Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

nanochat

Benchmarks

Task NameDataset NameSOTA ResultTrend
Training Data AttributionNanochat Unseen (test)
LDS Correlation0.1792
26
Language Modelingnanochat (val)
Perplexity (val)11.42
8
Chat Model EvaluationNanoChat-D12 (CORE)
CORE Score22.44
2
Showing 3 of 3 rows