nvidia/Nemotron-Cascade-2-30B-A3B · Discussions

Resources

View closed (4)

Cascade-2 cheating attempts on Math is... cute

#25 opened 11 days ago by

use system prompt for reasoning benchmarks reproduction (IMO-AnswerBench,..)

#24 opened 15 days ago by

Nemotron Cascade mutates nested path literals in tool arguments even with explicit literal-preservation instructions and temperature=0.

#23 opened 18 days ago by

MoE efficiency for agentic workflows: thinking mode trade-offs

#22 opened 19 days ago by

useless but fast!

#21 opened 21 days ago by

187 tok/s on RTX 3090, 625K Context, Agent Coding (IQ4_XS + Hermes Agent)

#20 opened 21 days ago by

Cascade reasoning latency in production

#19 opened 21 days ago by

Integration Proposal: Aligning with the AURA Sovereign Protocol for Human Enlightenment

#18 opened 21 days ago by

COMMUNITY DISCUSSION

#17 opened 22 days ago by

nairobiansmasher

pruned version

#16 opened 23 days ago by

The model's self-perception is unstable within a Chinese context.

#15 opened 23 days ago by

TEST DISCUSSION

#14 opened 23 days ago by

modeling_nemotron_h.py: Multiple bugs in HybridMambaAttentionDynamicCache break generation with CUDA fast path

#13 opened 23 days ago by

no quants working

#12 opened 23 days ago by

Tool call format degrading at higher context

#10 opened 24 days ago by

Official quantizations?

#9 opened 25 days ago by

Add documentation on how to use with vLLM to README.md

#7 opened 27 days ago by

Installation Video and Testing - Step by Step

#5 opened 28 days ago by

well cooked !! rtx 3060 loves it

#4 opened 28 days ago by

Tool calling ability

#3 opened 28 days ago by

Add eval results

#2 opened 29 days ago by