New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa by denru

 ยป  All LLMs  ยป  denru  ยป  New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa   URL Share it on

  Merged Model   4-bit   Autotrain compatible Base model:meta-llama/llama-3.... Base model:sophosympatheia/new...   Conversational   Endpoints compatible   Exl2   Instruct   Llama   Not-for-all-audiences   Quantized   Region:us   Safetensors   Sharded   Tensorflow

New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa (denru/New-Dawn-Llama-3.1-70B-v1.1-4_0bpw-h6-exl2-pippa)

New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa Parameters and Internals

Model Type 
roleplaying, storytelling
Use Cases 
Areas:
research, roleplaying, storytelling
Limitations:
uncensored, contains mature themes and content
Additional Notes 
This model is a merge using a specially developed technique by jukofyork for preserving context capabilities.
Input Output 
Input Format:
Prompt template JSON for customizations.
Accepted Modalities:
text
Output Format:
Storytelling and roleplaying responses
Performance Tips:
Use Quadratic Sampling with smoothing factor ~0.2, Min-P between 0 and 0.1, DRY repetition penalty for best results.
LLM NameNew Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa
Repository ๐Ÿค—https://huggingface.co/denru/New-Dawn-Llama-3.1-70B-v1.1-4_0bpw-h6-exl2-pippa 
Base Model(s)  sophosympatheia/New-Dawn-Llama-3-70B-32K-v1.0   meta-llama/Meta-Llama-3.1-70B-Instruct   sophosympatheia/New-Dawn-Llama-3-70B-32K-v1.0   meta-llama/Meta-Llama-3.1-70B-Instruct
Merged ModelYes
Model Size70b
Required VRAM37.3 GB
Updated2025-01-17
Maintainerdenru
Model Typellama
Instruction-BasedYes
Model Files  8.5 GB: 1-of-5   8.6 GB: 2-of-5   8.6 GB: 3-of-5   8.6 GB: 4-of-5   3.0 GB: 5-of-5
Quantization Typeexl2
Model ArchitectureLlamaForCausalLM
Licensellama3.1
Context Length131072
Model Max Length131072
Transformers Version4.43.3
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size128256
Torch Data Typefloat16

Best Alternatives to New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa

Best Alternatives
Context / RAM
Downloads
Likes
...B Instruct Gradient 1048K 8bit1024K / 75 GB175
...B Instruct Gradient 1048K 4bit1024K / 39.7 GB203
...B Instruct Gradient 1048K 2bit1024K / 21.9 GB232
...0B Instruct Gradient 262K 4bit256K / 39.7 GB183
...0B Instruct Gradient 262K 2bit256K / 21.9 GB251
...0B Instruct Gradient 262K 8bit256K / 75 GB142
... Gradient 262K 2.25bpw H6 EXL2256K / 22.2 GB180
...t Gradient 262K 4.0bpw H6 EXL2256K / 37.2 GB171
...t Gradient 262K 2.4bpw H6 EXL2256K / 23.5 GB190
...t Gradient 262K 3.5bpw H6 EXL2256K / 32.9 GB160

Rank the New Dawn Llama 3.1 70B V1.1 4 0bpw H6 EXL2 Pippa Capabilities

๐Ÿ†˜ 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! ๐ŸŒŸ

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? 41473 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227