V0.1 | 2K / 5.6 GB | 35 | 2 |
| EEVE Korean Instruct 2.8B V1.0 | 2K / 5.7 GB | 2035 | 28 |
| Att Model | 2K / 5.7 GB | 5 | 0 |
| Eeve2.8 Base | 2K / 5.7 GB | 5 | 0 |
| YanoljaNEXT EEVE Instruct 2.8B | 2K / 5.7 GB | 1790 | 29 |
| Eeve2.8 Ko | 2K / 5.7 GB | 5 | 0 |
| ... Instruct 2.8B V1.0 20240430 2 | 2K / 2.9 GB | 7 | 0 |
| Phi 2 Code Instruct | 2K / 5.6 GB | 6 | 4 |
| Dolphin 2 6 Phi 2 | 0K / 5.6 GB | 1200 | 191 |
Note: green Score (e.g. "73.2") means that the model is better than ArtifactAI/phi-2-arxiv-math-instruct .
Rank the Phi 2 Arxiv Math Instruct 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? 51538 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124