top | item 45984800 (no title) gac3 | 3 months ago Was this trained on the same data as Dia 1? discuss order hn newest gac3|3 months ago Would be interesting to know what improvements come from arch, data, and different tokenizer.
gac3|3 months ago Would be interesting to know what improvements come from arch, data, and different tokenizer.
gac3|3 months ago