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

Llama 3 Lumimaid 8B V0.1 GTPQ by mikudev

 »  All LLMs  »  mikudev  »  Llama 3 Lumimaid 8B V0.1 GTPQ   URL Share it on

  18   
  4-bit   4bit   Autotrain compatible   Conversational   Endpoints compatible   Gptq   Llama   Not-for-all-audiences   Nsfw   Quantized   Region:us
Model Card on HF 🤗: https://huggingface.co/mikudev/Llama-3-Lumimaid-8B-v0.1-GTPQ 

Llama 3 Lumimaid 8B V0.1 GTPQ Benchmarks

ARC: 61.09 vs 96.7 (so35)-36.8%
HellaSwag: 79.63 vs 95.3 (gpt4)-16.4%
MMLU: 67.52 vs 88.3 (so35)-23.5%
TruthfulQA: 49.82 vs 59 (gpt4)-15.6%
WinoGrande: 75.06 vs 87.5 (gpt4)-14.2%
GSM8K: 66.19 vs 96.4 (so35)-31.3%
LLME Score: 0.19819
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Llama 3 Lumimaid 8B V0.1 GTPQ (mikudev/Llama-3-Lumimaid-8B-v0.1-GTPQ)

Llama 3 Lumimaid 8B V0.1 GTPQ Parameters and Internals

Model Type 
text generation
Additional Notes 
This is the first quantized version by mikudev. Feedback is encouraged.
Training Details 
Data Sources:
Aesir datasets, NoRobots, limarp, toxic-dpo-v0.1-sharegpt, ToxicQAFinal, Luminae-i2, Squish42/bluemoon-fandom-1-1-rp-cleaned, NobodyExistsOnTheInternet/PIPPAsharegptv2test, cgato/SlimOrcaDedupCleaned, Airoboros, Capybara, Initial LumiMaid 8B Finetune, Undi95/Llama-3-Unholy-8B-e4, Undi95/Llama-3-LewdPlay-8B
Methodology:
Quantized using AutoGPTQ with 8192 seqlen using wikitext2 example
Context Length:
8192
Input Output 
Input Format:
<|begin_of_text|><|start_header_id|>system<|end_header_id|>{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>{input}<|eot_id|><|start_header_id|>assistant<|end_header_id|>{output}<|eot_id|>
LLM NameLlama 3 Lumimaid 8B V0.1 GTPQ
Repository 🤗https://huggingface.co/mikudev/Llama-3-Lumimaid-8B-v0.1-GTPQ 
Model Size8b
Required VRAM5.7 GB
Updated2025-02-22
Maintainermikudev
Model Typellama
Model Files  5.7 GB
GPTQ QuantizationYes
Quantization Typegptq|4bit
Model ArchitectureLlamaForCausalLM
Licensecc-by-nc-4.0
Context Length8192
Model Max Length8192
Transformers Version4.39.0.dev0
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size128256
Torch Data Typefloat16

Best Alternatives to Llama 3 Lumimaid 8B V0.1 GTPQ

Best Alternatives
Context / RAM
Downloads
Likes
...a 3 8B Instruct 262K 4bit GPTQ256K / 5.8 GB171
... 8B Instruct 262K 4bit GPTQ 02256K / 5.7 GB60
...lama 3.1 8B Instruct GPTQ INT4128K / 5.8 GB1052922
...Instruct 80K Qlora Merged GPTQ80K / 5.8 GB820
...oLeo Instruct 8B 32K V0.1 GPTQ64K / 5.7 GB60
Llama3 German 8B 32K GPTQ64K / 5.7 GB50
Tsukasa Llama 3 8B Qlora Gptq32K / 5.8 GB100
Llama 3 Soliloquy 8B V2 GPTQ24K / 5.7 GB101
Llama 3 Soliloquy 8B GPTQ16K / 5.7 GB105
Meta Llama 3 8B Instruct GPTQ8K / 5.7 GB19683
Note: green Score (e.g. "73.2") means that the model is better than mikudev/Llama-3-Lumimaid-8B-v0.1-GTPQ.

Rank the Llama 3 Lumimaid 8B V0.1 GTPQ 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