Random data fails. Plan is to test vs jax.nn.softmax .
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
- Re: [ot][spam][crazy][data] transformer model 'attention' improvement k
