0xSero

24 14

HorrorSciFiAi's profile picture

as4020's profile picture

Sportsandfragrance's profile picture

https://sybilsolutions.ai

0xSero
0xSero

AI & ML interests

Quantizing, benchmarking, training, and building.

Recent Activity

published a model about 10 hours ago

0xSero/deepseek-v4-flash-0731-spark

updated a model about 11 hours ago

0xSero/deepseek-v4-flash-0731-spark

updated a model about 12 hours ago

0xSero/DeepSeek-V4-Flash-0731-EXL3-3.0bpw

View all activity

Organizations

0xSero 's collections 11

Inkling-Small EXL3 Quantization Suite

Public EXL3 trellis quantizations of thinkingmachines/Inkling-Small from 2.0 through 5.0 bpw.

0xSero/Inkling-Small-EXL3-2.0bpw

Image-Text-to-Text • 41B • Updated 1 day ago • 13
0xSero/Inkling-Small-EXL3-2.5bpw

Image-Text-to-Text • 49B • Updated 1 day ago • 14 • 1
0xSero/Inkling-Small-EXL3-3.0bpw

Image-Text-to-Text • 57B • Updated 1 day ago • 18
0xSero/Inkling-Small-EXL3-3.5bpw

Image-Text-to-Text • 65B • Updated 1 day ago • 17

Gemma — REAP

REAP-pruned Gemma-4 MoE.

0xSero/Gemma-4-19B

Text Generation • 19B • Updated May 30 • 92 • 20
0xSero/Gemma-4-21B

Text Generation • 21B • Updated May 30 • 831 • 102
0xSero/gemma-moe-reap

Text Generation • Updated May 30 • 1

Proven REAPs

Benchmarked REAP checkpoints with >=500 all-time downloads. GLM/Qwen/MiniMax/DeepSeek/Kimi/gemma.

0xSero/GLM-4.7-185B-W4A16

Text Generation • 2B • Updated May 30 • 72 • 69
0xSero/Qwen3.5-99B-GGUF

Text Generation • 99B • Updated May 30 • 6.08k • 11
0xSero/Gemma-4-21B

Text Generation • 21B • Updated May 30 • 831 • 102
0xSero/GLM-4.7-REAP-40-W4A16

Text Generation • 2B • Updated May 30 • 85 • 7

MiniMax — REAP

REAP-pruned & quantized MiniMax-M2.1 / M2.7.

0xSero/MiniMax-M2.1-139B

Text Generation • 139B • Updated May 30 • 40 • 10
0xSero/MiniMax-M2.1-162B

Text Generation • 162B • Updated May 30 • 15 • 4

Trinity — REAP

REAP-pruned & quantized Trinity-Large-Thinking.

0xSero/Trinity-337B

Text Generation • 337B • Updated May 30 • 15
0xSero/Trinity-337B-W4A16

Text Generation • 46B • Updated May 30 • 8

Datasets — observations & calibration

REAP layerwise observations, calibration sets, and training data.

0xSero/deepseek-v4-flash-reap-observations-v1

Preview • Updated Apr 25 • 52 • 2
0xSero/gemma-moe-reap

Updated Apr 14 • 17
0xSero/glm-calibration-pile640-v2

Updated Apr 14 • 19
0xSero/glm47-calibration-1360

Viewer • Updated Apr 14 • 1.38k • 41 • 1

Qwen — REAP

REAP-pruned & quantized Qwen3.5 / 3.6 / Coder variants.

0xSero/Qwen3-Coder-57B

Text Generation • 57B • Updated May 30 • 11 • 2
0xSero/Qwen3-Coder-64B

Text Generation • 64B • Updated May 30 • 21 • 2
0xSero/Qwen3.5-264B

Text Generation • 263B • Updated May 30 • 16 • 1
0xSero/Qwen3.5-264B-FP8

Text Generation • 264B • Updated May 30 • 11

GLM — REAP

REAP-pruned & quantized GLM-4.x / 5 / 5.1 (+ Flash fine-tunes).

0xSero/GLM-4.6-218B-W4A16

Text Generation • 2B • Updated May 30 • 11 • 8
0xSero/GLM-4.7-185B

Text Generation • 185B • Updated May 30 • 26 • 19
0xSero/GLM-4.7-185B-W4A16

Text Generation • 2B • Updated May 30 • 72 • 69
0xSero/GLM-4.7-202B

Text Generation • 202B • Updated May 30 • 20 • 2

DeepSeek — REAP

REAP-pruned & quantized DeepSeek-V4-Flash / V3.2.

0xSero/DeepSeek-V3.2-345B-W3A16

Text Generation • 2B • Updated May 30 • 18 • 10
0xSero/DeepSeek-V3.2-508B-NVFP4

Text Generation • 2B • Updated May 30 • 17
0xSero/DeepSeek-V4-Flash-162B

Text Generation • 92B • Updated May 30 • 1.88k • 31
0xSero/DeepSeek-V4-Flash-162B-GGUF

Text Generation • 163B • Updated May 30 • 3k • 29

Nemotron — REAP

REAP-pruned & quantized NVIDIA Nemotron-3-Super.

0xSero/Nemotron-3-Super-64B

Text Generation • 64B • Updated May 30 • 130 • 7
0xSero/Nemotron-3-Super-64B-W4A16

Text Generation • 6B • Updated May 30 • 26 • 3
0xSero/Nemotron-3-Super-92B

Text Generation • 92B • Updated May 30 • 119 • 2
0xSero/Nemotron-3-Super-92B-W4A16

Text Generation • 6B • Updated May 30 • 21 • 2

Other models

Hy3, Kimi-K2.5, INTELLECT-3, NousCoder.

0xSero/Hy3-299B-FP8

Text Generation • 299B • Updated May 30 • 47
0xSero/Hy3-299B-NVFP4

Text Generation • 169B • Updated May 30 • 384 • 4
0xSero/INTELLECT-3-57B

Text Generation • 57B • Updated May 30 • 24 • 3
0xSero/NousCoder-14B-SFT

Text Generation • Updated May 30 • 20 • 5

Inkling-Small EXL3 Quantization Suite

Public EXL3 trellis quantizations of thinkingmachines/Inkling-Small from 2.0 through 5.0 bpw.

0xSero/Inkling-Small-EXL3-2.0bpw

Image-Text-to-Text • 41B • Updated 1 day ago • 13
0xSero/Inkling-Small-EXL3-2.5bpw

Image-Text-to-Text • 49B • Updated 1 day ago • 14 • 1
0xSero/Inkling-Small-EXL3-3.0bpw

Image-Text-to-Text • 57B • Updated 1 day ago • 18
0xSero/Inkling-Small-EXL3-3.5bpw

Image-Text-to-Text • 65B • Updated 1 day ago • 17

Qwen — REAP

REAP-pruned & quantized Qwen3.5 / 3.6 / Coder variants.

0xSero/Qwen3-Coder-57B

Text Generation • 57B • Updated May 30 • 11 • 2
0xSero/Qwen3-Coder-64B

Text Generation • 64B • Updated May 30 • 21 • 2
0xSero/Qwen3.5-264B

Text Generation • 263B • Updated May 30 • 16 • 1
0xSero/Qwen3.5-264B-FP8

Text Generation • 264B • Updated May 30 • 11

Gemma — REAP

REAP-pruned Gemma-4 MoE.

0xSero/Gemma-4-19B

Text Generation • 19B • Updated May 30 • 92 • 20
0xSero/Gemma-4-21B

Text Generation • 21B • Updated May 30 • 831 • 102
0xSero/gemma-moe-reap

Text Generation • Updated May 30 • 1

GLM — REAP

REAP-pruned & quantized GLM-4.x / 5 / 5.1 (+ Flash fine-tunes).

0xSero/GLM-4.6-218B-W4A16

Text Generation • 2B • Updated May 30 • 11 • 8
0xSero/GLM-4.7-185B

Text Generation • 185B • Updated May 30 • 26 • 19
0xSero/GLM-4.7-185B-W4A16

Text Generation • 2B • Updated May 30 • 72 • 69
0xSero/GLM-4.7-202B

Text Generation • 202B • Updated May 30 • 20 • 2

Proven REAPs

Benchmarked REAP checkpoints with >=500 all-time downloads. GLM/Qwen/MiniMax/DeepSeek/Kimi/gemma.

0xSero/GLM-4.7-185B-W4A16

Text Generation • 2B • Updated May 30 • 72 • 69
0xSero/Qwen3.5-99B-GGUF

Text Generation • 99B • Updated May 30 • 6.08k • 11
0xSero/Gemma-4-21B

Text Generation • 21B • Updated May 30 • 831 • 102
0xSero/GLM-4.7-REAP-40-W4A16

Text Generation • 2B • Updated May 30 • 85 • 7

DeepSeek — REAP

REAP-pruned & quantized DeepSeek-V4-Flash / V3.2.

0xSero/DeepSeek-V3.2-345B-W3A16

Text Generation • 2B • Updated May 30 • 18 • 10
0xSero/DeepSeek-V3.2-508B-NVFP4

Text Generation • 2B • Updated May 30 • 17
0xSero/DeepSeek-V4-Flash-162B

Text Generation • 92B • Updated May 30 • 1.88k • 31
0xSero/DeepSeek-V4-Flash-162B-GGUF

Text Generation • 163B • Updated May 30 • 3k • 29

MiniMax — REAP

REAP-pruned & quantized MiniMax-M2.1 / M2.7.

0xSero/MiniMax-M2.1-139B

Text Generation • 139B • Updated May 30 • 40 • 10
0xSero/MiniMax-M2.1-162B

Text Generation • 162B • Updated May 30 • 15 • 4

Nemotron — REAP

REAP-pruned & quantized NVIDIA Nemotron-3-Super.

0xSero/Nemotron-3-Super-64B

Text Generation • 64B • Updated May 30 • 130 • 7
0xSero/Nemotron-3-Super-64B-W4A16

Text Generation • 6B • Updated May 30 • 26 • 3
0xSero/Nemotron-3-Super-92B

Text Generation • 92B • Updated May 30 • 119 • 2
0xSero/Nemotron-3-Super-92B-W4A16

Text Generation • 6B • Updated May 30 • 21 • 2

Trinity — REAP

REAP-pruned & quantized Trinity-Large-Thinking.

0xSero/Trinity-337B

Text Generation • 337B • Updated May 30 • 15
0xSero/Trinity-337B-W4A16

Text Generation • 46B • Updated May 30 • 8

Other models

Hy3, Kimi-K2.5, INTELLECT-3, NousCoder.

0xSero/Hy3-299B-FP8

Text Generation • 299B • Updated May 30 • 47
0xSero/Hy3-299B-NVFP4

Text Generation • 169B • Updated May 30 • 384 • 4
0xSero/INTELLECT-3-57B

Text Generation • 57B • Updated May 30 • 24 • 3
0xSero/NousCoder-14B-SFT

Text Generation • Updated May 30 • 20 • 5

Datasets — observations & calibration

REAP layerwise observations, calibration sets, and training data.

0xSero/deepseek-v4-flash-reap-observations-v1

Preview • Updated Apr 25 • 52 • 2
0xSero/gemma-moe-reap

Updated Apr 14 • 17
0xSero/glm-calibration-pile640-v2

Updated Apr 14 • 19
0xSero/glm47-calibration-1360

Viewer • Updated Apr 14 • 1.38k • 41 • 1