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

Revert workaround for TF safetensors loading (#30128)

Browse Source 
* See if we can get tests to pass with the fixed weights

* See if we can get tests to pass with the fixed weights

* Replace the revisions now that we don't need them anymore
This commit is contained in:
Matt
2024-04-09 11:04:18 +01:00
committed by GitHub
parent 841e87ef4f
commit ec59a42192
3 changed files with 5 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/generation/test_tf_utils.py
View File

@@ -194,7 +194,7 @@ class TFGenerationIntegrationTests(unittest.TestCase, GenerationIntegrationTests
tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-gpt2")
text = """Hello, my dog is cute and"""
tokens = tokenizer(text, return_tensors="tf")
model = TFAutoModelForCausalLM.from_pretrained("hf-internal-testing/tiny-random-gpt2", use_safetensors=False)
model = TFAutoModelForCausalLM.from_pretrained("hf-internal-testing/tiny-random-gpt2")
eos_token_id = 638
# forces the generation to happen on CPU, to avoid GPU-related quirks