Newton 7B 8.0bpw H8 EXL2 by LoneStriker

 ยป  All LLMs  ยป  LoneStriker  ยป  Newton 7B 8.0bpw H8 EXL2   URL Share it on

  Arxiv:2305.14314   Autotrain compatible   Axolotl Base model:finetune:openchat/o... Base model:openchat/openchat-3...   Conversational   Dataset:allenai/ai2 arc   Dataset:camel-ai/biology   Dataset:camel-ai/chemistry   Dataset:camel-ai/math   Dataset:camel-ai/physics   Dataset:derek-thomas/scienceqa Dataset:hendrycks/competition ...   Dataset:mandyyyyii/scibench   Dataset:metaeval/reclor   Dataset:openbookqa   Dataset:piqa   Dataset:sciq Dataset:stem-ai-mtl/electrical...   Dataset:tiger-lab/scienceeval   Endpoints compatible   Exl2   Finetuned   Mistral   Qlora   Quantized   Region:us   Safetensors

Newton 7B 8.0bpw H8 EXL2 Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Newton 7B 8.0bpw H8 EXL2 (LoneStriker/Newton-7B-8.0bpw-h8-exl2)

Newton 7B 8.0bpw H8 EXL2 Parameters and Internals

Model Type 
MistralForCausalLM
Additional Notes 
Model is a fine-tuned version using various scientific datasets for improved performance in science-related tasks.
Training Details 
Data Sources:
hendrycks/competition_math, allenai/ai2_arc, camel-ai/physics, camel-ai/chemistry, camel-ai/biology, camel-ai/math, STEM-AI-mtl/Electrical-engineering, openbookqa, piqa, metaeval/reclor, mandyyyyii/scibench, derek-thomas/ScienceQA, sciq, TIGER-Lab/ScienceEval
Methodology:
Fine-tuned using QLoRa and axolotl
Context Length:
8192
Model Architecture:
Mistral derived model with QLoRa adaptation
Input Output 
Input Format:
GPT4 Correct User: {instruction}<|end_of_turn|>GPT4 Correct Assistant
Accepted Modalities:
text
Output Format:
GPT4 Correct Assistant
LLM NameNewton 7B 8.0bpw H8 EXL2
Repository ๐Ÿค—https://huggingface.co/LoneStriker/Newton-7B-8.0bpw-h8-exl2 
Base Model(s)  openchat/openchat-3.5-0106   openchat/openchat-3.5-0106
Model Size7b
Required VRAM7.4 GB
Updated2024-12-26
MaintainerLoneStriker
Model Typemistral
Model Files  7.4 GB
Quantization Typeexl2
Model ArchitectureMistralForCausalLM
Licenseother
Context Length8192
Model Max Length8192
Transformers Version4.37.0
Tokenizer ClassLlamaTokenizer
Padding Token</s>
Vocabulary Size32002
Torch Data Typebfloat16

Best Alternatives to Newton 7B 8.0bpw H8 EXL2

Best Alternatives
Context / RAM
Downloads
Likes
...al Nemo Instruct 2407 Bnb 4bit1000K / 8.3 GB1288425
...istral Nemo Base 2407 Bnb 4bit1000K / 8.3 GB502612
...t 3.5 0106 128K 4.0bpw H6 EXL2128K / 3.9 GB181
...t 3.5 0106 128K 8.0bpw H8 EXL2128K / 7.4 GB181
...tral 7B Instruct V0.3 Bnb 4bit32K / 4.1 GB35913715
Mistral 7B V0.3 Bnb 4bit32K / 4.1 GB6271413
NaturalLM 7B Instruct32K / 14.5 GB8210
Mistral 7B Instruct V0.2 4bit32K / 4.3 GB4461
Mistral 7B Instruct V0.2 Fp1632K / 14.4 GB100
...tral 7B Instruct V0.2 Bnb 4bit32K / 4.1 GB1104831
Note: green Score (e.g. "73.2") means that the model is better than LoneStriker/Newton-7B-8.0bpw-h8-exl2.

Rank the Newton 7B 8.0bpw H8 EXL2 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? 40248 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241217