Fix typos

Found via `codespell -L therefrom,siz` and `typos --format brief`.

ldm/models/diffusion/plms.py

@@ -233,7 +233,7 @@ class PLMSSampler(object):
             # 2nd order Pseudo Linear Multistep (Adams-Bashforth)
             e_t_prime = (3 * e_t - old_eps[-1]) / 2
         elif len(old_eps) == 2:
-            # 3nd order Pseudo Linear Multistep (Adams-Bashforth)
+            # 3rd order Pseudo Linear Multistep (Adams-Bashforth)
             e_t_prime = (23 * e_t - 16 * old_eps[-1] + 5 * old_eps[-2]) / 12
         elif len(old_eps) >= 3:
             # 4nd order Pseudo Linear Multistep (Adams-Bashforth)

ldm/modules/attention.py

@@ -12,9 +12,9 @@ from ldm.modules.diffusionmodules.util import checkpoint
 try:
     import xformers
     import xformers.ops
-    XFORMERS_IS_AVAILBLE = True
+    XFORMERS_IS_AVAILABLE = True
 except:
-    XFORMERS_IS_AVAILBLE = False
+    XFORMERS_IS_AVAILABLE = False
 
 # CrossAttn precision handling
 import os
@@ -251,7 +251,7 @@ class BasicTransformerBlock(nn.Module):
     def __init__(self, dim, n_heads, d_head, dropout=0., context_dim=None, gated_ff=True, checkpoint=True,
                  disable_self_attn=False):
         super().__init__()
-        attn_mode = "softmax-xformers" if XFORMERS_IS_AVAILBLE else "softmax"
+        attn_mode = "softmax-xformers" if XFORMERS_IS_AVAILABLE else "softmax"
         assert attn_mode in self.ATTENTION_MODES
         attn_cls = self.ATTENTION_MODES[attn_mode]
         self.disable_self_attn = disable_self_attn

ldm/modules/diffusionmodules/model.py

@@ -11,9 +11,9 @@ from ldm.modules.attention import MemoryEfficientCrossAttention
 try:
     import xformers
     import xformers.ops
-    XFORMERS_IS_AVAILBLE = True
+    XFORMERS_IS_AVAILABLE = True
 except:
-    XFORMERS_IS_AVAILBLE = False
+    XFORMERS_IS_AVAILABLE = False
     print("No module 'xformers'. Proceeding without it.")
 
 
@@ -279,7 +279,7 @@ class MemoryEfficientCrossAttentionWrapper(MemoryEfficientCrossAttention):
 
 def make_attn(in_channels, attn_type="vanilla", attn_kwargs=None):
     assert attn_type in ["vanilla", "vanilla-xformers", "memory-efficient-cross-attn", "linear", "none"], f'attn_type {attn_type} unknown'
-    if XFORMERS_IS_AVAILBLE and attn_type == "vanilla":
+    if XFORMERS_IS_AVAILABLE and attn_type == "vanilla":
         attn_type = "vanilla-xformers"
     print(f"making attention of type '{attn_type}' with {in_channels} in_channels")
     if attn_type == "vanilla":

ldm/modules/diffusionmodules/openaimodel.py

@@ -345,7 +345,7 @@ def count_flops_attn(model, _x, y):
 
 class QKVAttentionLegacy(nn.Module):
     """
-    A module which performs QKV attention. Matches legacy QKVAttention + input/ouput heads shaping
+    A module which performs QKV attention. Matches legacy QKVAttention + input/output heads shaping
     """
 
     def __init__(self, n_heads):

ldm/modules/ema.py

@@ -3,14 +3,14 @@ from torch import nn
 
 
 class LitEma(nn.Module):
-    def __init__(self, model, decay=0.9999, use_num_upates=True):
+    def __init__(self, model, decay=0.9999, use_num_updates=True):
         super().__init__()
         if decay < 0.0 or decay > 1.0:
             raise ValueError('Decay must be between 0 and 1')
 
         self.m_name2s_name = {}
         self.register_buffer('decay', torch.tensor(decay, dtype=torch.float32))
-        self.register_buffer('num_updates', torch.tensor(0, dtype=torch.int) if use_num_upates
+        self.register_buffer('num_updates', torch.tensor(0, dtype=torch.int) if use_num_updates
         else torch.tensor(-1, dtype=torch.int))
 
         for name, p in model.named_parameters():

ldm/modules/image_degradation/bsrgan.py

@@ -170,7 +170,7 @@ def gen_kernel(k_size=np.array([15, 15]), scale_factor=np.array([4, 4]), min_var
     [X, Y] = np.meshgrid(range(k_size[0]), range(k_size[1]))
     Z = np.stack([X, Y], 2)[:, :, :, None]
 
-    # Calcualte Gaussian for every pixel of the kernel
+    # Calculate Gaussian for every pixel of the kernel
     ZZ = Z - MU
     ZZ_t = ZZ.transpose(0, 1, 3, 2)
     raw_kernel = np.exp(-0.5 * np.squeeze(ZZ_t @ INV_SIGMA @ ZZ)) * (1 + noise)
@@ -613,7 +613,7 @@ def degradation_bsrgan_variant(image, sf=4, isp_model=None):
     return example
 
 
-# TODO incase there is a pickle error one needs to replace a += x with a = a + x in add_speckle_noise etc...
+# TODO in case there is a pickle error one needs to replace a += x with a = a + x in add_speckle_noise etc...
 def degradation_bsrgan_plus(img, sf=4, shuffle_prob=0.5, use_sharp=True, lq_patchsize=64, isp_model=None):
     """
     This is an extended degradation model by combining

ldm/modules/image_degradation/bsrgan_light.py

@@ -169,7 +169,7 @@ def gen_kernel(k_size=np.array([15, 15]), scale_factor=np.array([4, 4]), min_var
     [X, Y] = np.meshgrid(range(k_size[0]), range(k_size[1]))
     Z = np.stack([X, Y], 2)[:, :, :, None]
 
-    # Calcualte Gaussian for every pixel of the kernel
+    # Calculate Gaussian for every pixel of the kernel
     ZZ = Z - MU
     ZZ_t = ZZ.transpose(0, 1, 3, 2)
     raw_kernel = np.exp(-0.5 * np.squeeze(ZZ_t @ INV_SIGMA @ ZZ)) * (1 + noise)

ldm/modules/image_degradation/utils_image.py

@@ -59,7 +59,7 @@ def surf(Z, cmap='rainbow', figsize=None):
 
 '''
 # --------------------------------------------
-# get image pathes
+# get image paths
 # --------------------------------------------
 '''
 
@@ -122,14 +122,14 @@ def imssave(imgs, img_path):
         cv2.imwrite(new_path, img)
 
 
-def split_imageset(original_dataroot, taget_dataroot, n_channels=3, p_size=800, p_overlap=96, p_max=1000):
+def split_imageset(original_dataroot, target_dataroot, n_channels=3, p_size=800, p_overlap=96, p_max=1000):
     """
     split the large images from original_dataroot into small overlapped images with size (p_size)x(p_size),
-    and save them into taget_dataroot; only the images with larger size than (p_max)x(p_max)
-    will be splitted.
+    and save them into target_dataroot; only the images with larger size than (p_max)x(p_max)
+    will be split.
     Args:
         original_dataroot:
-        taget_dataroot:
+        target_dataroot:
         p_size: size of small images
         p_overlap: patch size in training is a good choice
         p_max: images with smaller size than (p_max)x(p_max) keep unchanged.
@@ -139,8 +139,8 @@ def split_imageset(original_dataroot, taget_dataroot, n_channels=3, p_size=800,
         # img_name, ext = os.path.splitext(os.path.basename(img_path))
         img = imread_uint(img_path, n_channels=n_channels)
         patches = patches_from_image(img, p_size, p_overlap, p_max)
-        imssave(patches, os.path.join(taget_dataroot,os.path.basename(img_path)))
-        #if original_dataroot == taget_dataroot:
+        imssave(patches, os.path.join(target_dataroot,os.path.basename(img_path)))
+        #if original_dataroot == target_dataroot:
         #del img_path
 
 '''
@@ -180,7 +180,7 @@ def mkdir_and_rename(path):
 
 
 # --------------------------------------------
-# get uint8 image of size HxWxn_channles (RGB)
+# get uint8 image of size HxWxn_channels (RGB)
 # --------------------------------------------
 def imread_uint(path, n_channels=3):
     #  input: path
@@ -215,7 +215,7 @@ def imwrite(img, img_path):
 
 
 # --------------------------------------------
-# get single image of size HxWxn_channles (BGR)
+# get single image of size HxWxn_channels (BGR)
 # --------------------------------------------
 def read_img(path):
     # read image by cv2

ldm/modules/midas/midas/transforms.py

@@ -125,7 +125,7 @@ class Resize(object):
                     # fit height
                     scale_width = scale_height
             elif self.__resize_method == "minimal":
-                # scale as least as possbile
+                # scale as least as possible
                 if abs(1 - scale_width) < abs(1 - scale_height):
                     # fit width
                     scale_height = scale_width

ldm/util.py

@@ -23,7 +23,7 @@ def log_txt_as_img(wh, xc, size=10):
         try:
             draw.text((0, 0), lines, fill="black", font=font)
         except UnicodeEncodeError:
-            print("Cant encode string for logging. Skipping.")
+            print("Can't encode string for logging. Skipping.")
 
         txt = np.array(txt).transpose(2, 0, 1) / 127.5 - 1.0
         txts.append(txt)

modelcard.md

@@ -112,7 +112,7 @@ In addition to the textual input, it receives a `noise_level` as an input parame
 - **Optimizer:** AdamW
 - **Gradient Accumulations**: 1
 - **Batch:** 32 x 8 x 2 x 4 = 2048
-- **Learning rate:** warmup to 0.0001 for 10,000 steps and then kept constant
+- **Learning rate:** warm up to 0.0001 for 10,000 steps and then kept constant
 
 ## Evaluation Results 
 Evaluations with different classifier-free guidance scales (1.5, 2.0, 3.0, 4.0,