Noah Zelezny

TheDrainFlorist

hozinwang's profile picture

Chelal233's profile picture

dongnhdev's profile picture

NoahZelezny
NoahZelezny

AI & ML interests

None yet

Recent Activity

updated a model about 1 month ago

TheDrainFlorist/gemma-4-e4b-it-VQ-PLE

updated a model about 1 month ago

TheDrainFlorist/gemma-4-26b-a4b-it-VQ-6.2bpw

updated a model about 1 month ago

TheDrainFlorist/Qwen3.8-Flash-Next-VQ-5.5bpw

View all activity

Organizations

None yet

TheDrainFlorist 's collections 7

DeepSeek-V4-Flash — Vector Quantized (MLX)

Data-free VQ of DeepSeek-V4-Flash for Apple Silicon. One rung, 3.2 bpw, 6-bit MTP head. Runs on Knurlogic (macOS). Built with VQLab.

TheDrainFlorist/DeepSeek-V4-Flash-VQ-3.2bpw

Text Generation • 36B • Updated Oct 11 • 151
TheDrainFlorist/DeepSeek-V4-Flash-Vision-Exp-VQ-3.2bpw

Image-Text-to-Text • 37B • Updated Oct 11 • 256

Qwen3.8-Flash-Next — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.8-Flash-Next for Apple Silicon. Four rungs, 2.1-5.5 bpw, MTP sidecar. Needs mlx-lm PR #1788. Built with VQLab.

TheDrainFlorist/Qwen3.8-Flash-Next-VQ-2.1bpw

Text Generation • 24B • Updated Oct 11 • 610 • 5
TheDrainFlorist/Qwen3.8-Flash-Next-VQ-3.2bpw

Text Generation • 43B • Updated Oct 11 • 271 • 1
TheDrainFlorist/Qwen3.8-Flash-Next-VQ-4.4bpw

Text Generation • 89B • Updated Oct 11 • 335 • 2
TheDrainFlorist/Qwen3.8-Flash-Next-VQ-5.5bpw

Text Generation • 60B • Updated Oct 11 • 254

Gemma-4 — Vector Quantized (MLX)

Data-free VQ builds of gemma-4 for Apple Silicon: the 26B MoE and the e4b with VQ per-layer embeddings. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/gemma-4-26b-a4b-it-VQ-6.2bpw

Image-Text-to-Text • 7B • Updated Oct 11 • 115
TheDrainFlorist/gemma-4-e4b-it-VQ-PLE

Image-Text-to-Text • 6B • Updated Oct 11 • 77

Qwen3.5-397B-A17B — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.5-397B-A17B for Apple Silicon, from one 128 GB Mac up. Five rungs, 2.2-3.5 bpw. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/Qwen3.5-397B-A17B-VQ-2.2bpw

Text Generation • 63B • Updated Oct 11 • 188
TheDrainFlorist/Qwen3.5-397B-A17B-VQ-2.4bpw

Text Generation • 120B • Updated Oct 11 • 159
TheDrainFlorist/Qwen3.5-397B-A17B-VQ-2.6bpw

Text Generation • 59B • Updated Oct 11 • 180
TheDrainFlorist/Qwen3.5-397B-A17B-VQ-3.1bpw

Text Generation • 64B • Updated Oct 11 • 191

GLM-5.3-Flash-VQ — Vector Quantized (MLX)

Data-free VQ rungs of GLM-5.3-Flash for Apple Silicon. Three rungs, 2.7-3.6 bpw, MTP sidecars. Stock mlx-vlm. Built with VQLab.

TheDrainFlorist/GLM-5.3-Flash-VQ-2.7bpw

Image-Text-to-Text • 37B • Updated Oct 11 • 266
TheDrainFlorist/GLM-5.3-Flash-VQ-3.1bpw

Image-Text-to-Text • 40B • Updated Oct 11 • 226
TheDrainFlorist/GLM-5.3-Flash-VQ-3.6bpw

Image-Text-to-Text • 45B • Updated Oct 11 • 294

Qwen3.8-27B — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.8-27B for Apple Silicon. Three rungs, 3.9-4.8 bpw, vision included. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/Qwen3.8-27B-VQ-3.9bpw

Text Generation • 12B • Updated Oct 11 • 281
TheDrainFlorist/Qwen3.8-27B-VQ-4.5bpw

Text Generation • 19B • Updated Oct 11 • 145 • 1
TheDrainFlorist/Qwen3.8-27B-VQ-4.8bpw

Text Generation • 13B • Updated Oct 11 • 160

Qwen3.6-35B-A3B — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.6-35B-A3B for Apple Silicon, from 32 GB Macs up. Four rungs, 3.4-5.4 bpw. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/Qwen3.6-35B-A3B-VQ-3.4bpw

Text Generation • 6B • Updated Oct 11 • 242 • 1
TheDrainFlorist/Qwen3.6-35B-A3B-VQ-3.8bpw

Text Generation • 7B • Updated Oct 11 • 244
TheDrainFlorist/Qwen3.6-35B-A3B-VQ-4.6bpw

Text Generation • 7B • Updated Oct 11 • 191
TheDrainFlorist/Qwen3.6-35B-A3B-VQ-5.4bpw

Text Generation • 8B • Updated Oct 11 • 229

DeepSeek-V4-Flash — Vector Quantized (MLX)

Data-free VQ of DeepSeek-V4-Flash for Apple Silicon. One rung, 3.2 bpw, 6-bit MTP head. Runs on Knurlogic (macOS). Built with VQLab.

TheDrainFlorist/DeepSeek-V4-Flash-VQ-3.2bpw

Text Generation • 36B • Updated Oct 11 • 151
TheDrainFlorist/DeepSeek-V4-Flash-Vision-Exp-VQ-3.2bpw

Image-Text-to-Text • 37B • Updated Oct 11 • 256

GLM-5.3-Flash-VQ — Vector Quantized (MLX)

Data-free VQ rungs of GLM-5.3-Flash for Apple Silicon. Three rungs, 2.7-3.6 bpw, MTP sidecars. Stock mlx-vlm. Built with VQLab.

TheDrainFlorist/GLM-5.3-Flash-VQ-2.7bpw

Image-Text-to-Text • 37B • Updated Oct 11 • 266
TheDrainFlorist/GLM-5.3-Flash-VQ-3.1bpw

Image-Text-to-Text • 40B • Updated Oct 11 • 226
TheDrainFlorist/GLM-5.3-Flash-VQ-3.6bpw

Image-Text-to-Text • 45B • Updated Oct 11 • 294

Qwen3.8-Flash-Next — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.8-Flash-Next for Apple Silicon. Four rungs, 2.1-5.5 bpw, MTP sidecar. Needs mlx-lm PR #1788. Built with VQLab.

TheDrainFlorist/Qwen3.8-Flash-Next-VQ-2.1bpw

Text Generation • 24B • Updated Oct 11 • 610 • 5
TheDrainFlorist/Qwen3.8-Flash-Next-VQ-3.2bpw

Text Generation • 43B • Updated Oct 11 • 271 • 1
TheDrainFlorist/Qwen3.8-Flash-Next-VQ-4.4bpw

Text Generation • 89B • Updated Oct 11 • 335 • 2
TheDrainFlorist/Qwen3.8-Flash-Next-VQ-5.5bpw

Text Generation • 60B • Updated Oct 11 • 254

Qwen3.8-27B — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.8-27B for Apple Silicon. Three rungs, 3.9-4.8 bpw, vision included. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/Qwen3.8-27B-VQ-3.9bpw

Text Generation • 12B • Updated Oct 11 • 281
TheDrainFlorist/Qwen3.8-27B-VQ-4.5bpw

Text Generation • 19B • Updated Oct 11 • 145 • 1
TheDrainFlorist/Qwen3.8-27B-VQ-4.8bpw

Text Generation • 13B • Updated Oct 11 • 160

Gemma-4 — Vector Quantized (MLX)

Data-free VQ builds of gemma-4 for Apple Silicon: the 26B MoE and the e4b with VQ per-layer embeddings. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/gemma-4-26b-a4b-it-VQ-6.2bpw

Image-Text-to-Text • 7B • Updated Oct 11 • 115
TheDrainFlorist/gemma-4-e4b-it-VQ-PLE

Image-Text-to-Text • 6B • Updated Oct 11 • 77

Qwen3.6-35B-A3B — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.6-35B-A3B for Apple Silicon, from 32 GB Macs up. Four rungs, 3.4-5.4 bpw. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/Qwen3.6-35B-A3B-VQ-3.4bpw

Text Generation • 6B • Updated Oct 11 • 242 • 1
TheDrainFlorist/Qwen3.6-35B-A3B-VQ-3.8bpw

Text Generation • 7B • Updated Oct 11 • 244
TheDrainFlorist/Qwen3.6-35B-A3B-VQ-4.6bpw

Text Generation • 7B • Updated Oct 11 • 191
TheDrainFlorist/Qwen3.6-35B-A3B-VQ-5.4bpw

Text Generation • 8B • Updated Oct 11 • 229

Qwen3.5-397B-A17B — Vector Quantized (MLX)

Data-free VQ rungs of Qwen3.5-397B-A17B for Apple Silicon, from one 128 GB Mac up. Five rungs, 2.2-3.5 bpw. Stock mlx-lm. Built with VQLab.

TheDrainFlorist/Qwen3.5-397B-A17B-VQ-2.2bpw

Text Generation • 63B • Updated Oct 11 • 188
TheDrainFlorist/Qwen3.5-397B-A17B-VQ-2.4bpw

Text Generation • 120B • Updated Oct 11 • 159
TheDrainFlorist/Qwen3.5-397B-A17B-VQ-2.6bpw

Text Generation • 59B • Updated Oct 11 • 180
TheDrainFlorist/Qwen3.5-397B-A17B-VQ-3.1bpw

Text Generation • 64B • Updated Oct 11 • 191