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

BigWeave V15 103B by llmixer

 »  All LLMs  »  llmixer  »  BigWeave V15 103B   URL Share it on

  3   
  103b   Autotrain compatible   Conversational   En   Endpoints compatible   Frankenmerge   Llama   Model-index   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/llmixer/BigWeave-v15-103b 

BigWeave V15 103B Benchmarks

ARC: 69.71 vs 96.7 (so35)-27.9%
HellaSwag: 86.41 vs 95.3 (gpt4)-9.3%
MMLU: 71.25 vs 88.3 (so35)-19.3%
TruthfulQA: 66.1 vs 59 (gpt4)12%
WinoGrande: 80.35 vs 87.5 (gpt4)-8.2%
GSM8K: 56.18 vs 96.4 (so35)-41.7%
LLME Score: 0.16333
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
BigWeave V15 103B (llmixer/BigWeave-v15-103b)

BigWeave V15 103B Parameters and Internals

Model Type 
text-generation
Use Cases 
Areas:
research, commercial applications
Additional Notes 
The version number tracks attempts, not a quality indicator. Only results with good performance are shared.
Training Details 
Methodology:
Self-merge of 152334H/miqu-1-70b-sf, exl2 measurements to identify relevant layers which are then duplicated in pairs.
Input Output 
Input Format:
Prompting Format: Mistral, Vicuna and Alpaca.
LLM NameBigWeave V15 103B
Repository 🤗https://huggingface.co/llmixer/BigWeave-v15-103b 
Model Size103b
Required VRAM206.6 GB
Updated2025-04-19
Maintainerllmixer
Model Typellama
Model Files  9.6 GB: 1-of-22   9.8 GB: 2-of-22   9.8 GB: 3-of-22   9.9 GB: 4-of-22   9.8 GB: 5-of-22   10.0 GB: 6-of-22   9.8 GB: 7-of-22   10.0 GB: 8-of-22   9.8 GB: 9-of-22   9.8 GB: 10-of-22   9.8 GB: 11-of-22   9.8 GB: 12-of-22   9.8 GB: 13-of-22   10.0 GB: 14-of-22   9.7 GB: 15-of-22   9.9 GB: 16-of-22   9.6 GB: 17-of-22   10.0 GB: 18-of-22   9.8 GB: 19-of-22   9.7 GB: 20-of-22   9.9 GB: 21-of-22   0.3 GB: 22-of-22
Supported Languagesen
Model ArchitectureLlamaForCausalLM
Licenseunknown
Context Length32764
Model Max Length32764
Transformers Version4.36.2
Tokenizer ClassLlamaTokenizer
Padding Token<unk>
Vocabulary Size32000
Torch Data Typefloat16

Best Alternatives to BigWeave V15 103B

Best Alternatives
Context / RAM
Downloads
Likes
Dark Miqu 103B31K / 206.4 GB302
Midnight Miqu 103B V1.531K / 206.4 GB16918
Midnight Miqu 103B V1.031K / 206.4 GB3713
OmegaDolphin 103B V0.14K / 175.6 GB221
Midnight Rose 103B V2.0.34K / 206.6 GB119
KitchenSink 103b4K / 178.6 GB29
Midnight Rose 103B V1.04K / 206.6 GB163
Aurora Nights 103B V1.04K / 206.6 GB514
Rogue Rose 103B V0.24K / 206.5 GB3631
Lila 103B L24K / 206.2 GB00
Note: green Score (e.g. "73.2") means that the model is better than llmixer/BigWeave-v15-103b.

Rank the BigWeave V15 103B 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? 46490 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