noctrex

noctrex's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in noctrex/Qwen3.6-35B-A3B-MXFP4_MOE-GGUF 7 days ago

alternative mxfp4 with mixed shared experts and non-experts

7

#1 opened 11 days ago by

hugypufy

New activity in noctrex/gemma-4-26B-A4B-it-MXFP4_MOE-GGUF 26 days ago

MXFP4 of huihui-ai/Huihui-gemma-4-26B-A4B-it-abliterated

👍 1

4

#2 opened 26 days ago by

ampersandru

New activity in noctrex/Huihui-Qwen3.5-35B-A3B-abliterated-MXFP4_MOE-GGUF 26 days ago

Is the uploaded model MXFP4/Q8 or BF16?

2

#1 opened 26 days ago by

ampersandru

New activity in noctrex/gemma-4-26B-A4B-it-MXFP4_MOE-GGUF 27 days ago

Incorrect output during inference

1

#1 opened about 1 month ago by

Ike

New activity in noctrex/Qwen3.5-122B-A10B-MXFP4_MOE-GGUF about 1 month ago

Kind request for Qwen3.5-397B-A17B MXFP4 BF16

7

#2 opened 2 months ago by

dehnhaide

Hello noctrex, could you please tell me whether the MXFP4 has better quality or the Q4KM + imatrix quantization?

1

#3 opened about 1 month ago by

Luck0078

New activity in noctrex/Qwen3.5-35B-A3B-MXFP4_MOE-GGUF about 1 month ago

Comparing with Official GPTQ-Int4 quantized model?

1

#6 opened about 1 month ago by

haili-tian

New activity in noctrex/Mistral-Small-4-119B-2603-MXFP4_MOE-GGUF about 2 months ago

Poor performance and pretty lobotomized

2

#1 opened about 2 months ago by

mancub

New activity in noctrex/Qwen3-Coder-Next-MXFP4_MOE-GGUF about 2 months ago

MXFP4 vs other 4-bit quant algos?

2

#3 opened about 2 months ago by

dinerburger

New activity in noctrex/Qwen3.5-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-MXFP4_MOE-GGUF about 2 months ago

can you please do an f16 version

1

#1 opened about 2 months ago by

Shuasimodo

New activity in noctrex/Qwen3-Coder-Next-MXFP4_MOE-GGUF about 2 months ago

It would be neat to see a Heretical version of this.

1

#1 opened 3 months ago by

SabinStargem

New activity in noctrex/MiroThinker-1.7-mini-MXFP4_MOE-GGUF about 2 months ago

AI Model Evaluation Report: MiroThinker-1.7-Mini (GGUF/Ollama)

1

#1 opened about 2 months ago by

phanthai12

New activity in noctrex/Huihui-Qwen3-VL-4B-Instruct-abliterated-GGUF about 2 months ago

Would it make sense to get Qwen3-VL MXFP4 quants?

20

#2 opened 4 months ago by

ampersandru

New activity in noctrex/Qwen3.5-35B-A3B-MXFP4_MOE-GGUF 2 months ago

command to create GGUF MXFP4 mixed with BF16

1

#5 opened 2 months ago by

ghit72

It's really good.

👀 1

26

#3 opened 2 months ago by

Shuasimodo

New activity in noctrex/LFM2-24B-A2B-MXFP4_MOE-GGUF 2 months ago

Model performance

2

#1 opened 2 months ago by

spanspek

New activity in noctrex/Qwen3-Coder-Next-MXFP4_MOE-GGUF 2 months ago

Increasing the precision of some of the weights when quantizing

👍 4

57

#2 opened 3 months ago by

Shuasimodo

New activity in unsloth/Qwen3.5-35B-A3B-GGUF 2 months ago

Is there some helpful regex to offload all MoE layers to the CPU?

4

#7 opened 2 months ago by

hdnh2006

New activity in noctrex/Qwen3.5-122B-A10B-MXFP4_MOE-GGUF 2 months ago

BF16 version?

1

#1 opened 2 months ago by

Kackliqur

New activity in noctrex/Qwen3.5-35B-A3B-MXFP4_MOE-GGUF 2 months ago

"Use this model" wrong tag by default.

👍 2

1

#2 opened 2 months ago by

jorj2