Article
Speculative Decoding in Practice: How EAGLE3 Makes LLMs Faster Without Changing Their Outputs
•
8
Zhukov
Geximus
·
AI & ML interests
None yet
Recent Activity
upvoted an article about 23 hours ago
Speculative Decoding in Practice: How EAGLE3 Makes LLMs Faster Without Changing Their Outputs new activity 8 days ago
demon-zombie/MiniMax-M2.7-AWQ-4bit:These are NOT actual AWQ-quantized models. new activity 9 days ago
MiniMaxAI/MiniMax-M2.7:MiniMax-M2.7 is highly verbose and slowOrganizations
None yet
upvoted an article about 23 hours ago
These are NOT actual AWQ-quantized models.
2
#1 opened 9 days ago
by 
cai-cai
MiniMax-M2.7 is highly verbose and slow
2
#18 opened 10 days ago
by 
weisunding
thanks for 4bit awq!
🔥 3
1
#1 opened 10 days ago
by 
morlz
upvoted an article 10 days ago
Article
2x Faster on a 229B MoE: EAGLE3 Speculative Decoding for MiniMax-M2.5
•
2
liked a
model 11 days ago
Is minimax 2.7 on the way?
👍 1
1
#3 opened 11 days ago
by Geximus
Minimax 2.7???
49
#53 opened about 1 month ago
by 
rombodawg
Perfomance question
2
#4 opened 21 days ago
by Geximus
liked a
model about 2 months ago
liked a
model 3 months ago
Low generation speed and low GPU utilization (~12%) during inference
👀😔 2
7
#18 opened 3 months ago
by Geximus
liked 2
models 3 months ago
Why is Flux 2 so slow in Img2Img even though everything is in CUDA?
3
#22 opened 5 months ago
by 
Fredtt3
Perfomance of this model is one of the best
🤗👍 2
2
#13 opened 4 months ago
by Geximus
liked a
Space 5 months ago
why recently re-uploaded the core?
2
#7 opened 5 months ago
by Geximus
liked a
model 5 months ago
upvoted a collection 7 months ago