LLM EXPLORER 59,420 MODELS INDEXED

Midnight Miqu 70B V1.0 GGUF by ooooz

By ooooz · 398 downloads
HF Repo ↗ Copy URL

Midnight Miqu 70B V1.0 GGUF is an open-source language model by ooooz. Features: 70b LLM, VRAM: 29.9GB, Context: 31K, Quantized, LLM Explorer Score: 0.12.

  Conversational   Endpoints compatible   Gguf   Llama   Quantized   Region:us
Model Card on HF 🤗: https://huggingface.co/ooooz/midnight-miqu-70b-v1.0-GGUF 

Midnight Miqu 70B V1.0 GGUF Parameters and Internals

Model Type 
text generation, multimodal
Use Cases 
Areas:
Research, Commercial Applications
Applications:
Interactive AI, Content Generation
Primary Use Cases:
AI Assistants, Text generation tools
Limitations:
Not suitable for critical decision-making tasks
Considerations:
Ensure usage aligns with ethical guidelines.
Additional Notes 
Excels in creative tasks and information synthesis.
Supported Languages 
English (High Proficiency)
Training Details 
Data Sources:
Diverse internet datasets
Data Volume:
Thousands of GBs
Methodology:
Standard pre-training with fine-tuning on specific tasks
Context Length:
4096
Training Time:
Several months on dedicated HPC
Hardware Used:
A100 GPUs
Model Architecture:
Transformer-based architecture
Safety Evaluation 
Methodologies:
Red teaming
Findings:
Reduced propensity for harmful outputs
Risk Categories:
Bias, Misinformation
Ethical Considerations:
Model should not be used in applications where inaccurate responses could result in harm.
Responsible Ai Considerations 
Fairness:
Training datasets include diverse data to reduce bias.
Transparency:
Full architectural details available in the accompanying paper.
Accountability:
Developers are accountable for the training data and initial release.
Mitigation Strategies:
Continuous monitoring and updates for harmful data.
Input Output 
Input Format:
Text
Accepted Modalities:
text
Output Format:
Text Output
Performance Tips:
Prefer GPU deployment for real-time inference.
Release Notes 
Version:
1.0
Date:
2023-10-15
Notes:
Initial release, introducing superior efficiency in text generation tasks.
LLM NameMidnight Miqu 70B V1.0 GGUF
Repository 🤗https://huggingface.co/ooooz/midnight-miqu-70b-v1.0-GGUF 
Base Model(s)  sophosympatheia/Midnight-Miqu-70B-v1.0   sophosympatheia/Midnight-Miqu-70B-v1.0
Model Size70b
Required VRAM29.9 GB
Updated2026-07-22
Maintainerooooz
Model Typellama
Model Files  33.3 GB   29.9 GB   41.4 GB   39.2 GB   48.8 GB   47.5 GB
GGUF QuantizationYes
Quantization Typegguf
Model ArchitectureLlamaForCausalLM
Context Length32764
Model Max Length32764
Transformers Version4.36.2
Vocabulary Size32000
Torch Data Typefloat16

Best Alternatives to Midnight Miqu 70B V1.0 GGUF

Best Alternatives
Context / RAM
Downloads
Likes
Grok Oss Revenant 70B128K / 141.9 GB85032
...Seek R1 Distill Llama 70B GGUF128K / 15.9 GB27430120
Llama 3.3 70B Instruct GGUF128K / 15.9 GB23398124
R1 1776 Distill Llama 70B GGUF128K / 26.4 GB247824
Reflection Llama 3.1 70B Bf16128K / 141.9 GB2836
Reflection Llama 3.1 70B GGUF128K / 26.4 GB1155
...Horizon AI Korean Advanced 70B128K / 141.9 GB301
...qu 1 70B 24GB VRAM IQ2 XS SOTA31K / 20.3 GB681
Llama 3 70B Quantised8K / 48.7 GB132
...ma3 70B Chinese Chat GGUF 4bit8K / 40 GB19618
Note: green Score (e.g. "73.2") means that the model is better than ooooz/midnight-miqu-70b-v1.0-GGUF.