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

Fix model integration ci (#26322)

Browse Source 
* fix wav2vec2

* nit

* stash

* one more file to update

* fix byt5

* vocab size is 256, don't change that!

* use other revision

* test persimon in smaller size

* style

* tests

* nits

* update add tokens from pretrained

* test tokenization

* nits

* potential fnet fix?

* more nits

* nits

* correct test

* assert close

* udpate

* ouch

* fix it

* some more nits

* FINALLU

* use `adept` checkpoints

* more adept checkpoints

* that was invlved!
This commit is contained in:
Arthur
2023-10-02 13:55:46 +02:00
committed by GitHub
parent 6824461f2a
commit 63864e057f
9 changed files with 38 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
tests/models/fnet/test_modeling_fnet.py
View File

@@ -532,8 +532,6 @@ class FNetModelIntegrationTest(unittest.TestCase):
@slow
@require_tokenizers
def test_inference_long_sentence(self):
model = FNetForMaskedLM.from_pretrained("google/fnet-base")
model.to(torch_device)
tokenizer = FNetTokenizerFast.from_pretrained("google/fnet-base")
inputs = tokenizer(
@@ -543,8 +541,15 @@ class FNetModelIntegrationTest(unittest.TestCase):
padding="max_length",
max_length=512,
)
# fmt: off
torch.testing.assert_allclose(inputs["input_ids"], torch.tensor([[4, 13, 283, 2479, 106, 8, 6, 845, 5, 168, 65, 367, 6, 845, 5, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3, 3, 3, 3, 3, 3, 3, 3, 3, 3,3]]))
# fmt: on
inputs = {k: v.to(torch_device) for k, v in inputs.items()}
model = FNetForMaskedLM.from_pretrained("google/fnet-base")
model.to(torch_device)
logits = model(**inputs).logits
predictions_mask_1 = tokenizer.decode(logits[0, 6].topk(5).indices)
predictions_mask_2 = tokenizer.decode(logits[0, 12].topk(5).indices)