Patrick Devaney

patrickbdevaney

7 15 336

AI & ML interests

mamba s4, mixtral 8x-7b, candle, pytorch

Recent Activity

liked a model about 19 hours ago

Altworld/Hemmingway-1

reacted to onekq's post with 👀 about 19 hours ago

My takes on Jev 1. Very likely a small model. You can certainly pretrain, but I would grab an existing base model, say Qwen 3 class 2. The new RL method is a breakthrough, classification doesn't need to align with human preferences 3. The new output is an overstatement. It's just a new LM head. Of course autoregressive decoding can be used for classification: it takes just a few tokens to express the output. Think twice: are you sure classification doesn't need few-shot, CoT, or reasoning? All of these depend on auto-regressiveness 4. It carves out a market already existing, which is now served by oversized LLMs (hence overpaid), e.g. LLM as judge, labeling 5. Jevons effect will kick in, promoting more modeling efforts for small budget teams. It might even accelerate RSI

reacted to eaddario's post with 🔥 6 days ago

Experimental global target bits‑per‑weight quantization of openbmb/MiniCPM5-1B and openbmb/MiniCPM5-2B. Unlike standard llama.cpp quantization that rely on fixed type heuristics (e.g., Q4_K_M), the Target BPW approach automatically optimizes per-tensor precision where it matters the most, and produces high quality models that meet a precise global size target. Key Advantages: - VRAM Maximization: Can generate high quality models sized exactly to fit hardware constraints (e.g., fitting the model into exactly 24GB VRAM). - Data-Driven Precision: Quantization mix is determined by actual weight error sensitivity rather than hardcoded rules, often yielding better PPL/KLD size trade-offs. Full benchmarks (PPL, KLD, ARC, GPQA, MMLU, etc.) and methodology in the model's card. https://huggingface.co/eaddario/MiniCPM5-1B-GGUF https://huggingface.co/eaddario/MiniCPM5-2B-GGUF

View all activity

Organizations

liked a model about 19 hours ago

Altworld/Hemmingway-1

Text Generation • 27B • Updated about 21 hours ago • 3.79k • 537

reacted to onekq's post with 👀 about 19 hours ago

Post

3101

My takes on Jev

1. Very likely a small model. You can certainly pretrain, but I would grab an existing base model, say Qwen 3 class

2. The new RL method is a breakthrough, classification doesn't need to align with human preferences

3. The new output is an overstatement. It's just a new LM head. Of course autoregressive decoding can be used for classification: it takes just a few tokens to express the output. Think twice: are you sure classification doesn't need few-shot, CoT, or reasoning? All of these depend on auto-regressiveness

4. It carves out a market already existing, which is now served by oversized LLMs (hence overpaid), e.g. LLM as judge, labeling

5. Jevons effect will kick in, promoting more modeling efforts for small budget teams. It might even accelerate RSI

6 replies

reacted to eaddario's post with 🔥 6 days ago

Post

1196

Experimental global target bits‑per‑weight quantization of openbmb/MiniCPM5-1B and openbmb/MiniCPM5-2B.

Unlike standard llama.cpp quantization that rely on fixed type heuristics (e.g., Q4_K_M), the Target BPW approach automatically optimizes per-tensor precision where it matters the most, and produces high quality models that meet a precise global size target.

Key Advantages:
- VRAM Maximization: Can generate high quality models sized exactly to fit hardware constraints (e.g., fitting the model into exactly 24GB VRAM).
- Data-Driven Precision: Quantization mix is determined by actual weight error sensitivity rather than hardcoded rules, often yielding better PPL/KLD size trade-offs.

Full benchmarks (PPL, KLD, ARC, GPQA, MMLU, etc.) and methodology in the model's card.

eaddario/MiniCPM5-1B-GGUF
eaddario/MiniCPM5-2B-GGUF

liked 3 models 11 days ago

PixelML/Qwen3.8-Flash-Next-NVFP4-DFlash

Text Generation • 0.5B • Updated 13 days ago • 848 • 4

nerkyor/Qwen3.8-27B-EfficientThink-Uncensored-K3-Opus5-Grok4.6-GPT5.6Sol-SFT-SimPO-DFlash2

Text Generation • 2B • Updated 11 days ago • 49.3k • 44

nerkyor/Qwen3.8-27B-EfficientThink-Uncensored-K3-Opus5-Grok4.6-GPT5.6Sol-SFT-SimPO-DFlash2-GGUF

2B • Updated 15 days ago • 89.1k • 79

reacted to GoktugD's post with 👍 14 days ago

Post

2450

🇹🇷 One of our small Turkish models quietly reached **500+ monthly downloads** on Hugging Face.

**Werea-TR-TextRestore — only 300M parameters.**

Its job is simple:

istanbulda hava cok guzel
→ İstanbul'da hava çok güzel.

A lightweight model for restoring Turkish text:
• diacritics
• punctuation
• casing
• corrupted text

**96.5% word accuracy** on real Turkish news sentences.

And it runs without sending your text to a cloud API.

🤗 Try the model:
Werea-co/Werea-TR-TextRestore

🇹🇷 Built in Türkiye. Open source.

If you're working on Turkish NLP, I'd love to hear what we should build next.

#TurkishNLP #HuggingFace #OpenSourceAI #NLP