Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Can you add in Qwen3.5 and other series of models for testing?
#2
by TomLucidor - opened
Qwen3.5 seems to be a parting gift for some of the major researchers, and that their linear attention models got a lot of support. I wonder if those models can be tested.
There are also other model series like LFM, Falcon-H1, Inclusion AI's Ring, Granite-4.0, and the newer Nemotron models, will get covered as well.
Hey @TomLucidor ! If you're talking about swe-rebench leaderboard https://swe-rebench.com, a series of Qwen3.5 models will be added in the upcoming release
@djalexj Please include the smaller models too if you can! Cus I want to see if linear attention and SLMs can be "better than expected" for agentic problem-solving.