MCPcopy Create free account
hub / github.com/deepspeedai/DeepSpeed / test

Method test

tests/unit/linear/test_linear.py:24–40  ·  view source on GitHub ↗
(self)

Source from the content-addressed store, hash-verified

22 world_size = 2
23
24 def test(self):
25 lora_config = None
26 quantization_config = None
27
28 input_features = 64 # Number of input features
29 output_features = 64 # Number of output features
30 batch_size = 1 # Number of samples in a batch
31
32 linear_layer = OptimizedLinear(input_dim=input_features,
33 output_dim=output_features,
34 lora_config=lora_config,
35 quantization_config=quantization_config,
36 dtype=torch.bfloat16)
37
38 dummy_input = torch.rand(batch_size, input_features, dtype=torch.bfloat16)
39 output = linear_layer(dummy_input)
40 assert output.shape == (batch_size, output_features)
41
42
43@pytest.mark.parametrize("base_weight_sharding", [1, 2])

Callers

nothing calls this directly

Calls 1

OptimizedLinearClass · 0.90

Tested by

no test coverage detected