OpenELM 450M by apple

 ยป  All LLMs  ยป  apple  ยป  OpenELM 450M   URL Share it on

  Arxiv:2404.14619   Autotrain compatible   Custom code   Openelm   Region:us   Safetensors
Model Card on HF ๐Ÿค—: https://huggingface.co/apple/OpenELM-450M 

OpenELM 450M Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
OpenELM 450M (apple/OpenELM-450M)

OpenELM 450M Parameters and Internals

Model Type 
Language Model
Additional Notes 
The models were trained on public datasets with no safety guarantees, necessitating user-end safety testing and filtering.
Training Details 
Data Sources:
RefinedWeb, deduplicated PILE, subset of RedPajama, subset of Dolma v1.6
Data Volume:
approximately 1.8 trillion tokens
Methodology:
layer-wise scaling strategy
LLM NameOpenELM 450M
Repository ๐Ÿค—https://huggingface.co/apple/OpenELM-450M 
Model Size450m
Required VRAM1.8 GB
Updated2025-02-05
Maintainerapple
Model Typeopenelm
Model Files  1.8 GB
Model ArchitectureOpenELMForCausalLM
Licenseother
Transformers Version4.39.3
Vocabulary Size32000
Torch Data Typefloat32

Best Alternatives to OpenELM 450M

Best Alternatives
Context / RAM
Downloads
Likes
OpenELM 450M Instruct0K / 0.9 GB6983246
...450M Instruct Databaes SFT DPO0K / 0.9 GB1320
...Databaes SFT DPO 4bitquantized0K / 0.3 GB1040
DPO Reference Model0K / 1.8 GB1070
Note: green Score (e.g. "73.2") means that the model is better than apple/OpenELM-450M.

Rank the OpenELM 450M Capabilities

๐Ÿ†˜ 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! ๐ŸŒŸ

Instruction Following and Task Automation  
Factuality and Completeness of Knowledge  
Censorship and Alignment  
Data Analysis and Insight Generation  
Text Generation  
Text Summarization and Feature Extraction  
Code Generation  
Multi-Language Support and Translation  

What open-source LLMs or SLMs are you in search of? 42577 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227