Add TF weights
Model converted by the transformers' pt_to_tf CLI. All converted model outputs and hidden layers were validated against its Pytorch counterpart.
Maximum crossload output difference=1.621e-05; Maximum crossload hidden layer difference=7.210e-03;
Maximum conversion output difference=1.621e-05; Maximum conversion hidden layer difference=7.210e-03;
List of maximum output differences above the threshold (1e-19):
logits: 1.621e-05
List of maximum hidden layer differences above the threshold (1e-19):
hidden_states[0]: 1.287e-05
hidden_states[1]: 4.435e-05
hidden_states[2]: 5.460e-05
hidden_states[3]: 5.078e-05
hidden_states[4]: 6.038e-05
hidden_states[5]: 1.445e-04
hidden_states[6]: 7.858e-04
hidden_states[7]: 2.953e-03
hidden_states[8]: 5.280e-03
hidden_states[9]: 6.203e-03
hidden_states[10]: 5.585e-03
hidden_states[11]: 7.210e-03
hidden_states[12]: 5.981e-03