Luke Alonso
lukealonso
AI & ML interests
None yet
Recent Activity
new activity
about 3 hours ago
lukealonso/MiniMax-M2.5-NVFP4:VLLM error for kv weight scaling - workaround
updated
a model
about 4 hours ago
lukealonso/MiniMax-M2.5-NVFP4
updated
a model
about 4 hours ago
lukealonso/GLM-5-NVFP4
Organizations
None yet
VLLM error for kv weight scaling - workaround
3
#6 opened about 6 hours ago
by
ShaunEvansMD
Thanks for your effort
5
#5 opened about 13 hours ago
by
darkstar3537
fp8 kv cache
9
#4 opened 1 day ago
by
festr2
KeyError: '110.w1.input_scale' with TRT
2
#3 opened 2 days ago
by
guanwenyu1995
"w1_weight_scale_2 must match w3_weight_scale_2. Accuracy may be affected."
👍
1
14
#2 opened 3 days ago
by
zenmagnets
Here's the vLLM recipe I'm using with 2x RTX Pro 6000
👍
1
10
#1 opened 3 days ago
by
zenmagnets
one more time..? Minimax-M2.5 🔥
4
#4 opened 4 days ago
by
reneho
This is perfect! Thank you!
🔥
1
14
#1 opened about 2 months ago
by
ktsaou
MinimaxM2.1
3
#5 opened about 2 months ago
by
reneho
Devstral-2 NVFP4?
3
#3 opened 2 months ago
by
reneho
you know which nightly it worked with? because it does not with current one
31
#1 opened 3 months ago
by
willfalco