SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

[Config, Caching] Remove output_past everywhere and replace by use_cache argument (#3734)

Browse Source 
* remove output_past from pt

* make style

* add optional input length for gpt2

* add use cache to prepare input

* save memory in gpt2

* correct gpt2 test inputs

* make past input optional for gpt2

* finish use_cache for all models

* make style

* delete modeling_gpt2 change in test file

* correct docstring

* correct is true statements for gpt2
This commit is contained in:
Patrick von Platen
2020-04-14 20:40:28 +02:00
committed by GitHub
parent 092cf881a5
commit 01c37dcdb5
15 changed files with 342 additions and 168 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/summarization/bart/run_bart_sum.py
View File

@@ -19,7 +19,7 @@ class BartSystem(BaseTransformer):
mode = "language-modeling"
def __init__(self, hparams):
super().__init__(hparams, num_labels=None, mode=self.mode, output_past=False)
super().__init__(hparams, num_labels=None, mode=self.mode)
def forward(self, input_ids, attention_mask=None, decoder_input_ids=None, lm_labels=None):
return self.model(