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

Airophin 13B Pntk 16K Fp16 by bhenrym14

 »  All LLMs  »  bhenrym14  »  Airophin 13B Pntk 16K Fp16   URL Share it on

  1241   
  Arxiv:2306.15595   Autotrain compatible   Dataset:ehartford/dolphin Dataset:jondurbin/airoboros-gp...   Endpoints compatible   Ext 8k   Fp16   Llama   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/bhenrym14/airophin-13b-pntk-16k-fp16 

Airophin 13B Pntk 16K Fp16 Benchmarks

ARC: 61.18 vs 96.7 (so35)-36.7%
HellaSwag: 82.86 vs 95.3 (gpt4)-13.1%
MMLU: 55.19 vs 88.3 (so35)-37.5%
TruthfulQA: 43.2 vs 59 (gpt4)-26.8%
WinoGrande: 76.16 vs 87.5 (gpt4)-13%
GSM8K: 8.04 vs 96.4 (so35)-91.7%
LLME Score: 0.16131
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Airophin 13B Pntk 16K Fp16 (bhenrym14/airophin-13b-pntk-16k-fp16)

Airophin 13B Pntk 16K Fp16 Parameters and Internals

Model Type 
Text Generation
Use Cases 
Areas:
Research, Commercial Applications
Primary Use Cases:
Text generation, Question answering, Coding assistance
Additional Notes 
The model employs Partial NTK Rope Scaling, a method not yet natively implemented in Transformers.
Training Details 
Data Sources:
jondurbin/airoboros-gpt4-1.4.1, ehartford/dolphin
Data Volume:
110 mm tokens in first phase
Methodology:
QLoRA fine-tune with NTK Rope Scaling
Context Length:
16384
Training Time:
~35 hours total
Hardware Used:
1x RTX 6000 Ada
Input Output 
Input Format:
Airoboros-like prompting
LLM NameAirophin 13B Pntk 16K Fp16
Repository 🤗https://huggingface.co/bhenrym14/airophin-13b-pntk-16k-fp16 
Model Size13b
Required VRAM26 GB
Updated2025-02-05
Maintainerbhenrym14
Model Typellama
Model Files  9.9 GB: 1-of-3   9.9 GB: 2-of-3   6.2 GB: 3-of-3
Context Length8k
Quantization Typefp16
Model ArchitectureLlamaForCausalLM
Context Length16384
Model Max Length16384
Transformers Version4.31.0
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size32000
Torch Data Typefloat16

Best Alternatives to Airophin 13B Pntk 16K Fp16

Best Alternatives
Context / RAM
Downloads
Likes
Llama13b 32K Illumeet Finetune32K / 26 GB50
...Maid V3 13B 32K 6.0bpw H6 EXL232K / 10 GB41
...Maid V3 13B 32K 8.0bpw H8 EXL232K / 13.2 GB41
WhiteRabbitNeo 13B V116K / 26 GB2329411
CodeLlama 13B Python Fp1616K / 26 GB235725
CodeLlama 13B Instruct Fp1616K / 26 GB243528
Codellama 13B Bnb 4bit16K / 7.2 GB1031
CodeLlama 13B Fp1616K / 26 GB1866
...Llama 13B Instruct Hf 4bit MLX16K / 7.8 GB662
WhiteRabbitNeo 13B V1 4bit Mlx16K / 7.8 GB71
Note: green Score (e.g. "73.2") means that the model is better than bhenrym14/airophin-13b-pntk-16k-fp16.

Rank the Airophin 13B Pntk 16K Fp16 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? 42577 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