Yes, cross-entropy works for recsys but it's pointwise, it optimizes per-item probability not the order. So Precision@K / Recall@K are fine metrics for that setup, but if ranking order matters you'll get better results with BPR or a softmax over the candidate set.
vloplok PRO
RDTvlokip
AI & ML interests
None yet
Recent Activity
commentedon their article 27 days ago
📊 Cross-Entropy — La fonction de perte qui SAIT punir ! 🎯🔥 posted an update about 2 months ago
🧠 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-scratchOrganizations
commented on 📊 Cross-Entropy — La fonction de perte qui SAIT punir ! 🎯🔥 27 days ago
reacted to Aurelien-Morgan's post with 🔥 about 2 months ago
Post
1095
@retrain-pipelines v0.2.0 is out !
I'm at Station F at My booth with GOSIM Paris 2026 today & tomorrow.
Come meet me for a live in-person demo and a chat !
I'm at Station F at My booth with GOSIM Paris 2026 today & tomorrow.
Come meet me for a live in-person demo and a chat !
- 1 reply
posted an update about 2 months ago
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
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
upvoted an article about 2 months ago
Article
🧠 I trained my own French LLM from scratch — alone, with a 1080 Ti, and the power went out ⚡🇫🇷
RDTvlokip
• • 6 published an article about 2 months ago
Article
🧠 I trained my own French LLM from scratch — alone, with a 1080 Ti, and the power went out ⚡🇫🇷
RDTvlokip
• • 6 upvoted an article about 2 months ago
Article
🧠 J'ai entraîné mon propre LLM français from scratch — seul, avec une 1080 Ti, et le courant a coupé ⚡🇫🇷
RDTvlokip
• • 2 published an article about 2 months ago
Article
🧠 J'ai entraîné mon propre LLM français from scratch — seul, avec une 1080 Ti, et le courant a coupé ⚡🇫🇷
RDTvlokip
• • 2 upvoted an article 4 months ago
Article
🧲 Embeddings — When AI turns words into GPS coordinates! 📍🧠
RDTvlokip
• • 1 published an article 4 months ago
Article
🧲 Embeddings — When AI turns words into GPS coordinates! 📍🧠
RDTvlokip
• • 1 upvoted an article 4 months ago
Article
🧲 Embeddings — Quand l'IA transforme les mots en coordonnées GPS ! 📍🧠
RDTvlokip
• • 1 published an article 4 months ago
Article
🧲 Embeddings — Quand l'IA transforme les mots en coordonnées GPS ! 📍🧠
RDTvlokip
• • 1 upvoted a changelog 4 months ago
Hugging Face Changelog
Public Storage Add-ons
• 169
upvoted an article 4 months ago
Article
🎯 PCA (Principal Component Analysis) — Compresser les dimensions comme un boss ! 📊🔥
RDTvlokip
• • 1 published an article 4 months ago
Article
🎯 PCA (Principal Component Analysis) — Compresser les dimensions comme un boss ! 📊🔥
RDTvlokip
• • 1 upvoted an article 4 months ago
Article
🎯 PCA (Principal Component Analysis) — Compressing dimensions like a boss! 📊🔥
RDTvlokip
• • 1 published an article 4 months ago
Article
🎯 PCA (Principal Component Analysis) — Compressing dimensions like a boss! 📊🔥
RDTvlokip
• • 1 upvoted an article 5 months ago
Article
🎯 K-Means — Quand l'IA organise le chaos en boîtes bien rangées ! 📦✨
RDTvlokip
• • 1 published an article 5 months ago
Article
🎯 K-Means — Quand l'IA organise le chaos en boîtes bien rangées ! 📦✨
RDTvlokip
• • 1 upvoted an article 5 months ago
Article
🎯 K-Means — When AI organizes chaos into neat boxes! 📦✨
RDTvlokip
• • 1 published an article 5 months ago
Article
🎯 K-Means — When AI organizes chaos into neat boxes! 📦✨
RDTvlokip
• • 1