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

Dbrx Instruct by databricks

 »  All LLMs  »  databricks  »  Dbrx Instruct   URL Share it on

  13287   
  Arxiv:2211.15841   Arxiv:2304.11277   Autotrain compatible   Conversational   Dbrx   Instruct   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/databricks/dbrx-instruct 

Dbrx Instruct Benchmarks

MMLU Pro: 29.81
GPQA: 12.19
MUSR: 12.2
BBH: 35.96
IFEval: 54.16 vs 88 (so35)-38.5%
ARC: 67.83 vs 96.7 (so35)-29.9%
HellaSwag: 88.85 vs 95.3 (gpt4)-6.8%
MMLU: 73.72 vs 88.3 (so35)-16.5%
TruthfulQA: 67.02 vs 59 (gpt4)13.6%
WinoGrande: 82.08 vs 87.5 (gpt4)-6.2%
GSM8K: 67.32 vs 96.4 (so35)-30.2%
MATH Lvl 5: 6.87
LLME Score: 0.36389
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Dbrx Instruct (databricks/dbrx-instruct)

Dbrx Instruct Parameters and Internals

Model Type 
Large Language Model, Text Generation
Use Cases 
Areas:
Research, Commercial applications
Applications:
General English-language tasks, Coding tasks
Primary Use Cases:
Few-turn question answering
Limitations:
Not tested for non-English proficiency, No multimodal capabilities
Considerations:
Evaluation around safety for specific applications recommended
Supported Languages 
English (General Text Processing)
Training Details 
Data Sources:
12T tokens of text and code
Data Volume:
12T tokens
Methodology:
Curriculum learning, changed data mix during training
Context Length:
32768
Hardware Used:
Databricks infrastructure
Model Architecture:
Fine-grained MoE with 16 experts
Input Output 
Input Format:
Text-based, up to 32768 tokens
Accepted Modalities:
text
Output Format:
Text-based
Performance Tips:
Using FlashAttention2 is recommended for faster inference.
Release Notes 
Version:
1.0
Notes:
Instruction finetuned, Mixture-of-experts model
LLM NameDbrx Instruct
Repository 🤗https://huggingface.co/databricks/dbrx-instruct 
Model Size131.6b
Required VRAM181.1 GB
Updated2025-02-22
Maintainerdatabricks
Model Typedbrx
Instruction-BasedYes
Model Files  3.5 GB: 1-of-61   4.4 GB: 2-of-61   4.2 GB: 3-of-61   4.4 GB: 4-of-61   4.4 GB: 5-of-61   4.2 GB: 6-of-61   4.4 GB: 7-of-61   4.4 GB: 8-of-61   4.2 GB: 9-of-61   4.4 GB: 10-of-61   4.4 GB: 11-of-61   4.2 GB: 12-of-61   4.4 GB: 13-of-61   4.4 GB: 14-of-61   4.2 GB: 15-of-61   4.4 GB: 16-of-61   4.4 GB: 17-of-61   4.2 GB: 18-of-61   4.4 GB: 19-of-61   4.4 GB: 20-of-61   4.2 GB: 21-of-61   4.4 GB: 22-of-61   4.4 GB: 23-of-61   4.2 GB: 24-of-61   4.4 GB: 25-of-61   4.4 GB: 26-of-61   4.2 GB: 27-of-61   4.4 GB: 28-of-61   4.4 GB: 29-of-61   4.2 GB: 30-of-61   4.4 GB: 31-of-61   4.4 GB: 32-of-61   4.2 GB: 33-of-61   4.4 GB: 34-of-61   4.4 GB: 35-of-61   4.2 GB: 36-of-61   4.4 GB: 37-of-61   4.4 GB: 38-of-61   4.2 GB: 39-of-61   4.4 GB: 40-of-61   4.4 GB: 41-of-61   4.2 GB: 42-of-61
Gated ModelYes
Model ArchitectureDbrxForCausalLM
Licenseproprietary
Transformers Version4.40.0
Vocabulary Size100352
Torch Data Typebfloat16

Best Alternatives to Dbrx Instruct

Best Alternatives
Context / RAM
Downloads
Likes
Dbrx Instruct0K / 181.1 GB2220
...rx Instruct Quantization Fixed0K / 176.9 GB159

Rank the Dbrx 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? 43470 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 v20241227
  Support LLM Explorer