John Smith's picture

In a Training Loop 🔄

John Smith PRO

John6666

·

John6666cat

AI & ML interests

None yet

Recent Activity

reacted to eaddario's post with ❤️ about 5 hours ago

Experimental global target bits‑per‑weight quantization of ServiceNow-AI/Apriel-1.6-15b-Thinker and zai-org/GLM-4.6V-Flash Unlike standard llama.cpp quantizations that rely on fixed type heuristics (e.g., Q4_K_M), the Target BPW approach optimizes per-tensor precision where it matters the most, and produces high quality models that meet a precise global file 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, MMLU, etc.) and methodology in the models' cards https://huggingface.co/eaddario/Apriel-1.6-15b-Thinker-GGUF https://huggingface.co/eaddario/GLM-4.6V-Flash-GGUF

upvoted a collection about 5 hours ago

Arabic Speech Datasets

reacted to MohamedRashad's post with ❤️ about 5 hours ago

I have update my https://huggingface.co/collections/MohamedRashad/arabic-speech-datasets with new datasets, making the full audio data more than 3000 hours of good arabic speech. Feel Free to use it in your new innovations, And happy new year!

View all activity

Organizations

upvoted a collection about 5 hours ago

Arabic Speech Datasets

Best Datasets for Arabic Speech Tasks • 14 items • Updated about 18 hours ago • 9

upvoted 2 collections 1 day ago

Gemma 3 Collection

Some fun things I've made on Gemma 3 • 6 items • Updated Apr 18 • 2

TaiPhone LLM

TaiPhone: A Phone-Scale LLM Rooted in Taiwanese Knowledge • 4 items • Updated Jun 9 • 1

upvoted a collection 2 days ago

MolAct

6 items • Updated 6 days ago • 5

upvoted a collection 3 days ago

Mira & company

47 items • Updated 6 days ago • 1

upvoted an article 3 days ago

Article

The Optimal Architecture for Small Language Models

3 days ago

•

52

upvoted 3 collections 3 days ago

Josiefied and Abliterated Models

Other models Josiefied and Abliterated • 3 items • Updated 13 days ago • 2

J.O.S.I.E.-R1

My (Gökdeniz Gülmez) first reasoning model fine-tuned on a custom distill dataset • 2 items • Updated 10 days ago • 1

AdaSearch

Models used in AdaSearch: https://arxiv.org/abs/2512.16883 • 16 items • Updated 9 days ago • 1

upvoted an article 3 days ago

Article

Beyond the Wrapper: Building High-Throughput Reasoning Agents with Async Kernels

3 days ago

•

1

upvoted 2 collections 4 days ago

InverseCoder

Models and datasets of paper "InverseCoder: Unleashing the Power of Instruction-Tuned Code LLMs with Inverse-Instruct". • 7 items • Updated Dec 12, 2024 • 3

CodeV-SVA

CodeV-SVA: Training Specialized LLMs for Hardware Assertion Generation via RTL-Grounded Bidirectional Data Synthesis • 5 items • Updated 13 days ago • 1

upvoted a collection 5 days ago

Openhands Trajectories

Dataset of 67,074 OpenHands trajectories collected with Qwen3-Coder-480B-A35B-Instruct and two RFT checkpoints trained on the data • 3 items • Updated 6 days ago • 6

upvoted an article 6 days ago

Article

Encoding the World's Medical Knowledge into 970K

6 days ago

•

14

upvoted 2 collections 7 days ago

Skyra

Skyra: AI-Generated Video Detection via Grounded Artifact Reasoning • 5 items • Updated 11 days ago • 2

heretic-models

4 items • Updated about 8 hours ago • 1

upvoted an article 7 days ago

Article

Streaming datasets: 100x More Efficient

+3

Oct 27

•

75

upvoted a collection 8 days ago

Qwen3-abliterated

32 items • Updated 7 days ago • 40

upvoted 2 changelogs 9 days ago

Changelog

Duplicate Datasets

26 days ago

• 86

Changelog

Set your primary organization on your profile

10 days ago

• 55