diff --git a/imaginairy/modules/attention.py b/imaginairy/modules/attention.py
index 1144d8e..58f6d3a 100644
--- a/imaginairy/modules/attention.py
+++ b/imaginairy/modules/attention.py
@@ -1,5 +1,4 @@
 import math
-from inspect import isfunction
 
 import torch
 import torch.nn.functional as F
@@ -10,27 +9,6 @@ from imaginairy.modules.diffusion.util import checkpoint
 from imaginairy.utils import get_device
 
 
-def uniq(arr):
-    return {el: True for el in arr}.keys()
-
-
-def default(val, d):
-    if val is not None:
-        return val
-    return d() if isfunction(d) else d
-
-
-def max_neg_value(t):
-    return -torch.finfo(t.dtype).max
-
-
-def init_(tensor):
-    dim = tensor.shape[-1]
-    std = 1 / math.sqrt(dim)
-    tensor.uniform_(-std, std)
-    return tensor
-
-
 # feedforward
 class GEGLU(nn.Module):
     def __init__(self, dim_in, dim_out):
@@ -193,7 +171,7 @@ class CrossAttention(nn.Module):
         h = self.heads
 
         q_in = self.to_q(x)
-        context = default(context, x)
+        context = context if context is not None else x
         k_in = self.to_k(context)
         v_in = self.to_v(context)
         del context, x
@@ -227,7 +205,7 @@ class CrossAttention(nn.Module):
             max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64
             raise RuntimeError(
                 f"Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). "
-                f"Need: {mem_required/64/gb:0.1f}GB free, Have:{mem_free_total/gb:0.1f}GB free"
+                f"Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free"
             )
 
         slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1]
diff --git a/imaginairy/modules/find_noise.py b/imaginairy/modules/find_noise.py
deleted file mode 100644
index 7d394d0..0000000
--- a/imaginairy/modules/find_noise.py
+++ /dev/null
@@ -1,80 +0,0 @@
-"""
-I tried it with the DDIM sampler and it didn't work.
-
-Probably need to use the k-diffusion sampler with it
-from https://gist.githubusercontent.com/trygvebw/c71334dd127d537a15e9d59790f7f5e1/raw/a846393251f5be8289d4febc75a19f1f962aabcc/find_noise.py
-
-needs https://github.com/crowsonkb/k-diffusion
-"""
-from contextlib import nullcontext
-
-import torch
-from torch import autocast
-
-from imaginairy.img_utils import pillow_img_to_model_latent
-from imaginairy.utils import get_device
-from imaginairy.vendored import k_diffusion as K
-
-
-def find_noise_for_image(model, pil_img, prompt, steps=50, cond_scale=1.0, half=True):
-    img_latent = pillow_img_to_model_latent(model, pil_img, batch_size=1, half=half)
-    return find_noise_for_latent(
-        model,
-        img_latent,
-        prompt,
-        steps=steps,
-        cond_scale=cond_scale,
-    )
-
-
-def find_noise_for_latent(model, img_latent, prompt, steps=50, cond_scale=1.0):
-    x = img_latent
-
-    _autocast = autocast if get_device() in ("cuda", "cpu") else nullcontext
-    with (torch.no_grad(), _autocast(get_device())):
-        uncond = model.get_learned_conditioning([""])
-        cond = model.get_learned_conditioning([prompt])
-
-    s_in = x.new_ones([x.shape[0]])
-    dnw = K.external.CompVisDenoiser(model)
-    sigmas = dnw.get_sigmas(steps).flip(0)
-
-    with (torch.no_grad(), _autocast(get_device())):
-        for i in range(1, len(sigmas)):
-            x_in = torch.cat([x] * 2)
-            sigma_in = torch.cat([sigmas[i] * s_in] * 2)
-            cond_in = torch.cat([uncond, cond])
-
-            c_out, c_in = [
-                K.utils.append_dims(k, x_in.ndim) for k in dnw.get_scalings(sigma_in)
-            ]
-            t = dnw.sigma_to_t(sigma_in)
-
-            eps = model.apply_model(x_in * c_in, t, cond=cond_in)
-            denoised_uncond, denoised_cond = (x_in + eps * c_out).chunk(2)
-
-            denoised = denoised_uncond + (denoised_cond - denoised_uncond) * cond_scale
-
-            d = (x - denoised) / sigmas[i]
-            dt = sigmas[i] - sigmas[i - 1]
-
-            x = x + d * dt
-
-            # This shouldn't be necessary, but solved some VRAM issues
-            del (
-                x_in,
-                sigma_in,
-                cond_in,
-                c_out,
-                c_in,
-                t,
-            )
-            del eps, denoised_uncond, denoised_cond, denoised, d, dt
-            # collect_and_empty()
-
-        # return (x / x.std())
-        return (x / x.std()) * sigmas[-1]
-
-
-if __name__ == "__main__":
-    pass