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

Mpt 30B GGML by TheBloke

 »  All LLMs  »  TheBloke  »  Mpt 30B GGML   URL Share it on

  2   
  Arxiv:1909.08053   Arxiv:2010.04245   Arxiv:2108.12409   Arxiv:2205.14135   Arxiv:2302.06675   Arxiv:2302.13971   Composer   Dataset:allenai/c4   Dataset:allenai/s2orc Dataset:bigcode/the-stack-dedu...   Dataset:mc4 Dataset:togethercomputer/redpa...   Ggml   Llm-foundry   Mosaicml   Mpt   Quantized   Region:us   Streamingdatasets
Model Card on HF 🤗: https://huggingface.co/TheBloke/mpt-30B-GGML 

Mpt 30B GGML Benchmarks

LLME Score: 0.09552
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Mpt 30B GGML (TheBloke/mpt-30B-GGML)
🌟 Advertise your project 🚀

Mpt 30B GGML Parameters and Internals

Model Type 
decoder-style transformer
Use Cases 
Areas:
research, commercial applications
Applications:
text completion, dialogue generation, instruction following
Primary Use Cases:
Short-form instruction following, Chatbot dialogue generation
Limitations:
Not intended for deployment without finetuning, Potential to generate incorrect or offensive outputs
Considerations:
Should not be used for human-facing interactions without guardrails.
Additional Notes 
MPT-30B was trained primarily on English text and code using MosaicML's platform for efficient pretraining, finetuning, and inference.
Supported Languages 
English (proficient), Code (proficient)
Training Details 
Data Sources:
allenai/c4, mc4, togethercomputer/RedPajama-Data-1T, bigcode/the-stack-dedup, allenai/s2orc
Data Volume:
1T tokens
Methodology:
Trained from scratch with FlashAttention, ALiBi, QK LayerNorm optimizations
Context Length:
8192
Hardware Used:
440 A100-40GB GPUs, 216 A100-40GB GPUs, 256 H100-80GB GPUs
Model Architecture:
Modification of standard decoder-only transformer.
Input Output 
Accepted Modalities:
Text
Performance Tips:
Best used with MosaicML's llm-foundry for training and finetuning.
LLM NameMpt 30B GGML
Repository 🤗https://huggingface.co/TheBloke/mpt-30B-GGML 
Model Size30b
Required VRAM16.9 GB
Updated2025-08-18
MaintainerTheBloke
Model Typempt
Model Files  16.9 GB   18.7 GB   20.6 GB   22.5 GB   31.8 GB
GGML QuantizationYes
Quantization Typeggml
Model ArchitectureAutoModel
Licenseapache-2.0

Best Alternatives to Mpt 30B GGML

Best Alternatives
Context / RAM
Downloads
Likes
...e Llama 30B Instruct 2048 GGML0K / 13.7 GB420
30B Epsilon GGML0K / 13.7 GB29
Mpt 30B Chat GGML0K / 16.9 GB473
Mpt 30B Instruct GGML0K / 16.9 GB543
LLaMa 30B GGML0K / 13.6 GB424
Medalpaca Lora 30B 8bit0K / 0.2 GB015
Yayi2 30B Llama GGUF0K / 12.9 GB15710
LLaMA 30B GGUF0K / 13.5 GB7055
...e Llama 30B Instruct 2048 GGUF0K / 13.5 GB1764
Llama 30B Supercot GGUF0K / 13.5 GB1520
Note: green Score (e.g. "73.2") means that the model is better than TheBloke/mpt-30B-GGML.

Rank the Mpt 30B GGML 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? 50738 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 v20241124
  Support LLM Explorer