Newt Current

This notebook shows the current state of the newt re-identification models on our dataset.

import os
import sys
import torch
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from wildlife_tools.similarity import CosineSimilarity
from wildlife_datasets import analysis, datasets, splits
import pycocotools.mask as mask_util
from wildlife_tools.data import ImageDataset
from sklearn.metrics import average_precision_score
import numpy as np
import timm
from transformers import AutoModel
import torch
import numpy as np
from wildlife_tools.inference import TopkClassifier, KnnClassifier
from wildlife_tools.features import DeepFeatures
import torchvision.transforms as T
from PIL import Image
import kaggle
import pandas as pd
from wildlife_tools.data import ImageDataset
from gcn_reid.segmentation import decode_rle_mask

Create Dataset Class

---

source

get_newt_dataset

 get_newt_dataset ()

---

source

get_cropped_newt

 get_cropped_newt (path, rle)

---

source

get_cropping_image_dataset

 get_cropping_image_dataset ()

CroppingImageDataset = get_cropping_image_dataset()

Dataset already exists at data/newt_dataset

	path	image_name	identity	segmentation_mask_rle
0	original_images/GCN63-P6-S2/IMG_2725.JPEG	IMG_2725.JPEG	GCN63-P6-S2	2048x1536:bjlS15go17J6oIGT\N<`c1O^\N7]c1Ib\N>X...
1	original_images/GCN63-P6-S2/IMG_2727.JPEG	IMG_2727.JPEG	GCN63-P6-S2	2048x1536:XljT17do17N2M2THEQ`N<k_1KP`N=h_1FU`N...
2	original_images/GCN63-P6-S2/IMG_2728.JPEG	IMG_2728.JPEG	GCN63-P6-S2	2048x1536:chln02mo12N2M3E<M3N1OO4CQS2LXmM6L5K5...
3	original_images/GCN63-P6-S2/IMG_2726.JPEG	IMG_2726.JPEG	GCN63-P6-S2	2048x1536:l]\P1U1=]O4KVl1R1ZSN^O:IVl1m2K6[TNYL...
4	original_images/GCN63-P6-S2/IMG_2729.JPEG	IMG_2729.JPEG	GCN63-P6-S2	2048x1536:S]`92no11OO10ho31VPL3_PNKVo1=VOGQRN7...

dataset_path = "data/newt_dataset"
NewtDataset = get_newt_dataset()
NewtDataset._download(dataset_name="mshahoyi/barhill-newts-segmented", download_path=dataset_path)
dataset = NewtDataset(dataset_path)
dataset.df.head()

dataset.plot_grid()

analysis.display_statistics(dataset.df)

Number of identitites            207
Number of all animals            1253
Number of animals with one image 0
Number of unidentified animals   0

Create Query and Database Sets

splitter = splits.ClosedSetSplit(0.9)
for idx_database, idx_query in splitter.split(dataset.df):
    df_database, df_query = dataset.df.loc[idx_database], dataset.df.loc[idx_query]
    splits.analyze_split(dataset.df, idx_database, idx_query)

Split: time-unaware closed-set
Samples: train/test/unassigned/total = 1043/210/0/1253
Classes: train/test/unassigned/total = 207/207/0/207
Samples: train only/test only        = 0/0
Classes: train only/test only/joint  = 0/0/207

Fraction of train set     = 83.24%
Fraction of test set only = 0.00%

Test MegaDescriptor

transform = T.Compose([T.Resize([384, 384]), T.ToTensor(), T.Normalize(mean=(0.485, 0.456, 0.406), std=(0.229, 0.224, 0.225))])
dataset_database = datasets.WildlifeDataset(df=df_database, root=dataset.root, transform=transform)
dataset_query = datasets.WildlifeDataset(df=df_query, root=dataset.root, transform=transform)

name = 'hf-hub:BVRA/MegaDescriptor-L-384'
device = torch.device('cuda:1' if torch.cuda.is_available() else 'cpu')
extractor = DeepFeatures(timm.create_model(name, num_classes=0, pretrained=True), 
                         device=device,
                         batch_size=32,
                         num_workers=4,
                         )
print("model loaded to device:", device)

2025-05-27 17:23:22.237805: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:477] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1748366602.260419   47612 cuda_dnn.cc:8310] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1748366602.267161   47612 cuda_blas.cc:1418] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered

model loaded to device: cuda:1

query = extractor(CroppingImageDataset(dataset_query.df, root=dataset_query.root, transform=dataset_query.transform, crop_out=False))
database = extractor(CroppingImageDataset(dataset_database.df, root=dataset_database.root, transform=dataset_database.transform, crop_out=False))

100%|█████████████████████████████████████████████████████████████████| 7/7 [00:18<00:00,  2.63s/it]
100%|███████████████████████████████████████████████████████████████| 33/33 [01:30<00:00,  2.75s/it]

similarity_function = CosineSimilarity()
similarity = similarity_function(query, database)

top_5_classifier = TopkClassifier(k=5, database_labels=dataset_database.labels_string, return_all=True)
predictions_top_5, scores_top_5, _ = top_5_classifier(similarity)

accuracy_top_1 = np.mean(dataset_query.labels_string == predictions_top_5[:, 0])
accuracy_top_5 = np.mean(np.any(predictions_top_5 == dataset_query.labels_string[:, np.newaxis], axis=1))

accuracy_top_1, accuracy_top_5

(0.8523809523809524, 0.8523809523809524, 0.9714285714285714)

def calculate_map(query_labels, database_labels, similarity_matrix):
    """
    Calculate mean Average Precision (mAP) for retrieval task.
    
    Args:
        query_labels: Array of query labels
        database_labels: Array of database labels  
        similarity_matrix: Similarity scores between queries and database
    
    Returns:
        mAP: Mean Average Precision
    """
    aps = []
    
    for i, query_label in enumerate(query_labels):
        # Get similarity scores for this query
        scores = similarity_matrix[i]
        
        # Create binary relevance labels (1 if same identity, 0 otherwise)
        relevance = (database_labels == query_label).astype(int)
        
        # Calculate Average Precision for this query
        if np.sum(relevance) > 0:  # Only if there are relevant items
            ap = average_precision_score(relevance, scores)
            aps.append(ap)
    
    return np.mean(aps)

# Calculate mAP
map_score = calculate_map(dataset_query.labels_string, dataset_database.labels_string, similarity)
print(f"Mean Average Precision (mAP): {map_score:.4f}")

Mean Average Precision (mAP): 0.6209

---

source

plot_retrieval_results

 plot_retrieval_results (dataset_query, dataset_database,
                         similarity_matrix, crop_out=False,
                         mode='mistakes', num_results=4, num_queries=5,
                         figsize=(15, 20))

*Plot retrieval results showing query images and their most similar matches.

Args: dataset_query: Query dataset with images and labels dataset_database: Database dataset with images and labels similarity_matrix: Similarity scores between queries and database num_results: Number of top similar images to show per query num_queries: Number of query images to display figsize: Figure size for the plot*

# Plot retrieval results
plot_retrieval_results(dataset_query, dataset_database, similarity, num_results=4, num_queries=5)

Test MiewID

miew_id_model = AutoModel.from_pretrained("conservationxlabs/miewid-msv2", trust_remote_code=True)

miew_id_extractor = DeepFeatures(miew_id_model, 
                         device=device,
                         batch_size=32,
                         num_workers=4,
                         )

Building Model Backbone for efficientnetv2_rw_m model
config.model_name efficientnetv2_rw_m
model_name efficientnetv2_rw_m

/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for conv_stem.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.0.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.0.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.0.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.0.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.0.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.0.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.1.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.1.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.1.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.1.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.1.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.1.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.2.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.2.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.2.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.2.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.2.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.0.2.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.0.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.0.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.0.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.0.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.0.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.0.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.1.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.1.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.1.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.1.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.1.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.1.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.2.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.2.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.2.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.2.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.2.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.2.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.3.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.3.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.3.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.3.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.3.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.3.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.4.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.4.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.4.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.4.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.4.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.1.4.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.0.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.0.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.0.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.0.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.0.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.0.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.1.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.1.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.1.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.1.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.1.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.1.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.2.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.2.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.2.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.2.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.2.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.2.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.3.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.3.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.3.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.3.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.3.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.3.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.4.conv_exp.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.4.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.4.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.4.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.4.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.2.4.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.0.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.1.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.2.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.3.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.4.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.5.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.6.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.3.7.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.0.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.1.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.2.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.3.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.4.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.5.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.6.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.7.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.8.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.9.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.10.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.11.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.12.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.13.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.4.14.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.0.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.1.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.2.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.3.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.4.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.5.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.6.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.7.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.8.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.9.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.10.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.11.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.12.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.13.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.14.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.15.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.16.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.17.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.18.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.19.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.20.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.21.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.22.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.conv_pw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.bn1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.bn1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.conv_dw.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.se.conv_reduce.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.se.conv_reduce.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.se.conv_expand.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.se.conv_expand.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.conv_pwl.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.bn3.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for blocks.5.23.bn3.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for conv_head.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for bn2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(
/usr/local/lib/python3.11/dist-packages/torch/nn/modules/module.py:2397: UserWarning: for bn2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)
  warnings.warn(

final_in_features 2152

miew_id_query = miew_id_extractor(ImageDataset(dataset_query.df, root=dataset_query.root, transform=dataset_query.transform))
miew_id_database = miew_id_extractor(ImageDataset(dataset_database.df, root=dataset_database.root, transform=dataset_database.transform))

100%|█████████████████████████████████████████████████████████████████| 7/7 [00:08<00:00,  1.15s/it]
100%|███████████████████████████████████████████████████████████████| 33/33 [00:27<00:00,  1.22it/s]

miew_id_similarity = similarity_function(miew_id_query, miew_id_database)
miew_id_predictions_top_5, miew_id_scores_top_5, _ = top_5_classifier(miew_id_similarity)

miew_id_accuracy_top_1 = np.mean(dataset_query.labels_string == miew_id_predictions_top_5[:, 0])
miew_id_accuracy_top_5 = np.mean(np.any(miew_id_predictions_top_5 == dataset_query.labels_string[:, np.newaxis], axis=1))

miew_id_accuracy_top_1, miew_id_accuracy_top_5

(0.9333333333333333, 0.9333333333333333, 0.9761904761904762)

miew_id_map_score = calculate_map(dataset_query.labels_string, dataset_database.labels_string, miew_id_similarity)
miew_id_map_score

0.7022891902605303

plot_retrieval_results(dataset_query, dataset_database, miew_id_similarity, mode="mistakes", num_results=4, num_queries=5)

Test on Cropped Newts

cropped_mega_query = extractor(CroppingImageDataset(dataset_query.df, root=dataset_query.root, transform=dataset_query.transform))
cropped_mega_database = extractor(CroppingImageDataset(dataset_database.df, root=dataset_database.root, transform=dataset_database.transform))

100%|█████████████████████████████████████████████████████████████████| 7/7 [00:25<00:00,  3.64s/it]
100%|███████████████████████████████████████████████████████████████| 33/33 [01:36<00:00,  2.93s/it]

cropped_mega_similarity = similarity_function(cropped_mega_query, cropped_mega_database)
cropped_mega_predictions_top_5, cropped_mega_scores_top_5, _ = top_5_classifier(cropped_mega_similarity)

cropped_mega_accuracy_top_1 = np.mean(dataset_query.labels_string == cropped_mega_predictions_top_5[:, 0])
cropped_mega_accuracy_top_5 = np.mean(np.any(cropped_mega_predictions_top_5 == dataset_query.labels_string[:, np.newaxis], axis=1))

cropped_mega_accuracy_top_1, cropped_mega_accuracy_top_5

(0.6285714285714286, 0.6285714285714286, 0.7761904761904762)

cropped_mega_map_score = calculate_map(dataset_query.labels_string, dataset_database.labels_string, cropped_mega_similarity)
cropped_mega_map_score

0.3916774944167913

plot_retrieval_results(dataset_query, dataset_database, cropped_mega_similarity, crop_out=True, mode="mistakes", num_results=4, num_queries=5)

Test Cropped out newts on MiewID

miew_id_cropped_query = miew_id_extractor(CroppingImageDataset(dataset_query.df, root=dataset_query.root, transform=dataset_query.transform))
miew_id_cropped_database = miew_id_extractor(CroppingImageDataset(dataset_database.df, root=dataset_database.root, transform=dataset_database.transform))

100%|█████████████████████████████████████████████████████████████████| 7/7 [00:13<00:00,  1.88s/it]
100%|███████████████████████████████████████████████████████████████| 33/33 [00:51<00:00,  1.55s/it]

miew_id_cropped_similarity = similarity_function(miew_id_cropped_query, miew_id_cropped_database)
miew_id_cropped_predictions_top_5, miew_id_cropped_scores_top_5, _ = top_5_classifier(miew_id_cropped_similarity)

miew_id_cropped_accuracy_top_1 = np.mean(dataset_query.labels_string == miew_id_cropped_predictions_top_5[:, 0])
miew_id_cropped_accuracy_top_5 = np.mean(np.any(miew_id_cropped_predictions_top_5 == dataset_query.labels_string[:, np.newaxis], axis=1))

miew_id_cropped_accuracy_top_1, miew_id_cropped_accuracy_top_5

(0.6714285714285714, 0.7571428571428571)

miew_id_cropped_map_score = calculate_map(dataset_query.labels_string, dataset_database.labels_string, miew_id_cropped_similarity)
miew_id_cropped_map_score

0.3966437532673124

plot_retrieval_results(dataset_query, dataset_database, miew_id_cropped_similarity, crop_out=True, mode="mistakes", num_results=4, num_queries=5)

Create a dataframe of the results

results = pd.DataFrame({
    "model": ["MegaDescriptor-L-384", "MiewID", "MegaDescriptor-L-384 (cropped)", "MiewID (cropped)"],
    "accuracy_top_5": [accuracy_top_5, miew_id_accuracy_top_5, cropped_mega_accuracy_top_5, miew_id_cropped_accuracy_top_5],
    "accuracy_top_1": [accuracy_top_1, miew_id_accuracy_top_1, cropped_mega_accuracy_top_1, miew_id_cropped_accuracy_top_1],
    "map_score": [map_score, miew_id_map_score, cropped_mega_map_score, miew_id_cropped_map_score]
})

results

	model	accuracy_top_5	accuracy_top_1	map_score
0	MegaDescriptor-L-384	0.971429	0.852381	0.620877
1	MiewID	0.976190	0.933333	0.702289
2	MegaDescriptor-L-384 (cropped)	0.776190	0.628571	0.391677
3	MiewID (cropped)	0.757143	0.671429	0.396644

results.set_index("model").transpose().plot.bar(figsize=(10, 5), rot=0)

import nbdev
nbdev.export.nbdev_export()