Text2Text Generation
Transformers
PyTorch
English
switch_transformers
Kareem Gamal Mahmoud commited on
Commit
050016e
1 Parent(s): 65cc0fc

Update config.json

Browse files

The decoder_start_token_id has been forgotten for this model as well.

Files changed (1) hide show
  1. config.json +1 -0
config.json CHANGED
@@ -11,6 +11,7 @@
11
  "expert_capacity": 64,
12
  "feed_forward_proj": "relu",
13
  "initializer_factor": 1.0,
 
14
  "is_encoder_decoder": true,
15
  "is_full_sparse": true,
16
  "is_gated_act": false,
 
11
  "expert_capacity": 64,
12
  "feed_forward_proj": "relu",
13
  "initializer_factor": 1.0,
14
+ "decoder_start_token_id": 0,
15
  "is_encoder_decoder": true,
16
  "is_full_sparse": true,
17
  "is_gated_act": false,