auto-diffusers 2.0.25.dev0__py3-none-any.whl → 2.0.32__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,4 +1,4 @@
1
- __version__ = "2.0.25.dev0"
1
+ __version__ = "2.0.32"
2
2
 
3
3
  from .pipeline_easy import (
4
4
  search_huggingface,
@@ -16,53 +16,60 @@
16
16
  import os
17
17
  import re
18
18
  import types
19
- from collections import OrderedDict
20
19
  from dataclasses import asdict, dataclass, field
21
20
  from typing import Dict, List, Optional, Union
21
+ from collections import OrderedDict
22
+
22
23
 
23
24
  import requests
24
25
  import torch
26
+
25
27
  from diffusers.loaders.single_file_utils import (
26
28
  VALID_URL_PREFIXES,
27
29
  _extract_repo_id_and_weights_name,
28
30
  infer_diffusers_model_type,
29
31
  load_single_file_checkpoint,
30
32
  )
31
- from diffusers.pipelines.animatediff import AnimateDiffPipeline, AnimateDiffSDXLPipeline
33
+
34
+ from diffusers import (
35
+ StableDiffusionPipeline,
36
+ StableDiffusionImg2ImgPipeline,
37
+ StableDiffusionInpaintPipeline,
38
+ )
32
39
  from diffusers.pipelines.auto_pipeline import (
33
40
  AutoPipelineForImage2Image,
34
41
  AutoPipelineForInpainting,
35
42
  AutoPipelineForText2Image,
43
+ AUTO_TEXT2IMAGE_PIPELINES_MAPPING,
44
+ AUTO_IMAGE2IMAGE_PIPELINES_MAPPING,
45
+ AUTO_INPAINT_PIPELINES_MAPPING,
36
46
  )
37
- from diffusers.pipelines.controlnet import (
47
+ from diffusers.pipelines.pipeline_utils import DiffusionPipeline
48
+ from diffusers.utils import logging
49
+
50
+ from diffusers import (
51
+ AnimateDiffPipeline,
52
+ AnimateDiffSDXLPipeline,
53
+ StableDiffusionControlNetPipeline,
38
54
  StableDiffusionControlNetImg2ImgPipeline,
39
55
  StableDiffusionControlNetInpaintPipeline,
40
- StableDiffusionControlNetPipeline,
41
- StableDiffusionXLControlNetImg2ImgPipeline,
42
- StableDiffusionXLControlNetPipeline,
43
- )
44
- from diffusers.pipelines.flux import FluxImg2ImgPipeline, FluxPipeline
45
- from diffusers.pipelines.pipeline_utils import DiffusionPipeline
46
- from diffusers.pipelines.stable_diffusion import (
47
- StableDiffusionImg2ImgPipeline,
48
- StableDiffusionInpaintPipeline,
49
- StableDiffusionPipeline,
50
- StableDiffusionUpscalePipeline,
51
- )
52
- from diffusers.pipelines.stable_diffusion_3 import (
53
- StableDiffusion3Img2ImgPipeline,
54
- StableDiffusion3Pipeline,
55
- )
56
- from diffusers.pipelines.stable_diffusion_xl import (
57
56
  StableDiffusionXLImg2ImgPipeline,
58
- StableDiffusionXLInpaintPipeline,
59
57
  StableDiffusionXLPipeline,
58
+ StableDiffusionXLInpaintPipeline,
59
+ StableDiffusionUpscalePipeline,
60
+ StableDiffusion3Pipeline,
61
+ StableDiffusion3Img2ImgPipeline,
62
+ FluxPipeline,
63
+ FluxImg2ImgPipeline,
64
+ StableDiffusionXLControlNetPipeline,
65
+ StableDiffusionXLControlNetImg2ImgPipeline,
60
66
  )
61
- from diffusers.utils import logging
67
+
62
68
  from huggingface_hub import hf_api, hf_hub_download
63
69
  from huggingface_hub.file_download import http_get
64
70
  from huggingface_hub.utils import validate_hf_hub_args
65
71
 
72
+
66
73
  logger = logging.get_logger(__name__)
67
74
 
68
75
 
@@ -107,9 +114,14 @@ SINGLE_FILE_CHECKPOINT_TEXT2IMAGE_PIPELINE_MAPPING = OrderedDict(
107
114
  ("xl_base", StableDiffusionXLPipeline),
108
115
  ("xl_inpaint", None),
109
116
  ("xl_refiner", StableDiffusionXLPipeline),
110
- ]
117
+ ],
111
118
  )
112
119
 
120
+ # For single-file checkpoints we expose separate mappings for each task so callers
121
+ # can pick the most appropriate pipeline class set. The full, explicit mappings
122
+ # for image2image and inpaint are defined below; keep distinct names so they can
123
+ # be customized independently.
124
+
113
125
  SINGLE_FILE_CHECKPOINT_IMAGE2IMAGE_PIPELINE_MAPPING = OrderedDict(
114
126
  [
115
127
  ("animatediff_rgb", AnimateDiffPipeline),
@@ -198,7 +210,6 @@ SINGLE_FILE_CHECKPOINT_INPAINT_PIPELINE_MAPPING = OrderedDict(
198
210
  ]
199
211
  )
200
212
 
201
-
202
213
  CONFIG_FILE_LIST = [
203
214
  "pytorch_model.bin",
204
215
  "pytorch_model.fp16.bin",
@@ -213,11 +224,23 @@ CONFIG_FILE_LIST = [
213
224
  ]
214
225
 
215
226
  DIFFUSERS_CONFIG_DIR = [
227
+ # common pipeline component folders
216
228
  "safety_checker",
217
229
  "unet",
218
230
  "vae",
231
+ "vae_1_0",
232
+ "vae_decoder",
233
+ "vae_encoder",
219
234
  "text_encoder",
220
235
  "text_encoder_2",
236
+ "tokenizer",
237
+ "tokenizer_2",
238
+ "scheduler",
239
+ "transformer",
240
+ # legacy / other possible component folders
241
+ "controlnet",
242
+ "clip",
243
+ "encoder",
221
244
  ]
222
245
 
223
246
  TOKENIZER_SHAPE_MAP = {
@@ -244,6 +267,12 @@ EXTENSION = [".safetensors", ".ckpt", ".bin"]
244
267
  CACHE_HOME = os.path.expanduser("~/.cache")
245
268
 
246
269
 
270
+ # Set up pipeline mappings
271
+ AUTO_TEXT2IMAGE_PIPELINES_MAPPING["v1"] = StableDiffusionPipeline
272
+ AUTO_IMAGE2IMAGE_PIPELINES_MAPPING["v1"] = StableDiffusionImg2ImgPipeline
273
+ AUTO_INPAINT_PIPELINES_MAPPING["v1"] = StableDiffusionInpaintPipeline
274
+
275
+
247
276
  @dataclass
248
277
  class RepoStatus:
249
278
  r"""
@@ -394,13 +423,12 @@ def load_pipeline_from_single_file(
394
423
  if pipeline_class is None:
395
424
  raise ValueError(
396
425
  f"{model_type} is not supported in this pipeline."
397
- "For `Text2Image`, please use `AutoPipelineForText2Image.from_pretrained`, "
398
- "for `Image2Image` , please use `AutoPipelineForImage2Image.from_pretrained`, "
399
- "and `inpaint` is only supported in `AutoPipelineForInpainting.from_pretrained`"
426
+ "For `Text2Image`, please use `EasyPipelineForText2Image.from_pretrained`, "
427
+ "for `Image2Image` , please use `EasyPipelineForImage2Image.from_pretrained`, "
428
+ "and `inpaint` is only supported in `EasyPipelineForInpainting.from_pretrained`"
400
429
  )
401
430
 
402
431
  else:
403
- # Instantiate and return the pipeline with the loaded checkpoint and any additional kwargs
404
432
  return pipeline_class.from_single_file(pretrained_model_or_path, **kwargs)
405
433
 
406
434
 
@@ -529,19 +557,40 @@ def file_downloader(
529
557
 
530
558
  # Check if the file already exists at the save path
531
559
  if os.path.exists(save_path):
560
+ existing_size = os.path.getsize(save_path)
561
+ # If file exists and not forcing download
532
562
  if not force_download:
533
- # If the file exists and force_download is False, skip the download
534
- logger.info(f"File already exists: {save_path}, skipping download.")
535
- return None
536
- elif resume:
537
- # If resuming, set mode to append binary and get current file size
563
+ # If the file is non-empty, assume it's already downloaded
564
+ if existing_size > 0:
565
+ logger.info(f"File already exists and is non-empty: {save_path}, skipping download.")
566
+ return None
567
+ # If the file is zero bytes, remove it and proceed to download
568
+ logger.info(f"Found zero-byte file, removing and retrying download: {save_path}")
569
+ try:
570
+ os.remove(save_path)
571
+ except OSError:
572
+ # If removal fails, try to proceed by truncating the file
573
+ with open(save_path, "wb"):
574
+ pass
575
+ # If resuming is requested and file exists, append
576
+ if resume and os.path.exists(save_path):
538
577
  mode = "ab"
539
578
  file_size = os.path.getsize(save_path)
540
579
 
580
+ # Check if the URL is accessible before downloading. If HEAD returns 401,
581
+ # raise an HTTPError so callers can try the next candidate. Other HEAD
582
+ # failures are logged and the download is still attempted.
583
+ try:
584
+ response = requests.head(url, headers=headers, allow_redirects=True, timeout=2)
585
+ if response.status_code == 401:
586
+ raise requests.HTTPError(f"401 Unauthorized: {url}")
587
+ except requests.exceptions.RequestException as e:
588
+ logger.info(f"HEAD check failed (continuing to download): {url} -> {e}")
589
+
541
590
  # Open the file in the appropriate mode (write or append)
542
591
  with open(save_path, mode) as model_file:
543
592
  # Call the http_get function to perform the file download
544
- return http_get(
593
+ result = http_get(
545
594
  url=url,
546
595
  temp_file=model_file,
547
596
  resume_size=file_size,
@@ -549,6 +598,8 @@ def file_downloader(
549
598
  headers=headers,
550
599
  proxies=proxies,
551
600
  )
601
+ logger.info(f"Successfully downloaded: {url} -> {save_path}")
602
+ return result
552
603
 
553
604
 
554
605
  def search_huggingface(search_word: str, **kwargs) -> Union[str, SearchResult, None]:
@@ -582,6 +633,12 @@ def search_huggingface(search_word: str, **kwargs) -> Union[str, SearchResult, N
582
633
  """
583
634
  # Extract additional parameters from kwargs
584
635
  revision = kwargs.pop("revision", None)
636
+ model_type = kwargs.pop("model_type", "checkpoint")
637
+ # normalize model_type for simple comparisons
638
+ try:
639
+ model_type = str(model_type).lower()
640
+ except Exception:
641
+ model_type = ""
585
642
  checkpoint_format = kwargs.pop("checkpoint_format", "single_file")
586
643
  download = kwargs.pop("download", False)
587
644
  force_download = kwargs.pop("force_download", False)
@@ -603,7 +660,9 @@ def search_huggingface(search_word: str, **kwargs) -> Union[str, SearchResult, N
603
660
  search_word_status = get_keyword_types(search_word)
604
661
 
605
662
  if search_word_status["type"]["hf_repo"]:
606
- hf_repo_info = hf_api.model_info(repo_id=search_word, securityStatus=True)
663
+ hf_repo_info = asdict(
664
+ hf_api.model_info(repo_id=search_word, securityStatus=True)
665
+ )
607
666
  if download:
608
667
  model_path = DiffusionPipeline.download(
609
668
  search_word,
@@ -654,9 +713,11 @@ def search_huggingface(search_word: str, **kwargs) -> Union[str, SearchResult, N
654
713
  for repo_info in model_dicts:
655
714
  repo_id = repo_info["id"]
656
715
  file_list = []
657
- hf_repo_info = hf_api.model_info(repo_id=repo_id, securityStatus=True)
716
+ hf_repo_info = asdict(
717
+ hf_api.model_info(repo_id=repo_id, securityStatus=True)
718
+ )
658
719
  # Lists files with security issues.
659
- hf_security_info = hf_repo_info.security_repo_status
720
+ hf_security_info = hf_repo_info["security_repo_status"]
660
721
  exclusion = [issue["path"] for issue in hf_security_info["filesWithIssues"]]
661
722
 
662
723
  # Checks for multi-folder diffusers model or valid files (models with security issues are excluded).
@@ -674,8 +735,12 @@ def search_huggingface(search_word: str, **kwargs) -> Union[str, SearchResult, N
674
735
  any(file_path.endswith(ext) for ext in EXTENSION)
675
736
  and not any(config in file_path for config in CONFIG_FILE_LIST)
676
737
  and not any(exc in file_path for exc in exclusion)
677
- and os.path.basename(os.path.dirname(file_path))
678
- not in DIFFUSERS_CONFIG_DIR
738
+ and os.path.basename(os.path.dirname(file_path)) not in DIFFUSERS_CONFIG_DIR
739
+ # If caller requested checkpoint-only, exclude filenames that look like LoRA/Textual-Inversion
740
+ and not (
741
+ model_type == "checkpoint"
742
+ and any(kw in file_path.lower() for kw in ["lora", "textual", "inversion"])
743
+ )
679
744
  ):
680
745
  file_list.append(file_path)
681
746
 
@@ -740,7 +805,7 @@ def search_huggingface(search_word: str, **kwargs) -> Union[str, SearchResult, N
740
805
  loading_method=output_info["loading_method"],
741
806
  checkpoint_format=output_info["checkpoint_format"],
742
807
  repo_status=RepoStatus(
743
- repo_id=repo_id, repo_hash=hf_repo_info.sha, version=revision
808
+ repo_id=repo_id, repo_hash=hf_repo_info["sha"], version=revision
744
809
  ),
745
810
  model_status=ModelStatus(
746
811
  search_word=search_word,
@@ -816,7 +881,7 @@ def search_civitai(search_word: str, **kwargs) -> Union[str, SearchResult, None]
816
881
  params = {
817
882
  "query": search_word,
818
883
  "types": model_type,
819
- "limit": 20,
884
+ "limit": 100,
820
885
  }
821
886
  if base_model is not None:
822
887
  if not isinstance(base_model, list):
@@ -837,7 +902,10 @@ def search_civitai(search_word: str, **kwargs) -> Union[str, SearchResult, None]
837
902
  )
838
903
  response.raise_for_status()
839
904
  except requests.exceptions.HTTPError as err:
840
- raise requests.HTTPError(f"Could not get elements from the URL: {err}")
905
+ if skip_error:
906
+ return None
907
+ else:
908
+ raise requests.HTTPError(f"Could not get elements from the URL: {err}")
841
909
  else:
842
910
  try:
843
911
  data = response.json()
@@ -848,9 +916,12 @@ def search_civitai(search_word: str, **kwargs) -> Union[str, SearchResult, None]
848
916
  raise ValueError("Invalid JSON response")
849
917
 
850
918
  # Sort repositories by download count in descending order
851
- sorted_repos = sorted(
852
- data["items"], key=lambda x: x["stats"]["downloadCount"], reverse=True
853
- )
919
+ # sorted_repos = sorted(
920
+ # data["items"], key=lambda x: x["stats"]["downloadCount"], reverse=True
921
+ # )
922
+
923
+ # Since the Civitai API is broken, I will temporarily sort by name.
924
+ sorted_repos = sorted(data["items"], key=lambda x: x["name"], reverse=True)
854
925
 
855
926
  for selected_repo in sorted_repos:
856
927
  repo_name = selected_repo["name"]
@@ -889,7 +960,8 @@ def search_civitai(search_word: str, **kwargs) -> Union[str, SearchResult, None]
889
960
  sorted_models = sorted(
890
961
  models_list, key=lambda x: x["filename"], reverse=True
891
962
  )
892
- selected_model = next(
963
+ # Prefer safe/sfw models
964
+ candidate_model = next(
893
965
  (
894
966
  model_data
895
967
  for model_data in sorted_models
@@ -899,11 +971,27 @@ def search_civitai(search_word: str, **kwargs) -> Union[str, SearchResult, None]
899
971
  ),
900
972
  sorted_models[0],
901
973
  )
974
+
975
+ # Validate download URL accessibility (check first file only)
976
+ try:
977
+ # Since models requiring authentication do not return results, there is no need to determine their status via status codes.
978
+ _resp = requests.head(
979
+ candidate_model["download_url"],
980
+ headers=headers,
981
+ timeout=2,
982
+ allow_redirects=True
983
+ )
984
+ selected_model = candidate_model
985
+ logger.info(f"Validated download URL: {candidate_model['download_url']}")
986
+ break # Exit the version loop if a valid model is found
987
+ except requests.exceptions.RequestException as e:
988
+ logger.info(f"Failed to validate URL {candidate_model['download_url']}: {e}, trying next version...")
989
+ continue
990
+
991
+ # If we found a valid model in this repo, exit outer repo loop
992
+ if selected_model:
993
+ break
902
994
 
903
- break
904
- else:
905
- continue
906
- break
907
995
 
908
996
  # Exception handling when search candidates are not found
909
997
  if not selected_model:
@@ -920,20 +1008,45 @@ def search_civitai(search_word: str, **kwargs) -> Union[str, SearchResult, None]
920
1008
 
921
1009
  # Handle file download and setting model information
922
1010
  if download:
923
- # The path where the model is to be saved.
924
- model_path = os.path.join(
925
- str(civitai_cache_dir), str(repo_id), str(version_id), str(file_name)
926
- )
927
- # Download Model File
928
- file_downloader(
929
- url=download_url,
930
- save_path=model_path,
931
- resume=resume,
932
- force_download=force_download,
933
- displayed_filename=file_name,
934
- headers=headers,
935
- **kwargs,
1011
+ # Try downloading from the models list, retrying on 401 errors
1012
+ download_success = False
1013
+ # Create a sorted list of all available models to try
1014
+ sorted_models = sorted(
1015
+ models_list, key=lambda x: x["filename"], reverse=True
936
1016
  )
1017
+
1018
+ for model_candidate in sorted_models:
1019
+ try:
1020
+ file_name = model_candidate["filename"]
1021
+ download_url = model_candidate["download_url"]
1022
+
1023
+ # The path where the model is to be saved.
1024
+ model_path = os.path.join(
1025
+ str(civitai_cache_dir), str(repo_id), str(version_id), str(file_name)
1026
+ )
1027
+
1028
+ # Download Model File
1029
+ file_downloader(
1030
+ url=download_url,
1031
+ save_path=model_path,
1032
+ resume=resume,
1033
+ force_download=force_download,
1034
+ displayed_filename=file_name,
1035
+ headers=headers,
1036
+ **kwargs,
1037
+ )
1038
+ download_success = True
1039
+ logger.info(f"Successfully downloaded model from: {download_url}")
1040
+ break
1041
+ except requests.HTTPError as e:
1042
+ if "401" in str(e):
1043
+ logger.info(f"401 error for {download_url}, trying next candidate...")
1044
+ continue
1045
+ else:
1046
+ raise
1047
+
1048
+ if not download_success:
1049
+ raise ValueError(f"Failed to download any model file from {repo_name}. All URLs returned 401 errors.")
937
1050
 
938
1051
  else:
939
1052
  model_path = download_url
@@ -1113,12 +1226,12 @@ class AutoConfig:
1113
1226
  # Search for the model on Civitai and get the model status
1114
1227
  textual_inversion_path = search_civitai(search_word, **kwargs)
1115
1228
  logger.warning(
1116
- f"textual_inversion_path: {search_word} -> {textual_inversion_path.model_status.site_url}"
1229
+ f"textual_inversion_path: {search_word} -> {textual_inversion_path.model_status.site_url}" # type: ignore
1117
1230
  )
1118
1231
 
1119
1232
  pretrained_model_name_or_paths[
1120
1233
  pretrained_model_name_or_paths.index(search_word)
1121
- ] = textual_inversion_path.model_path
1234
+ ] = textual_inversion_path.model_path # type: ignore
1122
1235
 
1123
1236
  self.load_textual_inversion(
1124
1237
  pretrained_model_name_or_paths,
@@ -1132,6 +1245,7 @@ class AutoConfig:
1132
1245
  self,
1133
1246
  pretrained_model_name_or_path_or_dict: Union[str, Dict[str, torch.Tensor]],
1134
1247
  adapter_name=None,
1248
+ base_model: Optional[Union[str, List[str]]] = None,
1135
1249
  **kwargs,
1136
1250
  ):
1137
1251
  r"""
@@ -1155,6 +1269,9 @@ class AutoConfig:
1155
1269
  adapter_name (`str`, *optional*):
1156
1270
  Adapter name to be used for referencing the loaded adapter model. If not specified, it will use
1157
1271
  `default_{i}` where i is the total number of adapters being loaded.
1272
+ base_model (`str` or `List[str]`, *optional*):
1273
+ Base model tag(s) to filter LoRA models (e.g., "SD 1.5", "SDXL 1.0"). Used when searching Civitai
1274
+ to ensure LoRA compatibility with the current pipeline's base model.
1158
1275
  low_cpu_mem_usage (`bool`, *optional*):
1159
1276
  Speed up model loading by only loading the pretrained LoRA weights and not initializing the random
1160
1277
  weights.
@@ -1169,12 +1286,24 @@ class AutoConfig:
1169
1286
  "skip_error": False,
1170
1287
  "model_type": "LORA",
1171
1288
  }
1289
+ if base_model is not None:
1290
+ _status["base_model"] = base_model if isinstance(base_model, list) else [base_model]
1172
1291
  kwargs.update(_status)
1173
1292
  # Search for the model on Civitai and get the model status
1174
1293
  lora_path = search_civitai(pretrained_model_name_or_path_or_dict, **kwargs)
1175
- logger.warning(f"lora_path: {lora_path.model_status.site_url}")
1176
- logger.warning(f"trained_words: {lora_path.extra_status.trained_words}")
1177
- pretrained_model_name_or_path_or_dict = lora_path.model_path
1294
+ # search_civitai may return a SearchResult, a string (download url or path) or None.
1295
+ if isinstance(lora_path, SearchResult):
1296
+ logger.warning(f"lora_path: {lora_path.model_status.site_url}")
1297
+ logger.warning(f"trained_words: {lora_path.extra_status.trained_words}")
1298
+ pretrained_model_name_or_path_or_dict = lora_path.model_path
1299
+ elif lora_path is None:
1300
+ raise ValueError(
1301
+ f"No LORA model found for {pretrained_model_name_or_path_or_dict}"
1302
+ )
1303
+ else:
1304
+ # fallback when search_civitai returns a plain path/URL string
1305
+ logger.warning(f"lora_path: {lora_path}")
1306
+ pretrained_model_name_or_path_or_dict = lora_path
1178
1307
 
1179
1308
  self.load_lora_weights(
1180
1309
  pretrained_model_name_or_path_or_dict, adapter_name=adapter_name, **kwargs
@@ -1320,12 +1449,12 @@ class EasyPipelineForText2Image(AutoPipelineForText2Image):
1320
1449
  pretrained_model_link_or_path, **kwargs
1321
1450
  )
1322
1451
  logger.warning(
1323
- f"checkpoint_path: {hf_checkpoint_status.model_status.download_url}"
1452
+ f"checkpoint_path: {hf_checkpoint_status.model_status.download_url}" # type: ignore
1324
1453
  )
1325
- checkpoint_path = hf_checkpoint_status.model_path
1454
+ checkpoint_path = hf_checkpoint_status.model_path # type: ignore
1326
1455
 
1327
1456
  # Check the format of the model checkpoint
1328
- if hf_checkpoint_status.loading_method == "from_single_file":
1457
+ if hf_checkpoint_status.loading_method == "from_single_file": # type: ignore
1329
1458
  # Load the pipeline from a single file checkpoint
1330
1459
  pipeline = load_pipeline_from_single_file(
1331
1460
  pretrained_model_or_path=checkpoint_path,
@@ -1430,8 +1559,8 @@ class EasyPipelineForText2Image(AutoPipelineForText2Image):
1430
1559
 
1431
1560
  # Search for the model on Civitai and get the model status
1432
1561
  checkpoint_status = search_civitai(pretrained_model_link_or_path, **kwargs)
1433
- logger.warning(f"checkpoint_path: {checkpoint_status.model_status.site_url}")
1434
- checkpoint_path = checkpoint_status.model_path
1562
+ logger.warning(f"checkpoint_path: {checkpoint_status.model_status.site_url}") # type: ignore
1563
+ checkpoint_path = checkpoint_status.model_path # type: ignore
1435
1564
 
1436
1565
  # Load the pipeline from a single file checkpoint
1437
1566
  pipeline = load_pipeline_from_single_file(
@@ -1582,12 +1711,12 @@ class EasyPipelineForImage2Image(AutoPipelineForImage2Image):
1582
1711
  pretrained_model_link_or_path, **kwargs
1583
1712
  )
1584
1713
  logger.warning(
1585
- f"checkpoint_path: {hf_checkpoint_status.model_status.download_url}"
1714
+ f"checkpoint_path: {hf_checkpoint_status.model_status.download_url}" # type: ignore
1586
1715
  )
1587
- checkpoint_path = hf_checkpoint_status.model_path
1716
+ checkpoint_path = hf_checkpoint_status.model_path # type: ignore
1588
1717
 
1589
1718
  # Check the format of the model checkpoint
1590
- if hf_checkpoint_status.loading_method == "from_single_file":
1719
+ if hf_checkpoint_status.loading_method == "from_single_file": # type: ignore
1591
1720
  # Load the pipeline from a single file checkpoint
1592
1721
  pipeline = load_pipeline_from_single_file(
1593
1722
  pretrained_model_or_path=checkpoint_path,
@@ -1693,8 +1822,8 @@ class EasyPipelineForImage2Image(AutoPipelineForImage2Image):
1693
1822
 
1694
1823
  # Search for the model on Civitai and get the model status
1695
1824
  checkpoint_status = search_civitai(pretrained_model_link_or_path, **kwargs)
1696
- logger.warning(f"checkpoint_path: {checkpoint_status.model_status.site_url}")
1697
- checkpoint_path = checkpoint_status.model_path
1825
+ logger.warning(f"checkpoint_path: {checkpoint_status.model_status.site_url}") # type: ignore
1826
+ checkpoint_path = checkpoint_status.model_path # type: ignore
1698
1827
 
1699
1828
  # Load the pipeline from a single file checkpoint
1700
1829
  pipeline = load_pipeline_from_single_file(
@@ -1845,12 +1974,12 @@ class EasyPipelineForInpainting(AutoPipelineForInpainting):
1845
1974
  pretrained_model_link_or_path, **kwargs
1846
1975
  )
1847
1976
  logger.warning(
1848
- f"checkpoint_path: {hf_checkpoint_status.model_status.download_url}"
1977
+ f"checkpoint_path: {hf_checkpoint_status.model_status.download_url}" # type: ignore
1849
1978
  )
1850
- checkpoint_path = hf_checkpoint_status.model_path
1979
+ checkpoint_path = hf_checkpoint_status.model_path # type: ignore
1851
1980
 
1852
1981
  # Check the format of the model checkpoint
1853
- if hf_checkpoint_status.loading_method == "from_single_file":
1982
+ if hf_checkpoint_status.loading_method == "from_single_file": # type: ignore
1854
1983
  # Load the pipeline from a single file checkpoint
1855
1984
  pipeline = load_pipeline_from_single_file(
1856
1985
  pretrained_model_or_path=checkpoint_path,
@@ -1955,8 +2084,8 @@ class EasyPipelineForInpainting(AutoPipelineForInpainting):
1955
2084
 
1956
2085
  # Search for the model on Civitai and get the model status
1957
2086
  checkpoint_status = search_civitai(pretrained_model_link_or_path, **kwargs)
1958
- logger.warning(f"checkpoint_path: {checkpoint_status.model_status.site_url}")
1959
- checkpoint_path = checkpoint_status.model_path
2087
+ logger.warning(f"checkpoint_path: {checkpoint_status.model_status.site_url}") # type: ignore
2088
+ checkpoint_path = checkpoint_status.model_path # type: ignore
1960
2089
 
1961
2090
  # Load the pipeline from a single file checkpoint
1962
2091
  pipeline = load_pipeline_from_single_file(
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: auto_diffusers
3
- Version: 2.0.25.dev0
3
+ Version: 2.0.32
4
4
  Summary: diffusers with search engine
5
5
  Home-page: https://github.com/suzukimain/auto_diffusers
6
6
  Author: suzukimain
@@ -23,8 +23,7 @@ License-File: LICENSE
23
23
  Requires-Dist: diffusers >=0.32.0
24
24
  Requires-Dist: transformers >=4.41.2
25
25
 
26
- # auto_diffusers
27
-
26
+ ![auto_diffusers icon](https://github.com/suzukimain/auto_diffusers/blob/master/images/auto_diffusers_icon.png)
28
27
 
29
28
  <p>
30
29
  <a href="https://github.com/suzukimain/auto_diffusers/blob/main/LICENSE"><img alt="GitHub release" src="https://img.shields.io/badge/license-Apache%202.0-blue.svg?style=social"></a>
@@ -45,10 +44,20 @@ Requires-Dist: transformers >=4.41.2
45
44
  + [License](#License)
46
45
  + [Acknowledgement](#Acknowledgement)
47
46
 
47
+
48
48
  ## About The Project<a name = "About_The_Project"></a>
49
- Enhance the functionality of diffusers.
50
- * Search models from huggingface and Civitai.
51
49
 
50
+ Enhance the functionality of diffusers.
51
+ * Search models from Hugging Face and Civitai.
52
+ * Officially listed in the [Hugging Face Diffusers Community Projects](https://huggingface.co/docs/diffusers/community_projects) documentation.
53
+
54
+ <details>
55
+ <summary>Image of the listing location on Hugging Face</summary>
56
+
57
+ [<img src="images\huggingface_highlight.png" width="420">](https://github.com/suzukimain/auto_diffusers/blob/master/images/huggingface_highlight.png)
58
+
59
+
60
+ </details>
52
61
 
53
62
  ## How to use<a name = "How_to_use"></a>
54
63
 
@@ -0,0 +1,8 @@
1
+ auto_diffusers/__init__.py,sha256=c8RJ0wSpvziBqK5jq_Y9EOYkPXsEXF_eDnVyXu5jv0I,228
2
+ auto_diffusers/pipeline_easy.py,sha256=mVpapjENM-Djpmzdk_6r8e3ddZY2Jne4lM5AwWwnCGM,99561
3
+ auto_diffusers/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
4
+ auto_diffusers-2.0.32.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
5
+ auto_diffusers-2.0.32.dist-info/METADATA,sha256=Y4IZYwkNICxHKbVA8px9E70RbuEevJzz_XQ30WSJszE,10337
6
+ auto_diffusers-2.0.32.dist-info/WHEEL,sha256=iAkIy5fosb7FzIOwONchHf19Qu7_1wCWyFNR5gu9nU0,91
7
+ auto_diffusers-2.0.32.dist-info/top_level.txt,sha256=iYkbnc9T-VKVnoCP6AufdXcDUfUaLZ_yMG-USuYf4UQ,15
8
+ auto_diffusers-2.0.32.dist-info/RECORD,,
@@ -1,8 +0,0 @@
1
- auto_diffusers/__init__.py,sha256=XKmoQRuUnVyCjIz95HUrWXU83xGSbupM84mf-tEzVt0,233
2
- auto_diffusers/pipeline_easy.py,sha256=5lIEp864CMvBGZKSf2U_WQIC40USb7ebIlhsYlI5baA,93488
3
- auto_diffusers/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
4
- auto_diffusers-2.0.25.dev0.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
5
- auto_diffusers-2.0.25.dev0.dist-info/METADATA,sha256=j6mUUB7jJtPZBsBFrVw-utky4Uh0R5WN725S73ZFANM,9858
6
- auto_diffusers-2.0.25.dev0.dist-info/WHEEL,sha256=iAkIy5fosb7FzIOwONchHf19Qu7_1wCWyFNR5gu9nU0,91
7
- auto_diffusers-2.0.25.dev0.dist-info/top_level.txt,sha256=iYkbnc9T-VKVnoCP6AufdXcDUfUaLZ_yMG-USuYf4UQ,15
8
- auto_diffusers-2.0.25.dev0.dist-info/RECORD,,