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 8B ProLong 512K Instruct by princeton-nlp

 »  All LLMs  »  princeton-nlp  »  Llama 3 8B ProLong 512K Instruct   URL Share it on

  3184   
  Arxiv:2410.02660 Base model:finetune:princeton-... Base model:princeton-nlp/llama... Dataset:huggingfaceh4/ultracha... Dataset:princeton-nlp/prolong-... Dataset:princeton-nlp/prolong-...   Instruct   Llama   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/princeton-nlp/Llama-3-8B-ProLong-512k-Instruct 

Llama 3 8B ProLong 512K Instruct Benchmarks

MMLU Pro: 24.96
GPQA: 4.14
MUSR: 12.09
BBH: 28.67
IFEval: 39.78 vs 88 (so35)-54.8%
MATH Lvl 5: 5.82
LLME Score: 0.30392
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Llama 3 8B ProLong 512K Instruct (princeton-nlp/Llama-3-8B-ProLong-512k-Instruct)

Llama 3 8B ProLong 512K Instruct Parameters and Internals

Model Type 
long-context instruction-tuned
Additional Notes 
The model is the first open-source model to effectively support a 512K context window. Extensive ablations conducted to find effective ways to extend LM’s context length.
Training Details 
Data Sources:
Books, Textbooks, The Stack V1, StackExchange, Tulu-v2, Wikipedia, Arxiv, OpenWebMath, FineWeb, FineWeb-EDU
Data Volume:
40B tokens
Methodology:
long-context training, fine-tuning on UltraChat
Context Length:
512000
LLM NameLlama 3 8B ProLong 512K Instruct
Repository 🤗https://huggingface.co/princeton-nlp/Llama-3-8B-ProLong-512k-Instruct 
Base Model(s)  Llama 3 8B ProLong 512K Base   princeton-nlp/Llama-3-8B-ProLong-512k-Base
Model Size8b
Required VRAM32.1 GB
Updated2025-03-03
Maintainerprinceton-nlp
Model Typellama
Instruction-BasedYes
Model Files  4.9 GB: 1-of-7   4.8 GB: 2-of-7   5.0 GB: 3-of-7   5.0 GB: 4-of-7   4.8 GB: 5-of-7   5.0 GB: 6-of-7   2.6 GB: 7-of-7
Model ArchitectureLlamaForCausalLM
Licensellama3
Context Length524288
Model Max Length524288
Tokenizer ClassPreTrainedTokenizerFast
Padding Token!
Vocabulary Size128256
Torch Data Typefloat32

Best Alternatives to Llama 3 8B ProLong 512K Instruct

Best Alternatives
Context / RAM
Downloads
Likes
...a 3 8B Instruct Gradient 1048K1024K / 16.1 GB4185680
931024K / 16.1 GB840
1171024K / 16.1 GB330
93a1024K / 16.1 GB200
1261024K / 16.1 GB130
1361024K / 16.1 GB130
Mpasila Viking 8B1024K / 16.1 GB840
1331024K / 16.1 GB110
1371024K / 16.1 GB80
161024K / 16.1 GB1690
Note: green Score (e.g. "73.2") means that the model is better than princeton-nlp/Llama-3-8B-ProLong-512k-Instruct.

Rank the Llama 3 8B ProLong 512K Instruct 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? 44127 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