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

Test loading generation config with safetensor weights (#31550)

Browse Source 
fix test
This commit is contained in:
Joao Gante
2024-07-09 15:22:43 +01:00
committed by GitHub
parent cffa2b9c1d
commit 4c2538b863
1 changed files with 5 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
tests/utils/test_modeling_utils.py
View File

@@ -1424,20 +1424,15 @@ class ModelUtilsTest(TestCasePlus):
self.assertEqual(model.__class__.__name__, model_ref.__class__.__name__) self.assertEqual(model.__class__.__name__, model_ref.__class__.__name__)
def test_generation_config_is_loaded_with_model(self): def test_generation_config_is_loaded_with_model(self):
# Note: `joaogante/tiny-random-gpt2-with-generation-config` has a `generation_config.json` containing a dummy # Note: `TinyLlama/TinyLlama-1.1B-Chat-v1.0` has a `generation_config.json` containing `max_length: 2048`
# `transformers_version` field set to `foo`. If loading the file fails, this test also fails.
# 1. Load without further parameters # 1. Load without further parameters
model = AutoModelForCausalLM.from_pretrained( model = AutoModelForCausalLM.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0")
"joaogante/tiny-random-gpt2-with-generation-config", use_safetensors=False self.assertEqual(model.generation_config.max_length, 2048)
)
self.assertEqual(model.generation_config.transformers_version, "foo")
# 2. Load with `device_map` # 2. Load with `device_map`
model = AutoModelForCausalLM.from_pretrained( model = AutoModelForCausalLM.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0", device_map="auto")
"joaogante/tiny-random-gpt2-with-generation-config", device_map="auto", use_safetensors=False self.assertEqual(model.generation_config.max_length, 2048)
)
self.assertEqual(model.generation_config.transformers_version, "foo")
@require_safetensors @require_safetensors
def test_safetensors_torch_from_torch(self): def test_safetensors_torch_from_torch(self):