Language Modelling On 2000 Hub5 English

10-stage average accuracy

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	10-stage average accuracy
spirit-lm-interleaved-spoken-and-written	10