BAAR Logo

BAAR2-95M

🌐 Official Website: https://baar.uk/model/
πŸ“„ License: Non-Commercial Only (Contact for Commercial Pricing)


πŸ“Œ Model Overview

BAAR2-95M is an ultra-compact multilingual model designed to enable real-time chat and generation under minimal hardware conditions. With just 95 million (95M) parameters, it brings lightweight, interactive dialogue capabilities across 11 languages to resource-constrained environments, embedded systems, and edge on-device platforms where standard LLMs cannot operate.

BAAR2-95M은 μ΅œμ†Œν•œμ˜ ν•˜λ“œμ›¨μ–΄ 여건과 극단적인 λ¦¬μ†ŒμŠ€ μ œμ•½ ν™˜κ²½μ—μ„œλ„ 11개 μ–Έμ–΄λ‘œ μ‹€μ‹œκ°„ ν…μŠ€νŠΈ 생성 및 λŒ€ν™” μ²˜λ¦¬κ°€ κ°€λŠ₯ν•˜λ„λ‘ μ„€κ³„λœ 9,500만(95M) νŒŒλΌλ―Έν„° 규λͺ¨μ˜ μ΄ˆκ²½λŸ‰ λ‹€κ΅­μ–΄ μ–Έμ–΄ λͺ¨λΈμž…λ‹ˆλ‹€. 일반적인 λŒ€ν˜• λͺ¨λΈ ꡬ동이 λΆˆκ°€λŠ₯ν•œ μ΄ˆμ €μ‚¬μ–‘ μ—£μ§€ 및 μž„λ² λ””λ“œ κΈ°κΈ°μ—μ„œλ„ 즉각적인 μΆ”λ‘ κ³Ό μ΄ˆκ²½λŸ‰ μ‹€μ‹œκ°„ RAG μ‹œμŠ€ν…œ ꡬ좕을 μ§€μ›ν•©λ‹ˆλ‹€.


πŸ” Retrieval-Augmented Generation (RAG) & Edge AI

BAAR2-95M is optimized for low-latency vector retrieval integration and context-based response generation, ensuring sub-second response times even under strict compute and memory budgets.

μ‹€μ‹œκ°„ 벑터 검색(Retrieval)κ³Ό BAAR2-95M μ΄ˆκ²½λŸ‰ 생성 엔진을 κ²°ν•©ν•˜μ—¬ κ²½λŸ‰ RAG νŒŒμ΄ν”„λΌμΈμ„ ꡬ좕할 수 있으며, μ™ΈλΆ€ λ¬Έλ§₯(Context)을 전달받아 μ„œλΈŒ-μ„Έμ»¨λ“œ(Sub-second) λ‹¨μœ„μ˜ λΉ λ₯Έ λ°˜μ‘ 속도λ₯Ό μ œκ³΅ν•©λ‹ˆλ‹€.


πŸ“„ License & Commercial Terms (λΌμ΄μ„ μŠ€ 및 상업적 λ„μž… μ•ˆλ‚΄)

  • Non-Commercial Use Only (비상업적 μ—°κ΅¬Β·κ°œμΈ 이용 ν•œμ •):
    This model is strictly provided for non-commercial, research, and educational purposes only. Any commercial use without a separate commercial agreement is strictly prohibited.
    λ³Έ λͺ¨λΈμ€ 순수 연ꡬ, ꡐ윑 및 개인 λΉ„μ˜λ¦¬ λͺ©μ μ— ν•œν•΄μ„œλ§Œ 무료둜 μ œκ³΅λ©λ‹ˆλ‹€. 사전 ν—ˆκ°€ μ—†λŠ” 일체의 영리적 μ΄μš©μ€ μ—„κ²©νžˆ κΈˆμ§€λ©λ‹ˆλ‹€.

  • Commercial Licensing & Inquiries for Revenue-Generating Entities (맀좜 λ°œμƒ κΈ°μ—… 및 μƒμš©ν™” λ„μž… 문의):
    If your organization generates revenue, or if you plan to incorporate this model into commercial products, services, or paid APIs, you are required to purchase a commercial license. Please contact us directly via email for enterprise licensing and integration support.
    κΈ°μ—…/사업체에 맀좜이 λ°œμƒν•˜κ³  μžˆκ±°λ‚˜, λ³Έ λͺ¨λΈμ„ μƒμš© μ„œλΉ„μŠ€Β·μœ λ£Œ μ œν’ˆΒ·μˆ˜μ΅ 창좜 λͺ©μ μ˜ μ‹œμŠ€ν…œμ— λ„μž…ν•˜κ³ μž ν•˜μ‹œλŠ” 경우 λ°˜λ“œμ‹œ λ³„λ„μ˜ μƒμš© λΌμ΄μ„ μŠ€λ₯Ό μ·¨λ“ν•˜μ…”μ•Ό ν•©λ‹ˆλ‹€. λ„μž… 쑰건 및 κΈ°μ—…μš© λΌμ΄μ„ μŠ€ λ°œκΈ‰μ€ μ•„λž˜ μ΄λ©”μΌλ‘œ λ¬Έμ˜ν•΄ μ£Όμ‹œκΈ° λ°”λžλ‹ˆλ‹€.

    • λ„μž… 및 μƒμš© λΌμ΄μ„ μŠ€ 문의 (Commercial Inquiries): admin@099.kr
    • 곡식 μ›Ήμ‚¬μ΄νŠΈ (Official Website): https://baar.uk/model/

⚠️ Notes (μ•ˆλ‚΄ 사항)

  • Pre-trained Model: This model is a compact base pre-trained checkpoint. For specialized tasks, factual precision, or task completion, task-specific fine-tuning (SFT) or Retrieval-Augmented Generation (RAG) pipelines are recommended.
  • 기초 μ‚¬μ „ν•™μŠ΅ λͺ¨λΈ: λ³Έ λͺ¨λΈμ€ 95M 규λͺ¨μ˜ 베이슀 μ‚¬μ „ν•™μŠ΅ μ²΄ν¬ν¬μΈνŠΈμž…λ‹ˆλ‹€. μ •κ΅ν•œ μž‘μ—… μˆ˜ν–‰, 사싀 관계 검증 및 완성도 높은 ν…μŠ€νŠΈ 생성이 ν•„μš”ν•œ 경우 도메인 νŒŒμΈνŠœλ‹(SFT) λ˜λŠ” 검색 증강 생성(RAG) νŒŒμ΄ν”„λΌμΈ 결합을 ꢌμž₯ν•©λ‹ˆλ‹€.

πŸ’Ό Opportunities & Contact (μ±„μš© μ œμ•ˆ 및 투자 문의)

This project demonstrates practical competency in designing custom micro-scale architectures, end-to-end multi-language pre-training, and ultra-low-latency deployment under extreme compute constraints.

I am actively seeking AI Engineering / Research opportunities, recruitment offers, and technical partnerships.

  • 곡식 μ›Ήμ‚¬μ΄νŠΈ (Official Website): https://baar.uk/model/
  • κΈ°μ—… λ„μž… 및 λΌμ΄μ„ μŠ€ 문의 (Commercial Adoption): admin@099.kr
  • μ±„μš© 및 ν¬μ§€μ…˜ μ œμ•ˆ (Recruitment & Hiring): admin@099.kr
  • 투자 및 기술 ν˜‘λ ₯ 문의 (Investment & Partnerships): admin@099.kr

극단적인 저사양 λ¦¬μ†ŒμŠ€ ν™˜κ²½μ—μ„œλ„ 즉각 μž‘λ™ν•˜λŠ” 고효율 λ‹€κ΅­μ–΄ λͺ¨λΈ μ•„ν‚€ν…μ²˜λ₯Ό μ„€κ³„ν•˜κ³ , 데이터 μ •μ œλΆ€ν„° ν•™μŠ΅Β·μ„œλΉ™κΉŒμ§€ μ „ 과정을 직접 κ΅¬ν˜„ν•˜λŠ” AI μ—”μ§€λ‹ˆμ–΄μž…λ‹ˆλ‹€. 기술 λ„μž…, μ±„μš© μ œμ•ˆ 및 ν˜‘μ—…μ— 관심 μžˆλŠ” κΈ°μ—…κ³Ό νŒ€μ˜ 문의λ₯Ό ν™˜μ˜ν•©λ‹ˆλ‹€.

Downloads last month
-
Safetensors
Model size
95.6M params
Tensor type
F16
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support