MiniG by CausalLM

 ยป  All LLMs  ยป  CausalLM  ยป  MiniG   URL Share it on

  Chatglm   Co2 eq emissions   Custom code   De   En   Feature-extraction   Ja   Model-index   Region:us   Safetensors   Zh
Model Card on HF ๐Ÿค—: https://huggingface.co/CausalLM/miniG 

MiniG Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
MiniG (CausalLM/miniG)

MiniG Parameters and Internals

Model Type 
text-generation, multimodal
Additional Notes 
Model trained on unfiltered internet data, may contain objectionable content.
Supported Languages 
en (high), zh (high), ja (high), de (high)
Training Details 
Data Sources:
synthetic dataset generated using large context windows, retrieval-augmented generation, and knowledge graph integration
Data Volume:
20 billion tokens
Methodology:
fine-tuning using a synthesis dataset
Context Length:
1000000
Training Time:
< 1 day on 16 nodes of 8*A100-80G
Hardware Used:
16 nodes of 8*A100-80G GPUs
Input Output 
Input Format:
Accepts text and image modalities.
Accepted Modalities:
text, image
Performance Tips:
Use a standardized implementation for inference to avoid performance degradation. For fewer hallucinations, use top_p=0.8 and temperature=0.3, or temperature=0.2.
LLM NameMiniG
Repository ๐Ÿค—https://huggingface.co/CausalLM/miniG 
Model Size14b
Required VRAM28 GB
Updated2024-11-24
MaintainerCausalLM
Model Typechatglm
Model Files  28.0 GB
Supported Languagesen zh ja de
Gated ModelYes
Model ArchitectureChatGLMForConditionalGeneration
Licenseproprietary
Model Max Length8192
Transformers Version4.44.0
Tokenizer ClassChatGLM4Tokenizer
Padding Token<|endoftext|>
Vocabulary Size151552
Torch Data Typebfloat16

Rank the MiniG 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? 41363 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227