๐Ÿ”ง Unsloth launched Studio โ€” faster, cheaper LLM fine-tuning for small teams.

Custom model training used to be a big-lab privilege.

Not anymore.

#LLM #FineTuning #MLOps #Unsloth #AI

Unsloth AI (@UnslothAI)

Unsloth Studio๊ฐ€ ํ•œ ์ค„์˜ ์ฝ”๋“œ๋กœ ์„ค์น˜๋˜๋„๋ก ๊ฐœ์„ ๋๊ณ , macOSยทWindowsยทLinux๋ฅผ ์ง€์›ํ•ฉ๋‹ˆ๋‹ค. ๋˜ํ•œ NVIDIA์˜ ์ƒˆ ๋ชจ๋ธ Nemotron 3 4B๋ฅผ ๋กœ์ปฌ์—์„œ 4GB RAM๋งŒ์œผ๋กœ ์‹คํ–‰ํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ์†Œ๊ฐœํ•ด, ๊ฒฝ๋Ÿ‰ ์˜จ๋””๋ฐ”์ด์Šค AI ํ™œ์šฉ ๊ฐ€๋Šฅ์„ฑ์„ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค.

https://x.com/UnslothAI/status/2034992903010066740

#unsloth #nvidia #nemotron #localai #llm

Unsloth AI (@UnslothAI)

Unsloth Studio๊ฐ€ ์ด์ œ uv๋กœ ์„ค์น˜ ๊ฐ€๋Šฅํ•ด์กŒ๊ณ , ์–ด๋–ค ํ™˜๊ฒฝ์—์„œ๋„ ์„ค์น˜๊ฐ€ ์ž‘๋™ํ•˜๋„๋ก ๊ฐœ์„ ๋๋‹ค. Docker ์ด๋ฏธ์ง€๋„ ์—…๋ฐ์ดํŠธ๋์œผ๋ฉฐ, ์—ฌ๋Ÿฌ ๋ฒ„๊ทธ ์ˆ˜์ •๊ณผ ์ƒˆ ๊ธฐ๋Šฅ์ด ํ•จ๊ป˜ ๋ฐฐํฌ๋๋‹ค. ํŠนํžˆ ํˆด ํ˜ธ์ถœ ์ •ํ™•๋„๊ฐ€ ์•ฝ 30% ํ–ฅ์ƒ๋˜์–ด ๊ฐœ๋ฐœ์ž์šฉ AI ๋„๊ตฌ๋กœ์„œ ์‹ค์šฉ์„ฑ์ด ํฌ๊ฒŒ ๋†’์•„์กŒ๋‹ค.

https://x.com/UnslothAI/status/2034363137722352066

#unsloth #studio #toolcalling #docker #opensource

Unsloth AI (@UnslothAI) on X

Unsloth Studio now installs via uv. Installation works in any environment. We also updated our Docker image. And shipped lots of bug fixes and new features including ~30% more accurate tool calling. Info: https://t.co/iIZWnQsVxQ

X (formerly Twitter)

Unsloth Studio(Beta)๋Š” ์˜คํ”ˆ์†Œ์Šค ๋…ธ์ฝ”๋“œ ๋กœ์ปฌ ์›น UI๋กœ, GGUFยทsafetensor ๋ชจ๋ธ์„ ๋กœ์ปฌ์—์„œ ์‹คํ–‰ยทํ•™์Šตยท๋‚ด๋ณด๋‚ด๊ธฐ๊นŒ์ง€ ํ•œ๊ณณ์—์„œ ์ง€์›ํ•ฉ๋‹ˆ๋‹ค. 500+ ๋ชจ๋ธ์„ VRAM 70% ์ ˆ๊ฐยท2๋ฐฐ ๋น ๋ฅด๊ฒŒ ํ•™์Šต(์ •ํ™•๋„ ์œ ์ง€), ํ…์ŠคํŠธยท๋น„์ „ยทTTSยท์ž„๋ฒ ๋”ฉ ์ง€์›. PDF/CSV/DOCX์—์„œ ์ž๋™์œผ๋กœ ๋ฐ์ดํ„ฐ์…‹ ์ƒ์„ฑ, ์…€ํ”„ํžˆ์ผ๋ง ๋„๊ตฌํ˜ธ์ถœยท์›น๊ฒ€์ƒ‰ยท์ฝ”๋“œ ์‹คํ–‰ยท์ž๋™ ํŠœ๋‹, ๋ชจ๋ธ์„ GGUF/safetensors๋กœ ๋‚ด๋ณด๋‚ด๊ธฐ ๊ฐ€๋Šฅ. ์˜คํ”„๋ผ์ธ ์šฐ์„  ์„ค๊ณ„๋กœ ๊ฐœ์ธ์ •๋ณด ๋ณดํ˜ธ, Windows/Mac/Linux/Docker/Colab์—์„œ ์‚ฌ์šฉ(๋ฒ ํƒ€).

https://unsloth.ai/docs/new/studio

#unsloth #localai #modeltraining #gguf #opensource

Introducing Unsloth Studio | Unsloth Documentation

Run and train AI models locally with Unsloth Studio.

Akshay (@akshay_pachaar)

Unsloth ๊ด€๋ จ GitHub ๋ฆฌํฌ์ง€ํ† ๋ฆฌ์™€ ๋ธ”๋กœ๊ทธ/๊ฐ€์ด๋“œ๋ฅผ ์•ˆ๋‚ดํ•˜๋Š” ๋งํฌ ํŠธ์œ—์ž…๋‹ˆ๋‹ค. ์˜คํ”ˆ์†Œ์Šค ์›น UI์˜ ์ฝ”๋“œ์™€ ์„ค์น˜/์‚ฌ์šฉ ๊ฐ€์ด๋“œ, ๋ฌธ์„œ๋ฅผ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋Š” ์ž๋ฃŒ ๋งํฌ๋ฅผ ์ œ๊ณตํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

https://x.com/akshay_pachaar/status/2034253794490638422

#unsloth #github #opensource #llm

Akshay ๐Ÿš€ (@akshay_pachaar) on X

GitHub: https://t.co/smDHEA1tI8 Blog and Guide: https://t.co/XFYwMmKV57

X (formerly Twitter)

Akshay (@akshay_pachaar)

Unsloth๊ฐ€ ์˜คํ”ˆ์†Œ์Šค ์›น UI๋ฅผ ์ถœ์‹œํ•˜์—ฌ ์ฝ”๋“œ๋ฅผ ์ž‘์„ฑํ•˜์ง€ ์•Š๊ณ  500๊ฐœ ์ด์ƒ์˜ LLM์„ ์‹คํ–‰ํ•˜๊ณ  ํŒŒ์ธํŠœ๋‹ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ๋‹ค๊ณ  ๋ฐœํ‘œํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฃผ์š” ํŠน์ง•์œผ๋กœ ๋กœ์ปฌ(๋งฅ/์œˆ๋„์šฐ/๋ฆฌ๋ˆ…์Šค) ์‹คํ–‰ ์ง€์›, ๋ชจ๋ธ ํ•™์Šต 2๋ฐฐ ์†๋„ ๋ฐ VRAM 70% ์ ˆ๊ฐ, GGUF ๋ฐ ๋น„์ „ ๋ชจ๋ธ ์ง€์› ๋“ฑ์ด ์–ธ๊ธ‰๋˜์–ด ํŒŒ์ธํŠœ๋‹ ์›Œํฌํ”Œ๋กœ์šฐ๋ฅผ ํฌ๊ฒŒ ๋‹จ์ˆœํ™”ํ•ฉ๋‹ˆ๋‹ค.

https://x.com/akshay_pachaar/status/2034253782444589498

#unsloth #finetuning #llm #opensource #gguf

Akshay ๐Ÿš€ (@akshay_pachaar) on X

finetuning LLMs will never be the same! Unsloth just launched an open-source web UI to run and fine-tune 500+ LLMs without writing any code. key features - run models locally on Mac, Windows, Linux - train 500+ models 2x faster with 70% less VRAM - supports GGUF, vision,

X (formerly Twitter)

Johan Puisais (@mrgris)

Unsloth Studio๊ฐ€ ์ด์ œ Pinokio์—์„œ ์‚ฌ์šฉ ๊ฐ€๋Šฅํ•˜๋‹ค๋Š” ๋ฐœํ‘œ์ž…๋‹ˆ๋‹ค. ๋กœ์ปฌ์—์„œ ๋ชจ๋ธ์„ ๋ฏธ์„ธ์กฐ์ •(fine-tune)ํ•˜๋Š” ๊ณผ์ •์„ ๊ฐ„์†Œํ™”ํ•˜๋ฉฐ, ์ƒˆ ๋Ÿฐ์ฒ˜๊ฐ€ ์„ค์น˜๋ฅผ ์ž๋™ํ™”ํ•ด ์‚ฌ์šฉ์ž์˜ ๋จธ์‹ ์„ 'AI ๋นŒ๋“œ ์Šคํ…Œ์ด์…˜'์œผ๋กœ ๋ณ€ํ™˜ํ•ด ์ค€๋‹ค๊ณ  ์†Œ๊ฐœํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

https://x.com/mrgris/status/2034024733826580563

#unsloth #pinokio #finetuning #devtools

Johan Puisais ๐Ÿ‡บ๐Ÿ‡ฆ ๐Ÿ‡ช๐Ÿ‡บ ๐Ÿ‡ซ๐Ÿ‡ท (@mrgris) on X

โšก๏ธ Unsloth Studio est maintenant disponible sur Pinokio ! Finer-tuner vos modรจles localement n'a jamais รฉtรฉ aussi simple. Le nouveau launcher Unsloth Studio automatise toute l'installation pour transformer votre machine en station de build IA. ๐Ÿฆฅ๐Ÿ’ป ๐Ÿ› ๏ธ Ce que fait le launcher :

X (formerly Twitter)
Unsloth AI releases Studio, a no-code local interface for high-performance LLM fine-tuning. The open-source tool claims 70% less VRAM usage through hand-written Triton kernels, enabling developers to fine-tune models like Llama 3.1 and DeepSeek-R1 on consumer GPUs. https://www.marktechpost.com/2026/03/17/unsloth-ai-releases-studio-a-local-no-code-interface-for-high-performance-llm-fine-tuning-with-70-less-vram-usage/ #AIagent #AI #GenAI #AIInfrastructure #Unsloth
Unsloth AI Releases Unsloth Studio: A Local No-Code Interface For High-Performance LLM Fine-Tuning With 70% Less VRAM Usage

Unsloth AI Releases Unsloth Studio: A Local No-Code Interface For High-Performance LLM Fine-Tuning With 70% Less VRAM Usage

MarkTechPost
Introducing Unsloth Studio | Unsloth Documentation

Run and train AI models locally with Unsloth Studio.

้‡‘ใฎใƒ‹ใƒฏใƒˆใƒช (@gosrum)

unsloth๊ฐ€ ๊ณต๊ฐœํ•œ GGUF ํŒŒ์ผ(๋ชจ๋ธ ํฌ๋งท)์ด ์˜ฌ๋ผ์™€์„œ ์ž‘์„ฑ์ž๊ฐ€ ํ˜„์žฌ ๋‹ค์šด๋กœ๋“œ ์ค‘์ด๋ผ๊ณ  ์•Œ๋ฆผ. ์†Œํ˜•/๋น„๊ณต์‹ ๋ชจ๋ธ ํฌ๋งท ๋ฐฐํฌ๋‚˜ ๋ณ€ํ™˜๋ณธ ์ถœ์‹œ์— ๋Œ€ํ•œ ์งง์€ ๊ณต์ง€๋กœ ๋ณด์ž„.

https://x.com/gosrum/status/2033821384225525768

#gguf #model #download #unsloth

้‡‘ใฎใƒ‹ใƒฏใƒˆใƒช (@gosrum) on X

unslothใฎGGUFใŒๆฅใฆใŸใฎใงใƒ€ใ‚ฆใƒณใƒญใƒผใƒ‰ไธญ... https://t.co/xbSua4spjG

X (formerly Twitter)