Gemma 2 2B by google

 ยป  All LLMs  ยป  google  ยป  Gemma 2 2B   URL Share it on

  Arxiv:1705.03551   Arxiv:1804.06876   Arxiv:1804.09301   Arxiv:1809.02789   Arxiv:1811.00937   Arxiv:1903.00161   Arxiv:1904.09728   Arxiv:1905.07830   Arxiv:1905.10044   Arxiv:1907.10641   Arxiv:1911.01547   Arxiv:1911.11641   Arxiv:2009.03300   Arxiv:2009.11462   Arxiv:2101.11718   Arxiv:2103.03874   Arxiv:2107.03374   Arxiv:2108.07732   Arxiv:2109.07958   Arxiv:2110.08193   Arxiv:2110.14168   Arxiv:2203.09509   Arxiv:2206.04615   Arxiv:2304.06364   Arxiv:2403.13793   Autotrain compatible   Endpoints compatible   Gemma2   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF ๐Ÿค—: https://huggingface.co/google/gemma-2-2b 

Gemma 2 2B Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
๐ŸŒŸ Advertise your project ๐Ÿš€

Gemma 2 2B Parameters and Internals

Model Type 
text-to-text, decoder-only, large language model
Use Cases 
Areas:
Content Creation, Communication, Research, Education
Applications:
Text Generation, Chatbots, Text Summarization
Primary Use Cases:
Customer service, Virtual assistants, Interactive applications, NLP research
Limitations:
Bias in training data, Complex open-ended tasks, Language nuances
Considerations:
Continuous monitoring and content safety mechanisms encouraged
Additional Notes 
Offers open model with lightweight architecture for various text generation tasks.
Supported Languages 
English (Proficient)
Training Details 
Data Sources:
Web Documents, Code, Mathematics
Data Volume:
2 trillion tokens
Methodology:
Uses CSAM filtering and sensitive data filtering during preprocessing
Hardware Used:
TPUv5p
Model Architecture:
Open weights for both pre-trained and instruction-tuned variants
Safety Evaluation 
Methodologies:
Internal red-teaming testing, Structured evaluations
Findings:
Within acceptable thresholds for safety standards
Risk Categories:
Text-to-Text Content Safety, Text-to-Text Representational Harms, Memorization
Ethical Considerations:
Addressed content safety, representational harms, memorization, and large-scale harms
Responsible Ai Considerations 
Fairness:
Considered socio-cultural biases, evaluation, and pre-processing
Transparency:
Model card provides details on architecture and evaluation processes
Accountability:
Google and responsible AI toolkit recommendations
Mitigation Strategies:
Automation and manual evaluation for filtering and safety guidelines
Input Output 
Input Format:
Text string
Accepted Modalities:
text
Output Format:
Generated English-language text
Performance Tips:
Use CUDA for optimal performance
LLM NameGemma 2 2B
Repository ๐Ÿค—https://huggingface.co/google/gemma-2-2b 
Model Size2b
Required VRAM10.5 GB
Updated2024-12-04
Maintainergoogle
Model Typegemma2
Model Files  5.0 GB: 1-of-3   5.0 GB: 2-of-3   0.5 GB: 3-of-3
Model ArchitectureGemma2ForCausalLM
Licensegemma
Context Length8192
Model Max Length8192
Transformers Version4.42.4
Tokenizer ClassGemmaTokenizer
Padding Token<pad>
Vocabulary Size256000
Torch Data Typefloat32
Gemma 2 2B (google/gemma-2-2b)

Quantized Models of the Gemma 2 2B

Model
Likes
Downloads
VRAM
Gemma 2 2B Bnb 4bit6195052 GB

Best Alternatives to Gemma 2 2B

Best Alternatives
Context / RAM
Downloads
Likes
Gemma 2 2B It8K / 5.2 GB750121719
Gemma 2 2B Jpn It8K / 5.2 GB36305141
...emma 2 2B It Chinese Kyara DPO8K / 15.7 GB980116
Gemma 2 Baku 2B It8K / 10.5 GB726218
BgGPT Gemma 2 2.6B IT V1.08K / 5.2 GB5354
SauerkrautLM Gemma 2 2B It8K / 5.2 GB32248
2B Ad8K / 12.8 GB262
Shieldgemma 2B8K / 5.2 GB641649
Gemmasutra Mini 2B V18K / 5.2 GB23346
Gemma 2 2B It8K / 5.2 GB590810
Note: green Score (e.g. "73.2") means that the model is better than google/gemma-2-2b.

Rank the Gemma 2 2B 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? 38813 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124