prism-ml/Ternary-Bonsai-2-27B-gguf-dev
Text Generation • 27B • Updated • 7.2k • 32
Erik Scholz
Green-Sky
AI & ML interests
None yet
Recent Activity
liked a model about 18 hours ago
prism-ml/Ternary-Bonsai-2-27B-gguf-dev liked a model 1 day ago
WaveCut/Turbo-Image-2.1 liked a model 1 day ago
madebyollin/texture-fix-vae-for-qwen-image-2.1Organizations
liked a
model about 18 hours ago
liked 3
models 1 day ago
reacted to danielhanchen's post with 🔥 3 days ago
liked a
model 4 days ago
One Humble Suggestions
#1 opened 7 days ago
by Green-Sky
liked 2
models 7 days ago
Request: a Q2_0 group-64 (mainline-compatible) export for Bonsai 2 27B?
➕👀 2
6
#12 opened 12 days ago
by 
AlexGabbia
liked 2
models 9 days ago
liked a
model 10 days ago
liked a
model 12 days ago
upvoted a collection 12 days ago
🚩 Report: Spam
👍 6
6
#2 opened 13 days ago
by 
spanspek
liked a
model 13 days ago
reacted to eaddario's post with 🔥 15 days ago
Post
136
Experimental global target bits‑per‑weight quantization of XHToken/Spark-X2.5-1.7B and XHToken/Spark-X2.5-4B.
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/Spark-X2.5-1.7B-GGUF
eaddario/Spark-X2.5-4B-GGUF
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/Spark-X2.5-1.7B-GGUF
eaddario/Spark-X2.5-4B-GGUF
WaveCut/Qwen3.8-Flash-Next-REAM-288
1
#2914 opened 16 days ago
by Green-Sky
liked a
model 16 days ago