deftdawg/bitsandbytes-rocm
1
0
Fork 0
forked from mrq/bitsandbytes-rocm
Code Pull Requests Activity

cast before allclose

Browse Source
This commit is contained in:
justheuristic 2022-09-18 01:22:31 +03:00
parent 37f805bb44
commit 95dafc6475
1 changed files with 3 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
tests/test_modules.py
View File

@ -541,8 +541,8 @@ def test_linear8bitlt_no_fp16_weights(threshold, memory_efficient_backward):
mlp = MLP8bit(
32, 64, threshold=threshold, has_fp16_weights=False, memory_efficient_backward=memory_efficient_backward
)
w1, w2 = mlp.fc1.weight.clone(), mlp.fc2.weight.clone()
mlp = mlp.cuda().half()
w1, w2 = mlp.fc1.weight.clone(), mlp.fc2.weight.clone() # note: we grad original weights before quantization,
mlp = mlp.cuda().half() # and this line triggers quantization
for i in range(100):
b1 = torch.randn(16, 8, 32, device="cuda").half()
@ -567,8 +567,7 @@ def test_linear8bitlt_no_fp16_weights(threshold, memory_efficient_backward):
mlp.zero_grad()
(o1 * grad_proj).sum().backward()
assert False, (w1, w2)
grad_ref = grad_proj.flatten(2) @ w2 @ w1
grad_ref = grad_proj.flatten(2) @ w2.to(grad_proj.device) @ w1.to(grad_proj.device)
assert torch.allclose(b1.grad, grad_ref)