LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Mistral NeMo Minitron 8B Instruct by nvidia

 »  All LLMs  »  nvidia  »  Mistral NeMo Minitron 8B Instruct   URL Share it on

  2450   
  Arxiv:2406.11704   Arxiv:2407.14679   Autotrain compatible Base model:finetune:nvidia/mis... Base model:nvidia/mistral-nemo...   Conversational   Endpoints compatible   Instruct   Mistral   Nemo   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/nvidia/Mistral-NeMo-Minitron-8B-Instruct 

Mistral NeMo Minitron 8B Instruct Benchmarks

MMLU Pro: 33.23
GPQA: 5.03
MUSR: 7.37
BBH: 34.13
IFEval: 50.04 vs 88 (so35)-43.1%
MATH Lvl 5: 0.53
LLME Score: 0.38085
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Mistral NeMo Minitron 8B Instruct (nvidia/Mistral-NeMo-Minitron-8B-Instruct)

Mistral NeMo Minitron 8B Instruct Parameters and Internals

Model Type 
Transformer Decoder (Auto-regressive Language Model)
Use Cases 
Primary Use Cases:
Roleplaying, Retrieval augmented generation, Function calling
Limitations:
Model may amplify societal biases and return toxic responses, May produce inaccurate or unacceptable text
Considerations:
Validate the imported packages are from a trusted source for end-to-end security.
Training Details 
Methodology:
Multi-stage SFT and preference-based alignment with NeMo Aligner
Context Length:
8192
Model Architecture:
Transformer Decoder with Grouped-Query Attention (GQA) and Rotary Position Embeddings (RoPE), 40 layers, 32 attention heads.
Safety Evaluation 
Methodologies:
Garak automated LLM vulnerability scanner, AEGIS content safety evaluation, Human Content Red Teaming
Ethical Considerations:
NVIDIA encourages working with internal model team to ensure model meets specific industry and use case requirements.
Input Output 
Input Format:
System {system prompt} User {prompt} Assistant\n
Performance Tips:
The model may not perform optimally without the recommended prompt template.
LLM NameMistral NeMo Minitron 8B Instruct
Repository 🤗https://huggingface.co/nvidia/Mistral-NeMo-Minitron-8B-Instruct 
Base Model(s)  Mistral NeMo Minitron 8B Base   nvidia/Mistral-NeMo-Minitron-8B-Base
Model Size8b
Required VRAM16.8 GB
Updated2024-12-21
Maintainernvidia
Model Typemistral
Instruction-BasedYes
Model Files  4.9 GB: 1-of-4   5.0 GB: 2-of-4   4.9 GB: 3-of-4   2.0 GB: 4-of-4
Model ArchitectureMistralForCausalLM
Licenseother
Context Length8192
Model Max Length8192
Transformers Version4.44.0
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size131072
Torch Data Typebfloat16

Best Alternatives to Mistral NeMo Minitron 8B Instruct

Best Alternatives
Context / RAM
Downloads
Likes
Ministral 8B Instruct 2410 HF32K / 32 GB5009710
Ministrations 8B V132K / 16.1 GB14315
Ministral 8B Slerp32K / 29.2 GB270
...flect Mini8Bit Om2 460k Sft T132K / 16.1 GB1300
...ruct 2410 MetaMathQA DPO Iter132K / 16.1 GB3870
...t Ministral8Bit MMQA Mix Iter232K / 16.1 GB1320
...t Ministral8Bit MMQA DPO Iter132K / 60.8 GB890
...ect Ministral8Bit Mg DPO Psdp232K / 16.1 GB690
...t Ministral8Bit Math DPO Iter132K / 16.1 GB530
...t Ministral8Bit MMQA DPO Iter132K / 16.1 GB420
Note: green Score (e.g. "73.2") means that the model is better than nvidia/Mistral-NeMo-Minitron-8B-Instruct.

Rank the Mistral NeMo Minitron 8B 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! 🌟

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? 40013 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241217
  Support LLM Explorer