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

BioMistralMerged by skfrost19

 »  All LLMs  »  skfrost19  »  BioMistralMerged   URL Share it on

  3922   
  Merged Model   Autotrain compatible Base model:biomistral/biomistr... Base model:quantized:biomistra...   Conversational   Endpoints compatible   Gguf   Instruct   Mistral   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/skfrost19/BioMistralMerged 

BioMistralMerged Benchmarks

ARC: 62.46 vs 96.7 (so35)-35.4%
HellaSwag: 83.97 vs 95.3 (gpt4)-11.9%
MMLU: 60.06 vs 88.3 (so35)-32%
TruthfulQA: 60.19 vs 59 (gpt4)2%
WinoGrande: 77.9 vs 87.5 (gpt4)-11%
GSM8K: 0.08 vs 96.4 (so35)-99.9%
LLME Score: 0.25533
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
BioMistralMerged (skfrost19/BioMistralMerged)

BioMistralMerged Parameters and Internals

Model Type 
pre-trained language model, merged model
Additional Notes 
This model is a merge of two pre-trained language models using the SLERP method.
Training Details 
Methodology:
The model was merged using the SLERP merge method using mergekit, involving configurations across layers from two source models.
LLM NameBioMistralMerged
Repository 🤗https://huggingface.co/skfrost19/BioMistralMerged 
Base Model(s)  mohsenfayyaz/Mistral-7B-Instruct-v0.2_medical_bios_5000_5ep   BioMistral 7B   mohsenfayyaz/Mistral-7B-Instruct-v0.2_medical_bios_5000_5ep   BioMistral/BioMistral-7B
Merged ModelYes
Model Size7b
Required VRAM14.4 GB
Updated2024-12-22
Maintainerskfrost19
Model Typemistral
Instruction-BasedYes
Model Files  7.7 GB   2.0 GB: 1-of-8   1.9 GB: 2-of-8   2.0 GB: 3-of-8   2.0 GB: 4-of-8   1.9 GB: 5-of-8   1.9 GB: 6-of-8   1.9 GB: 7-of-8   0.8 GB: 8-of-8
GGUF QuantizationYes
Quantization Typegguf
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.40.0
Tokenizer ClassLlamaTokenizer
Vocabulary Size32000
Torch Data Typebfloat16

Best Alternatives to BioMistralMerged

Best Alternatives
Context / RAM
Downloads
Likes
Mistral 7B Instruct V0.2 GGUF32K / 4.4 GB1992
Mistral 7B Instruct V0.2.gguf32K / 14.5 GB203
...andle Mistral 7B Instruct V0.232K / 14.4 GB390
...tral 7B V0.3 Chinese Chat 8bit32K / 7.7 GB104573
Tamil Mistral 7B Instruct V0.132K / 14.8 GB25014
...stral 7B V0.3 Chinese Chat F1632K / 14.5 GB4711
...tral 7B V0.3 Chinese Chat 4bit32K / 4.1 GB1701
Mistral 7B Instruct V0.3 GGUF32K / 2.7 GB400
Mistral 7B Instruct V0.3 GGUF32K / 2.7 GB290
...7B Instruct V0.2 SFT Local 27632K / 14.4 GB100
Note: green Score (e.g. "73.2") means that the model is better than skfrost19/BioMistralMerged.

Rank the BioMistralMerged 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? 40066 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 v20241217
  Support LLM Explorer