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

1.5 Pints 16K V0.1 by pints-ai

 »  All LLMs  »  pints-ai  »  1.5 Pints 16K V0.1   URL Share it on

  4   
  Arxiv:2408.03506   Conversational Dataset:huggingfaceh4/deita-10... Dataset:huggingfaceh4/ultracha... Dataset:huggingfaceh4/ultrafee...   Dataset:ldjnr/capybara   Dataset:meta-math/metamathqa Dataset:open-orca/slimorca-ded... Dataset:pints-ai/expository-pr... Dataset:togethercomputer/llama... Dataset:wizardlm/wizardlm evol...   En   Instruct   Llama   Model-index   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/pints-ai/1.5-Pints-16K-v0.1 

1.5 Pints 16K V0.1 Benchmarks

MMLU Pro: 1.32
MUSR: 2.73
BBH: 3.66
IFEval: 16.36 vs 88 (so35)-81.4%
MATH Lvl 5: 1.44
LLME Score: 0.18526
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
1.5 Pints 16K V0.1 (pints-ai/1.5-Pints-16K-v0.1)
🌟 Advertise your project 🚀

1.5 Pints 16K V0.1 Parameters and Internals

Model Type 
Large Language Model, Text Generation
Use Cases 
Areas:
Research, Education
Primary Use Cases:
User assistance, Reasoning
Limitations:
Not suitable for full knowledge retrieval without documented retrieval augmented generation.
Considerations:
Finetune for domain adaptation for specialized tasks; use a repetition penalty of 1.3 for full performance.
Additional Notes 
The model has been preference-optimized using the ChatML template for specific multi-turn conversational tasks.
Supported Languages 
en (Proficient)
Training Details 
Data Sources:
pints-ai/Expository-Prose-V1, HuggingFaceH4/ultrachat_200k, Open-Orca/SlimOrca-Dedup, meta-math/MetaMathQA, HuggingFaceH4/deita-10k-v0-sft, WizardLM/WizardLM_evol_instruct_V2_196k, togethercomputer/llama-instruct, LDJnr/Capybara, HuggingFaceH4/ultrafeedback_binarized
Data Volume:
57 billion tokens
Methodology:
Pre-training emphasizes quality over quantity. Fine-tuning and DPO follow the ChatML template.
Context Length:
16384
Training Time:
9 days
Hardware Used:
GPU with at least 8GB of VRAM
Model Architecture:
Llama 2 Autoregressive Model with Mistral tokenizer and Float32 precision.
Input Output 
Input Format:
Chat representation using ChatML template.
Accepted Modalities:
Text
Output Format:
Generated text output from user prompts.
Performance Tips:
Use a repetition penalty of 1.3 to optimize output effectivity.
LLM Name1.5 Pints 16K V0.1
Repository 🤗https://huggingface.co/pints-ai/1.5-Pints-16K-v0.1 
Model Size1.6b
Required VRAM3.1 GB
Updated2025-09-07
Maintainerpints-ai
Model Typellama
Instruction-BasedYes
Model Files  3.1 GB
Supported Languagesen
Model ArchitectureLlamaForCausalLM
Licensemit
Context Length16384
Model Max Length16384
Transformers Version4.38.0
Tokenizer ClassLlamaTokenizer
Padding Token<|pad|>
Vocabulary Size32064
Torch Data Typebfloat16

Best Alternatives to 1.5 Pints 16K V0.1

Best Alternatives
Context / RAM
Downloads
Likes
Fox 1 1.6B Instruct V0.18K / 3.3 GB1014
1.5 Pints 2K V0.12K / 3.1 GB717

Rank the 1.5 Pints 16K V0.1 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? 51187 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