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

Bloomz 560M Sft Chat by cmarkea

 »  All LLMs  »  cmarkea  »  Bloomz 560M Sft Chat   URL Share it on

  945   

Bloomz 560M Sft Chat is an open-source language model by cmarkea. Features: 560m LLM, VRAM: 1.1GB, License: bigscience-bloom-rail-1.0, HF Score: 30.7, LLM Explorer Score: 0.15, Arc: 27.5, HellaSwag: 37.1, MMLU: 23.9, TruthfulQA: 42.4, WinoGrande: 53.5.

  Arxiv:2001.09977   Arxiv:2012.15613   Bloom Dataset:ehartford/wizard vicun...   Dataset:laion/oig   Dataset:shahules786/orca-chat Dataset:timdettmers/openassist...   En   Endpoints compatible   Fr   Pytorch   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/cmarkea/bloomz-560m-sft-chat 

Bloomz 560M Sft Chat Benchmarks

ARC: 27.47 vs 96.7 (so35)-71.6%
HellaSwag: 37.05 vs 95.3 (gpt4)-61.1%
MMLU: 23.93 vs 88.3 (so35)-72.9%
TruthfulQA: 42.35 vs 59 (gpt4)-28.2%
WinoGrande: 53.51 vs 87.5 (gpt4)-38.8%
LLME Score: 0.15109
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Bloomz 560M Sft Chat (cmarkea/bloomz-560m-sft-chat)
🌟 Advertise your project 🚀

Bloomz 560M Sft Chat Parameters and Internals

Model Type 
text generation
Use Cases 
Areas:
research, commercial applications
Primary Use Cases:
chatbot applications
Limitations:
Performance not guaranteed in languages other than French and English, Possibly reduced multilingual capabilities due to transposition from float16 to bfloat16
Considerations:
It's been tuned in a chatbot context and performs well on zero-shot tasks in English and French.
Additional Notes 
Performance degradation anticipated in non-supported languages.
Supported Languages 
French (High proficiency), English (High proficiency)
Training Details 
Data Sources:
ehartford/wizard_vicuna_70k_unfiltered, shahules7876/orca-chat, timdettmers/openassistant-guanaco, laion/OIG
Data Volume:
0.9 billion tokens
Methodology:
fine-tuned for chatbot applications from BigScience BLOOMZ-560m
Training Time:
41 hours
Hardware Used:
1 x A100 40GB
Model Architecture:
Transposition from float16 to bfloat16
Input Output 
Input Format:
Individual's prompt preceded by the EOS token (</s>).
Accepted Modalities:
text
Output Format:
Generated responses begin with BOS token (<s>).
Release Notes 
Version:
2023
Date:
October 12, 2023
Notes:
Release of bloomz-560m-sft-chat model, optimized for chat and instruction-based tasks in French and English.
LLM NameBloomz 560M Sft Chat
Repository 🤗https://huggingface.co/cmarkea/bloomz-560m-sft-chat 
Model Size560m
Required VRAM1.1 GB
Updated2026-03-19
Maintainercmarkea
Model Typebloom
Model Files  1.1 GB   1.1 GB
Supported Languagesfr en
Model ArchitectureBloomForCausalLM
Licensebigscience-bloom-rail-1.0
Transformers Version4.31.0
Tokenizer ClassBloomTokenizer
Padding Token<pad>
Vocabulary Size250880
Torch Data Typebfloat16

Best Alternatives to Bloomz 560M Sft Chat

Best Alternatives
Context / RAM
Downloads
Likes
Bloomz 560M0K / 1.1 GB1100330137
Train Test Bloom5600K / 2.2 GB50
Bloom 560M0K / 1.1 GB176740371
Promt Generator0K / 2.2 GB140942
Bloom 560M RLHF0K / 1.1 GB10571
Bloom 560M RLHF V20K / 1.1 GB9023
Train Test0K / 2.2 GB300
Guitester0K / 2.2 GB50
Product Description Fr0K / 2.2 GB50
Gogpt 560M0K / 1 GB13821
Note: green Score (e.g. "73.2") means that the model is better than cmarkea/bloomz-560m-sft-chat.

Rank the Bloomz 560M Sft Chat 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? 52721 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.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a
  Support LLM Explorer