search

PaddlePaddle / PaddleMIX

Paddle Multimodal Integration and eXploration, supporting mainstream multi-modal tasks, including end-to-end large-scale multi-modal pretrain models and diffusion model toolbox. Equipped with high performance and flexibility.
Apache License 2.0
353 stars 133 forks source link

huggingface_hub.errors.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name' #811

Open ZhijunLStudio opened 4 days ago

ZhijunLStudio commented 4 days ago

sd3模型加载的时候报错huggingface_hub.errors.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_3'. Use repo_type argument if needed.

下面是相关代码 import paddle from ppdiffusers import StableDiffusion3Pipeline pipe = StableDiffusion3Pipeline.from_pretrained( "stabilityai/stable-diffusion-3-medium-diffusers", paddle_dtype=paddle.float16 ) generator = paddle.Generator().manual_seed(42) prompt = "A cat holding a sign that says hello world" image = pipe(prompt, generator=generator).images[0] image.save("text_to_image_generation-stable_diffusion_3-result.png")

下面是相关包 paddlemix 0.1.0
ppdiffusers 0.29.0 paddlenlp 3.0.0b0.post0 huggingface-hub 0.23.0

以下是具体的报错信息 W1113 16:20:15.825760 50945 gpu_resources.cc:96] The GPU architecture in your current machine is Pascal, which is not compatible with Paddle installation with arch: 61 70 75 80 86 , it is recommended to install the corresponding wheel package according to the installation information on the official Paddle website. W1113 16:20:15.825991 50945 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 6.0, Driver API Version: 12.4, Runtime API Version: 11.4 W1113 16:20:15.830971 50945 gpu_resources.cc:164] device: 0, cuDNN Version: 8.7. (…)sion-3-medium-diffusers/model_index.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 889/889 [00:00<00:00, 36.7kB/s] (…)iffusers/scheduler/scheduler_config.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 138/138 [00:00<00:00, 8.54kB/s] (…)edium-diffusers/text_encoder/config.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 598/598 [00:00<00:00, 84.3kB/s] [2024-11-13 16:20:24,484] [ INFO] - Loading configuration file /home/lizhijun/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers/text_encoder/config.json | 0.00/598 [00:00<?, ?B/s] (…)diffusers/text_encoder/model.safetensors: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 495M/495M [02:34<00:00, 3.19MB/s] [2024-11-13 16:22:59,670] [ INFO] - Loading weights file from cache at /home/lizhijun/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers/text_encoder/model.safetensors█████████████████████████████████| 495M/495M [02:34<00:00, 3.07MB/s] [2024-11-13 16:23:01,193] [ INFO] - Loaded weights file from disk, setting weights to model. [2024-11-13 16:23:04,817] [ INFO] - All model checkpoint weights were used when initializing CLIPTextModelWithProjection.

[2024-11-13 16:23:04,819] [ INFO] - All the weights of CLIPTextModelWithProjection were initialized from the model checkpoint at stabilityai/stable-diffusion-3-medium-diffusers/text_encoder. If your task is similar to the task the model of the checkpoint was trained on, you can already use CLIPTextModelWithProjection for predictions without further training. (…)ium-diffusers/text_encoder_2/config.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 594/594 [00:00<00:00, 93.4kB/s] [2024-11-13 16:23:04,971] [ INFO] - Loading configuration file /home/lizhijun/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_2/config.json | 0.00/594 [00:00<?, ?B/s] (…)ffusers/text_encoder_2/model.safetensors: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2.78G/2.78G [13:56<00:00, 3.32MB/s] [2024-11-13 16:37:01,438] [ INFO] - Loading weights file from cache at /home/lizhijun/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_2/model.safetensors█████████████████████████████| 2.78G/2.78G [13:56<00:00, 3.98MB/s] [2024-11-13 16:37:11,467] [ INFO] - Loaded weights file from disk, setting weights to model. [2024-11-13 16:37:24,492] [ INFO] - All model checkpoint weights were used when initializing CLIPTextModelWithProjection.

[2024-11-13 16:37:24,493] [ INFO] - All the weights of CLIPTextModelWithProjection were initialized from the model checkpoint at stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_2. If your task is similar to the task the model of the checkpoint was trained on, you can already use CLIPTextModelWithProjection for predictions without further training. (…)um-diffusers/text_encoder_3//config.json: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 748/748 [00:00<00:00, 105kB/s] [2024-11-13 16:37:25,460] [ INFO] - Loading configuration file /home/lizhijun/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_3/config.json | 0.00/748 [00:00<?, ?B/s] (…)t_encoder_3/model.safetensors.index.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 20.0k/20.0k [00:00<00:00, 2.26MB/s] [2024-11-13 16:37:25,576] [ INFO] - Loading weights file from cache at /home/lizhijun/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_3/model.safetensors.index.json | 0.00/20.0k [00:00<?, ?B/s] Loading pipeline components...: 33%|█████████████████████████████████████████████████████████████████▎ | 3/9 [17:01<34:03, 340.55s/it] Traceback (most recent call last): File "/home/lizhijun/04.mix/PaddleMIX-2.1.0/sd3_infer.py", line 6, in pipe = StableDiffusion3Pipeline.from_pretrained( File "/home/lizhijun/04.mix/paddlemix2.1.0/ppdiffusers/ppdiffusers/pipelines/pipeline_utils.py", line 1112, in from_pretrained loaded_sub_model = load_sub_model( File "/home/lizhijun/04.mix/paddlemix2.1.0/ppdiffusers/ppdiffusers/pipelines/pipeline_utils.py", line 416, in load_sub_model loaded_sub_model = load_method(cached_folder, **loading_kwargs) File "/home/lizhijun/04.mix/paddlemix2.1.0/ppdiffusers/ppdiffusers/transformers/model_utils.py", line 355, in from_pretrained model = super().from_pretrained( File "/home/lizhijun/04.mix/paddlemix2.1.0/ppdiffusers/ppdiffusers/patches/paddle_patch.py", line 984, in new_from_pretrained return raw_from_pretrained( File "/home/lizhijun/anaconda3/envs/paddlex/lib/python3.9/site-packages/paddlenlp-3.0.0b0.post0-py3.9.egg/paddlenlp/transformers/model_utils.py", line 2213, in from_pretrained resolved_archive_file, resolved_sharded_files, sharded_metadata, is_sharded = cls._resolve_model_file_path( File "/home/lizhijun/anaconda3/envs/paddlex/lib/python3.9/site-packages/paddlenlp-3.0.0b0.post0-py3.9.egg/paddlenlp/transformers/model_utils.py", line 1673, in _resolve_model_file_path resolved_sharded_files, sharded_metadata = get_checkpoint_shard_files( File "/home/lizhijun/anaconda3/envs/paddlex/lib/python3.9/site-packages/paddlenlp-3.0.0b0.post0-py3.9.egg/paddlenlp/transformers/utils.py", line 679, in get_checkpoint_shard_files last_shard = try_to_load_from_cache( File "/home/lizhijun/anaconda3/envs/paddlex/lib/python3.9/site-packages/huggingface_hub/utils/_validators.py", line 106, in _inner_fn validate_repo_id(arg_value) File "/home/lizhijun/anaconda3/envs/paddlex/lib/python3.9/site-packages/huggingface_hub/utils/_validators.py", line 154, in validate_repo_id raise HFValidationError( huggingface_hub.errors.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'stabilityai/stable-diffusion-3-medium-diffusers/text_encoder_3'. Use repo_type argument if needed.

westfish commented 4 days ago

https://github.com/PaddlePaddle/PaddleMIX/blob/668091c82b358a20430da95f3a81d1f0567e507e/ppdiffusers/requirements.txt#L1C1-L1C19

请更新paddlenlp版本

ZhijunLStudio commented 4 days ago

paddlenlp 3.0.0b2.post20241113 ppdiffusers 0.29.0 huggingface-hub 0.23.0 还是报这个错

westfish commented 4 days ago

rm -rf ~/.cache/paddlenlp/ppdiffusers/stabilityai/stable-diffusion-3-medium-diffusers

删除一下缓存，重新下载一下paddle权重，再试试看呢

ZhijunLStudio commented 4 days ago

删了，也还是不行