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 Chat Q8 by Abzu

 »  All LLMs  »  Abzu  »  Mpt 30B Chat Q8   URL Share it on

  19   
  Arxiv:2010.04245   Arxiv:2108.12409   Arxiv:2205.14135   8-bit   Autotrain compatible   Codegen   Composer   Custom code   Instruct   Llm-foundry   Mosaicml   Mpt   Q8   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Abzu/mpt-30b-chat-q8 

Mpt 30B Chat Q8 Benchmarks

LMSys ELO: 1045 vs 1272 (so35)-17.8%
ARC: 58.7 vs 96.7 (so35)-39.3%
HellaSwag: 82.54 vs 95.3 (gpt4)-13.4%
MMLU: 51.16 vs 88.3 (so35)-42.1%
TruthfulQA: 52.42 vs 59 (gpt4)-11.2%
WinoGrande: 75.3 vs 87.5 (gpt4)-13.9%
GSM8K: 12.13 vs 96.4 (so35)-87.4%
LLME Score: 0.21821
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Mpt 30B Chat Q8 (Abzu/mpt-30b-chat-q8)

Mpt 30B Chat Q8 Parameters and Internals

Model Type 
text generation, chatbot
Additional Notes 
The model was trained with MPT-30B tokenizer, based on EleutherAI/gpt-neox-20b tokenizer. The model employed techniques like FlashAttention and ALiBi for hardware efficiency.
Training Details 
Data Sources:
camel-ai/code, ehartford/wizard_vicuna_70k_unfiltered, anon8231489123/ShareGPT_Vicuna_unfiltered, teknium1/GPTeacher/roleplay-instruct-v2-final, teknium1/GPTeacher/codegen-isntruct, timdettmers/openassistant-guanaco, camel-ai/math, project-baize/baize-chatbot/medical_chat_data, project-baize/baize-chatbot/quora_chat_data, project-baize/baize-chatbot/stackoverflow_chat_data, camel-ai/biology, camel-ai/chemistry, camel-ai/ai_society, jondurbin/airoboros-gpt4-1.2, LongConversations, camel-ai/physics
Data Volume:
various public datasets
Methodology:
finetuning
Context Length:
8192
Training Time:
7.6 hours
Hardware Used:
64 H100 GPUs
Model Architecture:
modified decoder-only transformer
Input Output 
Input Format:
sequence of tokens (text)
Accepted Modalities:
text
Output Format:
text sequence (text generation)
Performance Tips:
Use 'trust_remote_code=True' for model architecture adaptations
LLM NameMpt 30B Chat Q8
Repository 🤗https://huggingface.co/Abzu/mpt-30b-chat-q8 
Base Model(s)  Mpt 30B Chat   mosaicml/mpt-30b-chat
Model Size30b
Required VRAM30.4 GB
Updated2024-12-22
MaintainerAbzu
Model Typempt
Instruction-BasedYes
Model Files  10.0 GB: 1-of-4   9.9 GB: 2-of-4   9.9 GB: 3-of-4   0.6 GB: 4-of-4
Quantization Typeq8
Generates CodeYes
Model ArchitectureMPTForCausalLM
Licensecc-by-nc-sa-4.0
Model Max Length8192
Transformers Version4.30.2
Tokenizer ClassGPTNeoXTokenizer
Vocabulary Size50432
Torch Data Typebfloat16

Best Alternatives to Mpt 30B Chat Q8

Best Alternatives
Context / RAM
Downloads
Likes
Ct2fast Mpt 30B Chat0K / 30 GB102

Rank the Mpt 30B Chat Q8 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? 40066 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