Stan Jourdan

stanjourdan's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

liked a model about 14 hours ago

unsloth/Qwen3.6-27B-UD-MLX-MXFP4

Image-Text-to-Text • 10B • Updated Apr 23 • 1.56k • 7

liked 2 models about 15 hours ago

unsloth/Qwen3.6-35B-A3B-NVFP4

Image-Text-to-Text • 22B • Updated 28 days ago • 168k • 45

nvidia/Qwen3.6-35B-A3B-NVFP4

Text Generation • 19B • Updated 15 days ago • 5.02M • 367

liked a model 6 days ago

tvall43/Qwen3.5-14B-A3B-Claude-4.6-Opus-Reasoning-Distilled-reap-gguf

Text Generation • 14B • Updated Mar 9 • 5.58k • 48

New activity in unsloth/gemma-4-12B-it-qat-GGUF 7 days ago

Degenerate loops

2

#7 opened 15 days ago by

sverinn

liked a model 9 days ago

unsloth/Qwen3.5-2B-MTP-GGUF

Image-Text-to-Text • 2B • Updated May 16 • 23.6k • 30

liked 5 models 10 days ago

unsloth/DeepSeek-R1-Distill-Qwen-14B-GGUF

15B • Updated Jan 25, 2025 • 58.7k • 119

unsloth/Magistral-Small-2509-GGUF

24B • Updated Sep 17, 2025 • 6.52k • 101

mistralai/Magistral-Small-2509

24B • Updated Feb 23 • 23.6k • 304

barozp/Qwen3.6-28B-REAP20-A3B-GGUF

Text Generation • 28B • Updated Apr 19 • 27.3k • 48

0xSero/Qwen3.6-28B

Text Generation • 28B • Updated 29 days ago • 1.21k • 34

liked a model 12 days ago

unsloth/DeepSeek-R1-0528-Qwen3-8B-GGUF

Text Generation • 8B • Updated Jun 16, 2025 • 85.1k • 423

liked 3 models 16 days ago

Jackrong/Qwopus3.5-9B-Coder-MTP-GGUF

Image-Text-to-Text • Updated 25 days ago • 254k • 194

Jackrong/Qwen3.5-9B-Python-Coder-GGUF

9B • Updated Mar 15 • 5.76k • 7

unsloth/gemma-4-26B-A4B-it-GGUF

Image-Text-to-Text • 25B • Updated 19 days ago • 1.46M • 917

liked a model 18 days ago

unsloth/Qwen3.6-35B-A3B-MTP-GGUF

Image-Text-to-Text • 36B • Updated May 20 • 830k • 595

liked a model 19 days ago

Nekochu/Llama-3.1-8B-french-DPO

8B • Updated Sep 25, 2024 • 14 • 2

reacted to RDTvlokip's post with 👍 19 days ago

view post

Post

135

🧠 I trained a French LLM from scratch. Alone. On a 1080 Ti. And honestly… it was a lot.

4 months building the dataset before even touching the model. Custom crawler, custom extractor, custom BPE tokenizer, everything from zero. Then the architecture — RoPE, RMSNorm, SwiGLU, Flash Attention. Then a 3-phase trainer. Then debugging a causal mask bug that made the model generate "ïsïsïs" for hours.

Then the power went out at epoch 10/18.

The checkpoint survived. The model learned form perfectly — grammar, markdown, structure. Substance? Still working on it. Honest conclusion.

Full write-up here 👇

🔗 https://huggingface.co/blog/RDTvlokip/i-trained-my-own-french-llm-from-scratch

See translation

👍

2

2

+

Reply

liked a dataset 19 days ago

RDTvlokip/InfiniQA

Viewer • Updated Jun 26, 2025 • 101k • 48 • 2

liked a model 19 days ago

OpenLLM-France/Lucie-7B-Instruct-v1.1-gguf

Text Generation • 7B • Updated Mar 18 • 123 • 6