Replace print with logging (#6138)

* Replace print with logging

* nit

* nit

* nit

* nit

* nit

* nit

comfy/cldm/cldm.py

@@ -160,7 +160,6 @@ class ControlNet(nn.Module):
             if isinstance(self.num_classes, int):
                 self.label_emb = nn.Embedding(num_classes, time_embed_dim)
             elif self.num_classes == "continuous":
-                print("setting up linear c_adm embedding layer")
                 self.label_emb = nn.Linear(1, time_embed_dim)
             elif self.num_classes == "sequential":
                 assert adm_in_channels is not None

comfy/extra_samplers/uni_pc.py

@@ -2,6 +2,7 @@
 
 import torch
 import math
+import logging
 
 from tqdm.auto import trange
 
@@ -474,7 +475,7 @@ class UniPC:
             return self.multistep_uni_pc_vary_update(x, model_prev_list, t_prev_list, t, order, **kwargs)
 
     def multistep_uni_pc_vary_update(self, x, model_prev_list, t_prev_list, t, order, use_corrector=True):
-        print(f'using unified predictor-corrector with order {order} (solver type: vary coeff)')
+        logging.info(f'using unified predictor-corrector with order {order} (solver type: vary coeff)')
         ns = self.noise_schedule
         assert order <= len(model_prev_list)
 
@@ -518,7 +519,6 @@ class UniPC:
             A_p = C_inv_p
 
         if use_corrector:
-            print('using corrector')
             C_inv = torch.linalg.inv(C)
             A_c = C_inv
 

comfy/hooks.py

@@ -5,6 +5,7 @@ import math
 import torch
 import numpy as np
 import itertools
+import logging
 
 if TYPE_CHECKING:
     from comfy.model_patcher import ModelPatcher, PatcherInjection
@@ -575,7 +576,7 @@ def load_hook_lora_for_models(model: 'ModelPatcher', clip: 'CLIP', lora: dict[st
     k1 = set(k1)
     for x in loaded:
         if (x not in k) and (x not in k1):
-            print(f"NOT LOADED {x}")
+            logging.warning(f"NOT LOADED {x}")
     return (new_modelpatcher, new_clip, hook_group)
 
 def _combine_hooks_from_values(c_dict: dict[str, HookGroup], values: dict[str, HookGroup], cache: dict[tuple[HookGroup, HookGroup], HookGroup]):

comfy/ldm/aura/mmdit.py

@@ -381,7 +381,6 @@ class MMDiT(nn.Module):
         pe_new = pe_as_2d.squeeze(0).permute(1, 2, 0).flatten(0, 1)
         self.positional_encoding.data = pe_new.unsqueeze(0).contiguous()
         self.h_max, self.w_max = target_dim
-        print("PE extended to", target_dim)
 
     def pe_selection_index_based_on_dim(self, h, w):
         h_p, w_p = h // self.patch_size, w // self.patch_size

comfy/ldm/modules/diffusionmodules/util.py

@@ -9,6 +9,7 @@
 
 
 import math
+import logging
 import torch
 import torch.nn as nn
 import numpy as np
@@ -130,7 +131,7 @@ def make_ddim_timesteps(ddim_discr_method, num_ddim_timesteps, num_ddpm_timestep
     # add one to get the final alpha values right (the ones from first scale to data during sampling)
     steps_out = ddim_timesteps + 1
     if verbose:
-        print(f'Selected timesteps for ddim sampler: {steps_out}')
+        logging.info(f'Selected timesteps for ddim sampler: {steps_out}')
     return steps_out
 
 
@@ -142,8 +143,8 @@ def make_ddim_sampling_parameters(alphacums, ddim_timesteps, eta, verbose=True):
     # according the the formula provided in https://arxiv.org/abs/2010.02502
     sigmas = eta * np.sqrt((1 - alphas_prev) / (1 - alphas) * (1 - alphas / alphas_prev))
     if verbose:
-        print(f'Selected alphas for ddim sampler: a_t: {alphas}; a_(t-1): {alphas_prev}')
-        print(f'For the chosen value of eta, which is {eta}, '
+        logging.info(f'Selected alphas for ddim sampler: a_t: {alphas}; a_(t-1): {alphas_prev}')
+        logging.info(f'For the chosen value of eta, which is {eta}, '
               f'this results in the following sigma_t schedule for ddim sampler {sigmas}')
     return sigmas, alphas, alphas_prev
 

comfy/ldm/util.py

@@ -1,4 +1,5 @@
 import importlib
+import logging
 
 import torch
 from torch import optim
@@ -23,7 +24,7 @@ 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.")
+            logging.warning("Cant encode string for logging. Skipping.")
 
         txt = np.array(txt).transpose(2, 0, 1) / 127.5 - 1.0
         txts.append(txt)
@@ -65,7 +66,7 @@ def mean_flat(tensor):
 def count_params(model, verbose=False):
     total_params = sum(p.numel() for p in model.parameters())
     if verbose:
-        print(f"{model.__class__.__name__} has {total_params*1.e-6:.2f} M params.")
+        logging.info(f"{model.__class__.__name__} has {total_params*1.e-6:.2f} M params.")
     return total_params
 
 

comfy/model_base.py

@@ -770,7 +770,6 @@ class Flux(BaseModel):
             mask = torch.ones_like(noise)[:, :1]
 
         mask = torch.mean(mask, dim=1, keepdim=True)
-        print(mask.shape)
         mask = utils.common_upscale(mask.to(device), noise.shape[-1] * 8, noise.shape[-2] * 8, "bilinear", "center")
         mask = mask.view(mask.shape[0], mask.shape[2] // 8, 8, mask.shape[3] // 8, 8).permute(0, 2, 4, 1, 3).reshape(mask.shape[0], -1, mask.shape[2] // 8, mask.shape[3] // 8)
         mask = utils.resize_to_batch_size(mask, noise.shape[0])

comfy/model_management.py

@@ -1084,7 +1084,7 @@ def unload_all_models():
 
 
 def resolve_lowvram_weight(weight, model, key): #TODO: remove
-    print("WARNING: The comfy.model_management.resolve_lowvram_weight function will be removed soon, please stop using it.")
+    logging.warning("The comfy.model_management.resolve_lowvram_weight function will be removed soon, please stop using it.")
     return weight
 
 #TODO: might be cleaner to put this somewhere else

comfy/model_patcher.py

@@ -773,7 +773,7 @@ class ModelPatcher:
         return self.model.device
 
     def calculate_weight(self, patches, weight, key, intermediate_dtype=torch.float32):
-        print("WARNING the ModelPatcher.calculate_weight function is deprecated, please use: comfy.lora.calculate_weight instead")
+        logging.warning("The ModelPatcher.calculate_weight function is deprecated, please use: comfy.lora.calculate_weight instead")
         return comfy.lora.calculate_weight(patches, weight, key, intermediate_dtype=intermediate_dtype)
 
     def cleanup(self):
@@ -1029,7 +1029,7 @@ class ModelPatcher:
                 if cached_weights is not None:
                     for key in cached_weights:
                         if key not in model_sd_keys:
-                            print(f"WARNING cached hook could not patch. key does not exist in model: {key}")
+                            logging.warning(f"Cached hook could not patch. Key does not exist in model: {key}")
                             continue
                         self.patch_cached_hook_weights(cached_weights=cached_weights, key=key, memory_counter=memory_counter)
                 else:
@@ -1039,7 +1039,7 @@ class ModelPatcher:
                         original_weights = self.get_key_patches()
                     for key in relevant_patches:
                         if key not in model_sd_keys:
-                            print(f"WARNING cached hook would not patch. key does not exist in model: {key}")
+                            logging.warning(f"Cached hook would not patch. Key does not exist in model: {key}")
                             continue
                         self.patch_hook_weight_to_device(hooks=hooks, combined_patches=relevant_patches, key=key, original_weights=original_weights,
                                                             memory_counter=memory_counter)

comfy/sd.py

@@ -940,11 +940,11 @@ def load_diffusion_model(unet_path, model_options={}):
     return model
 
 def load_unet(unet_path, dtype=None):
-    print("WARNING: the load_unet function has been deprecated and will be removed please switch to: load_diffusion_model")
+    logging.warning("The load_unet function has been deprecated and will be removed please switch to: load_diffusion_model")
     return load_diffusion_model(unet_path, model_options={"dtype": dtype})
 
 def load_unet_state_dict(sd, dtype=None):
-    print("WARNING: the load_unet_state_dict function has been deprecated and will be removed please switch to: load_diffusion_model_state_dict")
+    logging.warning("The load_unet_state_dict function has been deprecated and will be removed please switch to: load_diffusion_model_state_dict")
     return load_diffusion_model_state_dict(sd, model_options={"dtype": dtype})
 
 def save_checkpoint(output_path, model, clip=None, vae=None, clip_vision=None, metadata=None, extra_keys={}):

comfy/sd1_clip.py

@@ -41,8 +41,7 @@ class ClipTokenWeightEncoder:
                 to_encode.append(self.gen_empty_tokens(self.special_tokens, max_token_len))
             else:
                 to_encode.append(gen_empty_tokens(self.special_tokens, max_token_len))
-            print(to_encode)
-                
+
         o = self.encode(to_encode)
         out, pooled = o[:2]