DEV Community

#quantization

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Unsloth Releases Qwen3.6-27B-NVFP4: Enhanced Throughput and Agentic Coding for Developers

Unsloth Releases Qwen3.6-27B-NVFP4: Enhanced Throughput and Agentic Coding for Developers

Comments
3 min read
Bonsai-27B: A 1-Bit LLM for On-Device Inference with Llama.cpp and MLX

Bonsai-27B: A 1-Bit LLM for On-Device Inference with Llama.cpp and MLX

Comments
3 min read
Quantizing MedGemma to INT4 (GPTQ/W4A16): Everything That Broke Along the Way

Quantizing MedGemma to INT4 (GPTQ/W4A16): Everything That Broke Along the Way

Comments
6 min read
Self-Hosting a Model Means Self-Hosting Its Evaluation Too

Self-Hosting a Model Means Self-Hosting Its Evaluation Too

1
Comments 2
5 min read
Gemma 4 QAT on a 1080 Ti: What 'Quantization-Aware' Actually Buys — and Fitting the 12B on 8 GB at 16k

Gemma 4 QAT on a 1080 Ti: What 'Quantization-Aware' Actually Buys — and Fitting the 12B on 8 GB at 16k

Comments
5 min read
Quantization formats compared: GGUF vs GPTQ vs AWQ vs NF4

Quantization formats compared: GGUF vs GPTQ vs AWQ vs NF4

Comments
7 min read
INT8 Q/DQ Calibration on Blackwell: 1.8 the TRT 10 + FP16 Baseline

INT8 Q/DQ Calibration on Blackwell: 1.8 the TRT 10 + FP16 Baseline

Comments
7 min read
LLM Quantization Levels Compared: Q4_K_M vs Q8_0 vs FP16 [2026]

LLM Quantization Levels Compared: Q4_K_M vs Q8_0 vs FP16 [2026]

Comments 1
15 min read
How to Pick a GGUF Quant Level for Your VRAM Budget

How to Pick a GGUF Quant Level for Your VRAM Budget

Comments
4 min read
Why your quantized LLM loses its MTP heads and how to keep them

Why your quantized LLM loses its MTP heads and how to keep them

1
Comments
5 min read
The Best Result This Week Was a Failed Prediction — Phase-3a Doesn't Transfer

The Best Result This Week Was a Failed Prediction — Phase-3a Doesn't Transfer

Comments
1 min read
Two Localizers, Both Wrong: Bounding a Quantization Cost That Wouldn't Close

Two Localizers, Both Wrong: Bounding a Quantization Cost That Wouldn't Close

Comments
1 min read
When the Sensitivity Metric Lies: A Drift-Inversion Smoking Gun in Mixed-Precision LLM Quantization

When the Sensitivity Metric Lies: A Drift-Inversion Smoking Gun in Mixed-Precision LLM Quantization

Comments
8 min read
GGUF Quantization Explained: Q4_K_M vs Q5_K_M vs Q8 — Which to Pick (2026)

GGUF Quantization Explained: Q4_K_M vs Q5_K_M vs Q8 — Which to Pick (2026)

Comments
5 min read
1-bit, 545 megabytes, zero API keys — local AI that beats GPT-5.4

1-bit, 545 megabytes, zero API keys — local AI that beats GPT-5.4

2
Comments 1
2 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.