FIX Faulty test that results in nan weights (#2448)

This specific test used a learning rate that is too high, resulting in
nan weights. Then, when weights are compared to assert that they're
different, the test passes trivially because nan != nan. The lr is now
reduced and there is a sanity check that none of the weights contain
non-finite values.

See discussion in
huggingface/peft#2433 (comment)
ff.

Author: csqaiub

tests/test_custom_models.py

@@ -4134,9 +4134,9 @@ def test_training_works(self, model_cls, custom_module_cls, custom_lora_cls):
         config._register_custom_module({custom_module_cls: custom_lora_cls})
 
         peft_model = get_peft_model(model, config)
-        sd_before = peft_model.state_dict()
+        sd_before = copy.deepcopy(peft_model.state_dict())
         inputs = torch.randn(16, 10)
-        optimizer = torch.optim.SGD(peft_model.parameters(), lr=1e-1)
+        optimizer = torch.optim.SGD(peft_model.parameters(), lr=1e-4)
 
         for _ in range(5):
             optimizer.zero_grad()
@@ -4146,6 +4146,13 @@ def test_training_works(self, model_cls, custom_module_cls, custom_lora_cls):
             optimizer.step()
 
         sd_after = peft_model.state_dict()
+
+        # sanity check that for finite results, since nan != nan, which would make the test pass trivially
+        for val in sd_before.values():
+            assert torch.isfinite(val).all()
+        for val in sd_after.values():
+            assert torch.isfinite(val).all()
+
         assert not torch.allclose(
             sd_before["base_model.model.my_module.lora_A.default.weight"],
             sd_after["base_model.model.my_module.lora_A.default.weight"],