Luke Alonso PRO

lukealonso's activity

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

New activity in nvidia/Qwen3.8-Flash-Next-FP4-Local 25 days ago

Recipe for VLLM dev (no docker container), and Humaneval benchmark results vs RadixArk

👍 11

2

#4 opened 25 days ago by

OverleapSecreted

New activity in nvidia/GLM-5.3-Flash-FP4-Local 25 days ago

Update README.md

#6 opened 25 days ago by

lukealonso

New activity in local-inference-lab/Qwen3.8-Flash-Next-NVFP4 about 1 month ago

Expand evaluation table with NVIDIA references and independent results

1

#4 opened about 2 months ago by

siertum

config.json in 6909a5be dropped `text_config.ple_embedding_dtype`, causing OOM at model construction

2

#6 opened about 1 month ago by

tingtu0721

Restore text_config.ple_embedding_dtype = nvfp4

#7 opened about 1 month ago by

tingtu0721

New activity in local-inference-lab/GLM-5.3-NVFP4 2 months ago

Upload chat_template.jinja

#1 opened 2 months ago by

bullpoint

New activity in local-inference-lab/GLM-5.3-Flash-NVFP4 2 months ago

2x 6kpro

5

#2 opened 2 months ago by

mtcl

New activity in canada-quant/DeepSeek-V4-Pro-NVFP4-FP8-MTP 6 months ago

Lossless?

👀 1

1

#1 opened 6 months ago by

lukealonso

New activity in lukealonso/MiMo-V2.5-NVFP4 6 months ago

Looping in OpenCode

👀 1

5

#4 opened 6 months ago by

Jon-Nielsen

The original repository has updated some files. Does this repository need to be updated?

1

#7 opened 6 months ago by

fanhed

Serving on two devices

3

#3 opened 6 months ago by

shadowlilac

Will it work on 2X6000 Pros

6

#1 opened 7 months ago by

mtcl

Why not GGUF?

#6 opened 6 months ago by

Nerdsking

New activity in lukealonso/GLM-5.1-NVFP4 6 months ago

Quantization of the Model

1

#9 opened 7 months ago by

shiva2022

New activity in lukealonso/MiMo-V2.5-NVFP4 6 months ago

Link to model and docker image

👍 1

1

#2 opened 6 months ago by

Jon-Nielsen

New activity in lukealonso/GLM-5.1-NVFP4 7 months ago

Fix tool calling: support array-formatted tool content (vLLM/SGLang)

#8 opened 7 months ago by

cudaoom

New activity in lukealonso/MiniMax-M2.7-NVFP4 7 months ago

w1 not matching w3 weight scales

12

#1 opened 7 months ago by

dareposte

tokenizer component mismatch and w1_weight_scale_2 must match w3_weight_scale_2. Accuracy may be affected issue

1

#5 opened 7 months ago by

mtcl

New activity in lukealonso/GLM-5.1-NVFP4 7 months ago

RuntimeError: The size of tensor a (3072) must match the size of tensor b (6144) at non-singleton dimension 1

3

#5 opened 7 months ago by

lianyouzao

From "Doesn't Work" to 641 tok/s: GLM-5.1 NVFP4 on 6× RTX PRO 6000 Blackwell

🔥 1

#4 opened 7 months ago by

sakamakismile