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

Alfred 40B 1023 GPTQ by TheBloke

 »  All LLMs  »  TheBloke  »  Alfred 40B 1023 GPTQ   URL Share it on

  14   
  Arxiv:2306.15595   Arxiv:2307.03172   Arxiv:2309.00071   4-bit   Autotrain compatible Base model:lightonai/alfred-40... Base model:quantized:lightonai...   Custom code   Dataset:ehartford/dolphin   Dataset:openassistant/oasst1   Dataset:tau/sled Dataset:tiiuae/falcon-refinedw...   De   En   Es   Falcon   Falcon-40b   Fr   Gptq   It   Long-context   Ntk-yarn   Quantized   Refinedweb   Region:us   Safetensors   Yarn
Model Card on HF 🤗: https://huggingface.co/TheBloke/alfred-40B-1023-GPTQ 

Alfred 40B 1023 GPTQ Benchmarks

LLME Score: 0.12372
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Alfred 40B 1023 GPTQ (TheBloke/alfred-40B-1023-GPTQ)
🌟 Advertise your project 🚀

Alfred 40B 1023 GPTQ Parameters and Internals

Model Type 
causal decoder-only
Use Cases 
Areas:
research, commercial applications
Applications:
chat models, instruction following
Primary Use Cases:
chat and instruct implementations
Limitations:
Not suitable for unsupported languages, carries potential biases due to training data.
Considerations:
Users are advised to assess risks and implement mitigation strategies for production deployments.
Additional Notes 
NTK-YaRN method improves extended context capabilities. Ongoing improvements planned for identified rare failure modes.
Supported Languages 
en (advanced), fr (advanced), de (advanced), es (advanced), it (intermediate), pt (limited), pl (limited), nl (limited), ro (limited), cz (limited), sv (limited)
Training Details 
Data Sources:
OpenAssistant/oasst1, ehartford/dolphin, tau/sled, tiiuae/falcon-refinedweb, internal, internal-long-context
Methodology:
Supervised fine-tuning with NTK-YaRN for extended context length, chat-specific tokens.
Context Length:
8192
Hardware Used:
128 A100 40GB GPUs
Model Architecture:
Falcon architecture with extended context handling through NTK-YaRN.
Responsible Ai Considerations 
Fairness:
Trained on diverse datasets but may inherit common online stereotypes and biases.
Mitigation Strategies:
Users are recommended to implement guardrails to prevent misuse.
Input Output 
Input Format:
Prompted with: 'You are Alfred...{user query}'
Accepted Modalities:
text
Output Format:
Textual responses
Performance Tips:
Use the included prompt template for optimal performance in chat or instruct mode.
LLM NameAlfred 40B 1023 GPTQ
Repository 🤗https://huggingface.co/TheBloke/alfred-40B-1023-GPTQ 
Model NameAlfred 40B 1023
Model CreatorLightOn AI
Base Model(s)  Alfred 40B 1023   lightonai/alfred-40b-1023
Model Size40b
Required VRAM22.5 GB
Updated2025-08-21
MaintainerTheBloke
Model TypeRefinedWeb
Model Files  22.5 GB
Supported Languagesen fr de es it
GPTQ QuantizationYes
Quantization Typegptq
Model ArchitectureRWForCausalLM
Licenseapache-2.0
Model Max Length8192
Transformers Version4.35.0
Is Biased0
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size65024
Torch Data Typebfloat16

Best Alternatives to Alfred 40B 1023 GPTQ

Best Alternatives
Context / RAM
Downloads
Likes
FalconLite0K / 22.3 GB349170
FalconLite0K / 22.3 GB284170
Falcon 40B Instruct GPTQ0K / 22.5 GB323197
Falcon 40B Gptq0K / 23.3 GB242
...st1 En 2048 Falcon 40B V2 GPTQ0K / 22.5 GB248
Falcon 40B Gptq0K / 23.4 GB5113
Falcon 40B Instruct GPTQ0K / 22.5 GB141
...truct GPTQ Inference Endpoints0K / 22.5 GB152
Falcon 40B 8bit0K / 41.8 GB131
...dLM Uncensored Falcon 40B GPTQ0K / 22.5 GB3660

Rank the Alfred 40B 1023 GPTQ 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? 50804 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 v20241124
  Support LLM Explorer