pathosethoslogos
pathosethoslogos
·
AI & ML interests
None yet
Recent Activity
new activity about 10 hours ago
RadixArk/Qwen3.8-Flash-Next-NVFP4:One-command deployment on a single DGX Spark (34-42 tok/s, prefix caching, 262K) new activity 1 day ago
unsloth/Qwen3.8-27B-unsloth-bnb-4bit:No explanation on what 'bnb' is in the model card new activity 3 days ago
inclusionAI/Ling-3.0-flash:Any plans for merging into mainline vLLM?Organizations
None yet
One-command deployment on a single DGX Spark (34-42 tok/s, prefix caching, 262K)
🔥👍 2
1
#7 opened 3 days ago
by
hasanbasbunar
No explanation on what 'bnb' is in the model card
#2 opened 1 day ago
by
pathosethoslogos
Any plans for merging into mainline vLLM?
➕ 1
#16 opened 3 days ago
by
pathosethoslogos
When will this DFlash model be compatible or come to mainline vLLM Git?
#12 opened 4 days ago
by
pathosethoslogos
When will it come to mainline vLLM Git?
#7 opened 5 days ago
by
pathosethoslogos
run on DGX SPARK
3
#23 opened 6 days ago
by
Muyanghao
Is this really a 1B model?
4
#2 opened 13 days ago
by
mindplay
model-00016-of-00016.safetensors was just updated... What?
3
#25 opened 8 days ago
by
pathosethoslogos
</think> every response
5
#36 opened 27 days ago
by
pathosethoslogos
Inferact/Qwen3.8-27B-NVFP4 is on vLLM's official documentation
10
#8 opened 17 days ago
by
pathosethoslogos
The model goes crazy and goes into loops
1
#1 opened 11 days ago
by
pathosethoslogos
Qwen3.8-27B Serving Configs: DGX Spark vLLM NVFP4
🤗🚀 10
7
#7 opened 17 days ago
by
erdal
Does not work with vllm 0.27.1 (latest)
7
#11 opened 16 days ago
by
flaviusburca
Bench maxed -- don't be fooled by the table presented in the model card
👀 1
4
#84 opened 16 days ago
by
pathosethoslogos
unsloth/Qwen3.8-27B-NVFP4 vs. Inferact/Qwen3.8-27B-NVFP4?
🔥➕ 3
10
#1 opened 17 days ago
by
pathosethoslogos
Custom vLLM merge request to main vLLM?
👍 1
7
#6 opened 18 days ago
by
pathosethoslogos
Please submit a PR to vLLM for upstream model support?
👍 1
2
#2 opened 17 days ago
by
GadflyII
DGX Spark - 66.7% token acceptance rate - let's push it higher together
5
#19 opened about 2 months ago
by
MavisFace
Quant request
➕ 1
#7 opened 24 days ago
by
pathosethoslogos