Benchmark
#1
by pirola - opened
Can it perform better than reagiriam tradicional models on the same parameters count?
It should be tested on several benchmarks to see.
Which model is it finally? Qwen2.5 or Qwen3.5? Do you have a technical report for the final use case?
The final model is Qwen 2.5.
I have just add the initial technical report in the readme currently I am doing evaluation o. MMLU dataset
drop me a message here when you have data! good luck